Global Global Economic Prospects Economic...

The eruption of the worldwide financial crisis has radically recast prospects for the world economy. Global Economic Prospects 2009: Commodities at the Crossroads analyzes the implications of

the crisis for low- and middle-income countries, including an in-depth look at long-term prospects for global commodity markets and the policies of both commodity producing and consuming nations.

Developing countries face sharply higher borrowing costs and reduced access to capital. This will cut into their capacity to finance investment spending—ending a five-year stretch of developing-country growth in excess of 6 percent annually. The looming recession presents new risks, coming as it does on the heels of the recent food and fuel crisis.

Commodity markets, meanwhile, are at a crossroads. Following decades of low prices and weak investment in supply capacity, commodity prices first spiked—spurred on by five years of very fast developing-country growth—and have now plummeted in response to the financial crisis.

In the longer run, commodities are not expected to be in short supply. Prices should be higher than they were in the 1990s but much lower than in the recent past. These higher prices should provide producers with sufficient incentive to discover new supplies, improve output from existing resources, and promote greater conservation and substitution with more abundant alternatives. At the same time, slower population growth will ease the pace at which commodity demand grows. Policies to limit carbon emissions and boost agricultural investment, along with the dissemination of efficient techniques, should also contribute to this long-term outcome.

This year’s Global Economic Prospects also looks at government responses to the recent price boom. Producing-country governments have saved more of their windfall revenues, and are therefore less likely to be forced to cut into spending now that prices have declined. The spike in food prices tipped more people into poverty, which led governments to expand social assistance programs. These programs need to be better targeted to the needs of the very poor so that governments can respond effectively the next time there is a crisis.

For additional information, please visit www.worldbank.org/prospects. An online companion to the prospects section of this report, including access to additional data and analysis not reported here, is also available at www.worldbank.org/globaloutlook.

“While developing countries entered this tumultuous

period with much improved fundamentals, this crisis is

expected to test severely both them and the international

financial system. In the longer run, even after developing-

country growth recovers, commodity supply should keep

pace with demand, but policy will need to foster conservation

efforts and technological progress. In particular, if poor

countries are to maintain domestic food self-sufficiency,

governments will need to strengthen investment in rural

infrastructure, agricultural research, and technological

outreach.”

—Justin Yifu LinSenior Vice President and

Chief EconomistThe World Bank

2009Commodities at the Crossroads 2009

Global Economic Prospects


SKU 17799

ISBN 978-0-8213-7799-4

Global E

conomic Prospects

2009

10363_Cover 1 11/29/08 7:15:41 AM

GlobalEconomicProspects

10363_Pg_i-xvi:10363_Pg_i-xvi 11/29/08 6:52 AM Page i

2009

GlobalEconomicProspectsCommodities at the Crossroads

10363_Pg_i-xvi:10363_Pg_i-xvi 11/29/08 6:52 AM Page iii

The cutoff date for the data used in this report was November 20, 2008. Dollars are currentU.S. dollars unless otherwise indicated.

© 2009 The International Bank for Reconstruction and Development / The World Bank1818 H Street NWWashington DC 20433Telephone: 202-473-1000Internet: www.worldbank.orgE-mail: [email protected]

All rights reserved

1 2 3 4 12 11 10 09

This volume is a product of the staff of the International Bank for Reconstruction andDevelopment / The World Bank. The findings, interpretations, and conclusions expressed inthis volume do not necessarily reflect the views of the Executive Directors of The World Bankor the governments they represent.

The World Bank does not guarantee the accuracy of the data included in this work. Theboundaries, colors, denominations, and other information shown on any map in this work donot imply any judgement on the part of The World Bank concerning the legal status of anyterritory or the endorsement or acceptance of such boundaries.

Rights and PermissionsThe material in this publication is copyrighted. Copying and/or transmitting portions or all ofthis work without permission may be a violation of applicable law. The International Bank forReconstruction and Development / The World Bank encourages dissemination of its work andwill normally grant permission to reproduce portions of the work promptly.

For permission to photocopy or reprint any part of this work, please send a request withcomplete information to the Copyright Clearance Center Inc., 222 Rosewood Drive, Danvers,MA 01923, USA; telephone: 978-750-8400; fax: 978-750-4470; Internet: www.copyright.com.

All other queries on rights and licenses, including subsidiary rights, should be addressed to theOffice of the Publisher, The World Bank, 1818 H Street NW, Washington, DC 20433, USA;fax: 202-522-2422; e-mail: [email protected].

ISBN: 978-0-8213-7799-4eISBN: 978-0-8213-7801-4DOI: 10.1596/978-0-8213-7799-4

ISSN: 1014-8906

Cover photos: Oil platform worker in Urucu, Brazil by Hervé Collart/Corbis (left); Molten steelbeing poured in Tangshan, China by Yang Liu/Corbis (top right); Farmers in Kenya by CurtCarnemark/The World Bank (bottom right)Cover design: Critical Stages

10363_Pg_i-xvi:10363_Pg_i-xvi 11/29/08 6:52 AM Page iv

Foreword xi

Acknowledgments xiii

Abbreviations xv

Overview 1

Chapter 1 Prospects for the Global Economy 15Financial markets 19Outlook for high-income OECD countries 24Outlook for the developing countries 27World trade 36Commodity markets 39Key risks and uncertainties 45Long-term prospects and poverty forecast 46

Chapter 2 The Commodity Boom: Longer-Term Prospects 51Characteristics of the current commodity price boom 53The roots of the boom in commodity prices 57Long-term demand prospects 64Long-term supply prospects 74Projections 85Conclusions 89

Chapter 3 Dealing with Changing Commodity Prices 95Commodity dependence and growth 98Managing primary commodity booms 102Poverty impacts of higher commodity prices 113Dealing with high food and fuel prices 121The international response to high commodity prices 127Conclusions 131Technical Annex: Sensitivity Analysis 132

v

Contents

10363_Pg_i-xvi:10363_Pg_i-xvi 11/29/08 6:52 AM Page v

Appendix: Regional Economic Prospects 141East Asia and the Pacific 141Europe and Central Asia 147Latin America and the Caribbean 153Middle East and North Africa 159South Asia 166Sub-Saharan Africa 171

FiguresO.1 The recent commodity boom was the largest and longest of any boom

since 1900 4O.2 Real commodity prices in local currency units increased by between 75 and

150 percent but have fallen since 4O.3 Slower growth should ease commodity demand 5O.4 Technological progress has reduced the quantity of commodities used per unit

of GDP 5O.5 Oil prices are having a direct impact on food prices 8O.6 On average, poor countries are dependent on commodities but relatively

resource poor 9O.7 Primary commodity exporters are exhibiting fewer signs of the behaviors linked to the

“resource curse” 10O.8 Exchange rates, inflation, and government expenditures in new versus established oil

exporters, 2001–06 101.1 GDP growth 181.2 Emerging market equities are hit hard as turbulence evolves to crisis 221.3 Emerging-market bond spreads widen, especially for corporates 221.4 Private debt and equity flows decline by a third in 2008 231.5 Change in GDP in the United States, Europe, and Japan 251.6 The contribution of U.S. domestic demand to GDP growth 251.7 U.S. household wealth falls sharply in the last quarters 261.8 GDP to decline across the OECD 261.9 East Asian countries show steep falloff in output growth 271.10 Output growth in Latin America, South Asia, and Europe and

Central Asia is fading 281.11 Investment was the driving force for growth in developing countries 281.12 Developing-country GDP growth is expected to fall below 5 percent in 2009 301.13 Headline inflation is easing across industrial countries 301.14 Inflation in emerging markets surged on higher food and energy prices 301.15 Key developments in 2008 for East Asia and the Pacific 331.16 Sovereign bond spreads widen across Europe and Central Asia 331.17 In Latin America and the Caribbean, current accounts of largest

economies diverge 341.18 Oil revenues, recovery from drought underpin growth in the Middle East and North

Africa in 2008 351.19 South Asian production slips in the last months 361.20 In Sub-Saharan Africa, primary commodity exports increased as prices surged 361.21 World trade is expected to decline in 2009 for the first time since 1982 37

C O N T E N T S

vi

10363_Pg_i-xvi:10363_Pg_i-xvi 11/29/08 6:52 AM Page vi

1.22 Decline in high-income import growth affects developing-country exports 371.23 Developing-country exports have been strong, even outside China 371.24 Developing-country current account surpluses to wane after 2008 381.25 Current account balances for commodity-exporting and -importing developing-country

groups (excluding China) 391.26 Almost all currencies have depreciated against the dollar 391.27 Commodity prices surged before retreating in the second half of 2008 401.28 Crude oil prices correct sharply after unprecedented run-up 401.29 Grains prices show sharp declines from recent peaks 431.30 Most vulnerable countries will benefit from the decline in grains and oil prices 451.31 First-round income impact of lower commodity prices will be positive in more than

half of developing countries 451.32 Revised poverty estimates following from new price survey 492.1 The recent commodity boom was the largest and longest of any

boom since 1900 552.2 The real local currency price of commodities rose much less than the real

dollar price 562.3 Oil and metal prices led this boom, with food prices rising only much later 562.4 Global growth lasted longer and was stronger during the recent commodity boom than

in earlier ones 572.5 Dormant capacity helped keep oil prices low in the 1990s 592.6 Real spending by major American multinational oil companies declined by 60 percent

in the 1980s 592.7 Global metal demand also fell during the transition 592.8 Real food prices were broadly stable in developing countries until mid-2007 612.9 Most of the decline in global grain stocks reflects lower stocks in China 632.10 Outside of China, only wheat stocks are unusually low 632.11 Demand for most commodities has grown less rapidly than GDP but more rapidly

than population 642.12 The quantity of most commodities used per unit of GDP was declining

until recently 652.13 Metal intensities have declined steadily in high-income countries but have reversed in

China since 1993 692.14 Metal intensities in China are much higher than elsewhere 702.15 Weaker population growth should slow demand for food 712.16 Per capita grain demand tends to stop rising when income reaches

around $5,000 722.17 Demand for edible oils grew much faster than population in Asia 722.18 Food crop prices have become sensitive to oil prices 732.19 Output of virtually all commodities has increased since 1965 742.20 Almost all of the additional oil supply since the 1970s has come from

nontraditional sources 752.21 Rather than declining, known oil reserves keep rising 752.22 Gas reserves are almost as large as oil reserves 762.23 Agricultural productivity has been rising rapidly over the past 20 years 802.24 For key crops, most of the increase in output was due to increased yield, not increased

area planted 812.25 Yield growth has decelerated recently 81

C O N T E N T S

vii

10363_Pg_i-xvi:10363_Pg_i-xvi 11/29/08 6:52 AM Page vii

2.26 The stock of unused but potentially arable land is enormous 822.27 Developing countries spend less on agricultural R&D than high-income countries 833.1 More-diversified developing countries grew more rapidly from 1980 to 2006 983.2 Poorer countries are more dependent on nonfuel primary commodities 983.3 On average, poor countries are dependent on commodities but relatively

resource poor 993.4 Government spending by primary commodity exporters responded less to export

booms in this decade than in the 1980s 1033.5 Public expenditures in Sub-Saharan Africa grew much less quickly in the 2000s than

in the 1980s 1033.6 Oil-exporting countries with large reserves spent a smaller portion of their revenue

from the recent boom in oil prices, 2000–06 1053.7 Imports and current account positions suggest more savings from commodity revenues

by oil exporters than by nonfuel commodity exporters 1063.8 Primary commodity exporters limited the real appreciation of their currencies during

the recent boom 1073.9 Many oil exporters are suffering significantly higher inflation 1073.10 New oil exporters are experiencing more macroeconomic volatility than established

producers 1083.11 Commercial bank lending to commodity-dependent economies in Sub-Saharan

Africa is rising 1083.12 Corruption is highest among fuel exporters, although the difference has

narrowed 1093.13 The increased grain bill could exceed 5 percent of GDP in more than

20 countries 1163.14 Real food prices in developing countries rose less than prices of internationally traded

foods 1193.15 Developing countries have responded to rising food prices with a variety

of policies 1223.16 Countries have tended to expand cash transfers and school feeding programs when

responding to higher food prices 1223.17 After India banned rice exports, international prices rose 124A1 GDP growth eases in several East Asian economies 142A2 Export growth in East Asia turns down on falling OECD demand 143A3 Exchange rates decline sharply as carry trades unwind 144A4 Gross capital inflows to East Asia contracted 40 percent in 2008 144A5 Deepening global financial crisis affects Europe and Central Asia 148A6 Core inflation is rising in several countries of Europe and Central Asia 152A7 Contributions to GDP growth in Latin America and the Caribbean 153A8 Bond spreads have increased sharply for many Latin American countries 155A9 Exchange rates in Latin America and the Caribbean fell sharply against the dollar in

late 2008 155A10 Inflation still high in Latin America and the Caribbean despite sharp falloff in food

and fuel prices 156A11 Current account positions in the Middle East and North Africa set to shift

dramatically 161A12 Markets in the Middle East and North Africa are hard hit by

financial crisis 161

C O N T E N T S

viii

10363_Pg_i-xvi:10363_Pg_i-xvi 11/29/08 6:52 AM Page viii

A13 Inflation rises across the Middle East and North Africa 162A14 Notable slowing of growth across the Middle East and North Africa in 2009 166A15 Key international finance links in South Asia, 2007 169A16 Bond spreads for African countries jumped after mid-September 173A17 African headline inflation jumps as food prices skyrocket 174A18 Economic growth to slow in Sub-Saharan Africa 178

TablesO.1 Food price hikes and consumption shares vary by region 11O.2 Higher food prices have increased both the incidence and severity of poverty

worldwide 121.1 The global outlook in summary 171.2 High-income OECD countries: growth and related indicators 251.3 Developing regions: growth and related indicators 291.4 Forecast of commodity prices 401.5 Poverty in developing countries by region, selected years 472.1 Principal characteristics of major commodity booms 552.2 Comovement among major commodity prices, 1960–2007 602.3 Impact of a 10 percent increase in incomes on commodity demand 652.4 Modern goods make less intensive use of commodities 652.5 Fundamental economic factors drive future commodity demand 662.6 Energy demand is projected to slow in the baseline scenario 692.7 Energy demand could decline further under more aggressive climate

change policies 692.8 Developing countries will account for most of the projected demand for various foods,

2000–30 722.9 Historically, estimates of oil reserves have kept pace with production 762.10 Increased investment has stabilized reserve-to-production ratios for some

commodities 772.11 Oil’s share in global energy supply is projected to decline 782.12 Potential gains from extending the green revolution remain large 822.13 With some exceptions, yield growth for key agricultural commodities has been highest

in South and East Asia 832.14 Agricultural sector simulation results, 2005–30 862.15 Energy sector simulation results, 2005–30 873.1 Non-oil or resource-rich countries have higher per capita incomes than resource-

dependent countries, 2006 993.2 Ratios of reserves to production vary greatly among oil exporters 1053.3 Assets in sovereign wealth funds grow in commodity-exporting countries 1073.4 Country studies suggest that high oil prices have large poverty impacts 1153.5 Higher food prices raise poverty more in urban areas than in rural areas 1173.6 Observed real price shocks and food shares of consumption vary across developing

regions 1193.7 Poverty effects of the changes in relative food prices 1203.8 Fiscal costs of selected antipoverty measures vary widely 1223.9 Increasing rice self-sufficiency can be more costly than relying on imports 1243A.1 Sensitivity analysis 132A1 East Asia and Pacific forecast summary 142

C O N T E N T S

ix

10363_Pg_i-xvi:10363_Pg_i-xvi 11/29/08 6:52 AM Page ix

A2 East Asia and Pacific country forecasts 146A3 Europe and Central Asia forecast summary 148A4 Europe and Central Asia country forecasts 151A5 Latin America and the Caribbean forecast summary 154A6 Latin America and the Caribbean country forecasts 158A7 Middle East and North Africa forecast summary 160A8 Middle East and North Africa country forecasts 164A9 South Asia forecast summary 168A10 South Asia country forecasts 170A11 Sub-Saharan Africa forecast summary 172A12 Sub-Saharan Africa country forecasts 176

Boxes1.1 Chronology of recent developments in the financial crisis 201.2 Commodity prices and inflation in developing countries 311.3 Impact of commodity prices on external balances and capital flows 411.4 The impact of the new price survey on poverty estimates 482.1 Commodity price cycles 542.2 Developing-country growth and global commodity demand in the recent past 582.3 The historical link between crude oil and other commodity prices 622.4 Alternative fuels for transportation 682.5 Understanding the rise in Chinese metal intensities 702.6 Declining costs of resource extraction 752.7 The rise of biofuel production 792.8 State-owned firms and output efficiency 802.9 Genetically modified crops—the next green revolution? 843.1 The impact of severe shocks on economic progress 1003.2 Efforts to capture a larger share of windfall commodity revenues 1043.3 Combating the corrupting influence of high commodity revenues 1093.4 Successful sovereign wealth funds 1103.5 National and international marketing strategies 1113.6 Malawi government hedging of maize price and supply risks, 2005–08 1133.7 Critical assumptions underlying the estimation of the poverty impact

of food price increases 1183.8 Conditional cash transfers are most effective in getting money to the poor 1263.9 Removing fuel subsidies in Ghana 1273.10 The international response to rising food prices 128

C O N T E N T S

x

10363_Pg_i-xvi:10363_Pg_i-xvi 11/29/08 6:52 AM Page x

xi

EACH YEAR, Global Economic Prospectsexplores critical “here and now” eco-nomic developments relevant to low-

and middle-income countries. Past editionshave examined the sustainability of developing-country growth over the long term, importancefor developing countries of international andregional trade liberalization, and migrationand remittances. Last year’s report looked atthe pace and determinants of technologicaldiffusion in developing countries.

This year’s Global Economic Prospectsfinds the global economy at a crossroads,transitioning from a sustained period of verystrong developing country–led growth to oneof substantial uncertainty as a financial crisisrooted in high-income countries has shaken fi-nancial markets worldwide. Commodity mar-kets too are at a crossroads with the very highprices of 2007 and early 2008 having fallen bymore than half in many instances.

Great uncertainty surrounds the implica-tions of this crisis for developing countries. Ini-tially, the repercussions for developing coun-tries of the financial turmoil that characterized2007 and the first half of 2008 were limited.However, since September 2008, the intensifi-cation of the banking crisis, the collapse of sev-eral global financial players, and the sharp in-crease in emerging market bond spreads havedramatically altered the outlook for develop-ing countries. These events constitute the kindof disorderly adjustment that has been dis-cussed in previous reports as a risk. Material-ized, it implies a sharp slowdown for develop-

ing countries and the possibility that seriouscrises may emerge.

While the measure of that slowdown and itsnear-term implications for growth and incomesare important, governments in developingcountries also need to be mindful of the longer-term implications of their policy response.Thus, while countercyclical policy may help re-duce the short-term costs of the slowdown,care must be exercised to react prudently so asnot to endanger longer-term fiscal sustainabil-ity and growth prospects. For as serious as thecoming slowdown may be, developing-countrygrowth is expected to recover after the crisis isover.

Commodity markets have seen spectacularswings over the past 24 months as enormoustensions first built up and were then released.The extended and sharp rise in commodityprices prompted concerns that the world wastransitioning into a new phase of commodityscarcity—a concern that the recent dramaticdrop in commodity prices has only partiallyalleviated. Long-term supply and demandprospects for commodities suggest that whilecommodity prices are likely to be higher thanthey were during the 1990s and early 2000s(when they were depressed by excess supply),the recent peaks that have been observed areunlikely to be the new norms. Over the longrun, demand for commodities is not expectedto outstrip supply. Even though per capita in-comes in developing countries are expected tocontinue rising rapidly, population growth isslowing and with it global GDP growth. As a

Foreword

10363_Pg_i-xvi:10363_Pg_i-xvi 11/29/08 6:52 AM Page xi

F O R E W O R D

xii

result, the pace at which commodity demandexpands should also ease. Assuming that effi-ciency with which commodities are both em-ployed and produced continues to improve asit has done over the past few decades, supplyshould keep pace with demand.

However, policy will need to be supportiveif such a positive result is to materialize. Inparticular, agricultural yields have declined inrecent years. Unless governments in develop-ing countries and aid agencies take concretesteps to increase investment in rural infra-structure, agricultural research and develop-ment, and agricultural extension services, it ispossible that global agricultural productivitygrowth will slow. Higher food prices wouldfollow and many countries that are now self-sufficient in food (notably those that still havefast growing populations) would become largenet importers of food. On the energy side,policies to combat carbon emissions wouldhelp slow the depletion of hydrocarbon re-sources, by speeding both demand-side andsupply-side substitution toward cleaner energysources. If successful in slowing global warm-ing, these could also help prevent the verylarge agricultural productivity losses predictedby some in the second half of this century.

The recent boom in commodity prices haschallenged policy makers in both producingand consuming countries. Encouragingly,commodity producers appear to have man-aged their windfall revenues more prudentlythan in the past. Instead of expanding spend-ing programs in line with increased revenues,many have saved a much larger share of theserevenues—reducing the likelihood that theywill need to cut back spending in a procyclicalmanner now that commodity prices (and

global growth) have declined. However,countries with new-found commodity wealthor newly independent commodity-rich stateshave not shown similar restraint and mayencounter more difficulties during the currentdownturn.

Higher food prices are estimated to haveincreased global poverty by some 130–155million people. Most countries responded tothe food crisis by expanding existing socialsafety net programs, which made good sensegiven the profound and long-term conse-quences that increased malnutrition couldgenerate. However, in many instances the re-sponse was poorly targeted and expensive.Now that food prices are declining, countriesneed to take steps to revamp their social wel-fare systems so that they are better targetedand that the next time a similar crisis comesalong, additional spending will be more effec-tive in limiting poverty impacts.

At the international level, steps need to betaken to prevent producing countries from ex-acerbating shortfalls by introducing exportbans or by withholding stocks from the globalmarket. An international scheme to share in-formation about private and public stocks andcoordinate their management during times ofcrisis is worth pursuing. Similarly, funding forinternational food aid programs should bemade more predictable, and agencies shouldbe endowed with a line of credit that wouldallow them to respond rapidly to future foodemergencies in a way they cannot at present.

Justin Yifu LinSenior Vice President and


10363_Pg_i-xvi:10363_Pg_i-xvi 11/29/08 6:52 AM Page xii

THIS REPORT WAS produced by staff from the World Bank’s Development Prospects Group.Andrew Burns was the lead author and manager of the report, with direction from UriDadush. The principal authors of the report were John Baffes, Donald Mitchell, Elliot

(Mick) Riordan, Shane Streifel, Hans Timmer, and William Shaw. The report was produced underthe general guidance of Justin Yifu Lin.

Several people contributed substantively to chapter 1. Elliott (Mick) Riordan and Hans Timmerwere its main authors. The Global Trends Team, under the leadership of Hans Timmer, wasresponsible for the projections. The projections and regional write-ups were produced by TengJiang, Annette De Kleine, Elliot (Mick) Riordan, Cristina Savescu, and Ani Silwal, in coordina-tion with country teams and the offices of the regional Chief Economists and PREM directorsincluding Luca Barbone, Marcello Guigale, August Kouame, Ernesto May, Vikram Nehru, RitvaReinikka, Sudir Shetty, and Augusto de la Torre. The short-term commodity price forecasts wereproduced by John Baffes, Betty Dow, Donald Mitchell, and Shane Streifel. The remittances fore-casts were produced by Sanket Mohapatra, while Shaohua Chen from the Development ResearchGroup and Dominique van der Mensbrugghe generated the long-term poverty forecast.

John Baffes, Andrew Burns, William Shaw, and Shane Streifel were the main authors of Chap-ter 2, with written contributions from Donald Mitchell, Marian Radetzki, Varun Kshirsagar,Denis Medvedev, and Dominique van der Mensbrugghe. Chapter 3 was written by DonaldMitchell and William Shaw with written contributions from Ataman Aksoy, Margaret Grosh, andRafael de Hoyos Navarro. Both Chapters 2 and 3 benefited from the expert research assistance ofVarun Kshirsagar and Teng Jiang.

The accompanying online publication, Prospects for the Global Economy (PGE), was producedby a team led by Cristina Savescu and comprised of Sarah Crow, Betty Dow, Kathy Rollins, AniSilwal, Cybele Arnaud, and Ying Yu, with technical support from Gauresh Rajadhyaksha. Thetranslation process was coordinated by Jorge del Rosario (French and Spanish) and Li Li(Chinese). A companion pamphlet highlighting the main messages of the commodities section ofthe report was prepared by Kavita Watsa and Roula Yazigi.

Several reviewers offered extensive advice and comments throughout the conceptualizationand writing stages. These included Charles Blitzer, Christopher Delgado, Sebastien Dessus,Christopher Gilbert Sheldon, Santiago Herrera, Justin Yifu Lin, William Maloney, William Martin,Celestin Monga, Vikram Nehru, Marian Radetzki, Ana Revenga, Alexander Sarris, KatherineSierra, Luiz Pereira da Silva, Claudia Paz Sepulveda, and Daniel Villar.

Marty Gottron edited the report. Hazel Macadangdang managed the publication processand Merrell Tuck-Primdahl managed the dissemination activities. Book design, editing, andproduction were coordinated by Aziz Gökdemir of the World Bank’s Office of the Publisher,along with Stephen McGroarty, Denise Bergeron, Andrés Meneses, and Susan Graham.

xiii

Acknowledgments

10363_Pg_i-xvi:10363_Pg_i-xvi 11/29/08 6:52 AM Page xiii

xv

Abbreviations

ASEAN Association of South Eastern Asian Nations

CEE Central and Eastern European countries

CIS Commonwealth of Independent States

CPI consumer price index

EMBI Emerging Markets Bond Index

EMBIG Emerging Markets Bond Index Global

EU European Union

FDI foreign direct investment

FSU former Soviet Union

GCC Gulf Cooperation Council

GDP gross domestic product

GFRP Global Food Crisis Response Program

GIDD Global Income Distribution Dynamics Model

IDA International Development Association (World Bank)

IEA International Energy Agency

IMF International Monetary Fund

IPO initial public equity offering

LSMS Living Standards Measurement Survey

MSCI Morgan-Stanley Composite Index

OECD Organisation for Economic Co-operation and Development

OPEC Organization of Petroleum Exporting Countries

PPP purchasing power parity

PCSC Programme Complémentaire de Soutien à la Croissance

RIGA Rural Income-Generating Activities

saar seasonally adjusted annualized rate

toe tonne of oil equivalent

UAE United Arab Emirates

WFP World Food Programme

10363_Pg_i-xvi:10363_Pg_i-xvi 11/29/08 6:52 AM Page xv

1

Overview

The release of this year’s Global EconomicProspects finds the world economy at a

crossroads. Markets all over the world are en-gulfed in a global economic crisis, with stockmarkets sharply down and volatile, almost allcurrencies having depreciated substantiallyagainst the dollar, and risk premiums on a widerange of debt having increased by 600 or morebasis points. Commodity markets too haveturned a corner. Following several years of in-crease, prices have plummeted, and althoughwell above their 1990s levels, they have givenup most of the increases of the past 24 months.

Chapter 1 of this report examines themedium-term implications of this crisis fordeveloping-country growth, inflation, andworld trade. Chapter 2 looks at longer-termsupply and demand prospects in commoditymarkets. It takes into account the long-termgrowth prospects of developing countriesand their rising share in world GDP (grossdomestic product), the declining quality ofnew pools of resources, and the influenceof technology on both demand and supply.Finally, chapter 3 reports on the poverty im-pacts of high commodity prices and examinesthe effectiveness of policies in both produc-ing and consuming countries in dealing withthe challenges posed by periodic bouts ofhigh commodity prices.

This report does not deal with water, fish,or timber, all commodities of critical impor-tance to developing countries and the globebut which fall outside the scope of this reporteither because of their public-goods characteror, in the case of timber, because of its treat-ment in a recent report (World Bank 2007).

The global financial crisis threatens short-term prospects in developing countriesThe banking crisis that erupted in September2008, following more than a year of less acutefinancial turmoil, has substantially reinforcedthe cyclical downturn that was already underway. Following the insolvency of a large num-ber of banks and financial institutions in theUnited States, Europe, and the developingworld, financial conditions have becomemuch tighter, capital flows to developingcountries have dried up, and huge amounts ofmarket capitalization have evaporated.

The crisis began in high-income countries,but developing countries have been caught upin its wake. As of mid-November, developing-country equity markets had given up almost allof their gains since the beginning of 2008 andinitial public offerings had disappeared. Riskpremiums, which had risen to more than 800basis points on sovereign bonds and 1,000 oncommercial debt, have declined but remainedwell above 600 basis points in every develop-ing region. As corporate bonds had been oneof the most important source of developing-country finance, these developments suggestthat a sharp slowing in developing-countryinvestment growth is to be expected. Banklending and foreign direct investment inflowswere also down, but less dramatically. The in-creased volatility and losses emanating fromthe banking sector have caused investorsworldwide to sell stocks and increase theirholdings of less risky assets, notably U.S. trea-suries. As a result, the currencies of virtuallyevery developing country in the world has de-preciated vis-à-vis the dollar.

10363_Pg1-14:10363_Pg1-14 11/29/08 6:56 AM Page 1

Following a series of efforts by central banksand governments to resolve the growing crisisthrough liquidity injections and various adhoc measures, policy makers have now actedforcefully to restore confidence in the interna-tional banking system, including the partialnationalization of nine banks and trillions ofdollars in rescue plans introduced by govern-ments in the United States and Europe andrecent multilateral meetings to address weak-nesses in the global financial architecture. Atthe time of this writing (November 20, 2008),it is too soon to judge the effectiveness of thesemeasures in restoring confidence in the bank-ing system. However, they do constitute thekind of forceful and credible action that hasbeen needed, and interbank lending rates havefallen substantially and although they remainvolatile, stock and bond markets have greetedthese measures favorably.

Notwithstanding these steps, growthprospects for both high-income and develop-ing countries have deteriorated substantially,and the possibility of a very deep global reces-sion cannot be ruled out.

Even before the emergence of a full-blownfinancial crisis in September 2008, globalgrowth showed significant weakening. Eco-nomic growth slowed sharply in Europe andJapan and in many developing countries inthe second quarter of 2008. In the UnitedStates, the continued disruption in financialmarkets and the fall in housing prices causeddomestic demand to fall in 6 of the past 12quarters. However, strong export growth—driven in part by developing-country importdemand—spared the U.S. economy from re-cession until recently when its GDP declined0.5 percent in the third quarter of 2008. Indeveloping countries, overall GDP growthalso remained robust in the first half of theyear. However, slower growth in high-incomecountries and the weakening of capital inflows,in combination with commodity-price-induced losses in real income, generated asharp deceleration in industrial production,investment, and international trade beginningin the third quarter.

At the same time, rising commodity pricesand tight capacity in many countries (followingyears of very fast growth fueled by ample liq-uidity) caused both headline and core inflationto pick up throughout the world, with headlineinflation rising by some 5 percentage pointsamong developing countries. Weaker growthand falling commodity prices have alreadycaused inflationary pressures to ease in somecountries. However, the significant losses in realincome endured by many people in developingcountries and the still overheated state of someof their economies could generate second-round price increases that either push inflationhigher or stabilize expectations at high levels.

The combination of a relatively strongfirst half and a much weaker second half isexpected to cause GDP growth to slow to 1.3percent in high-income countries and to 6.3 per-cent in developing countries in 2008. The slow-down is projected to intensify in 2009 becausemost of the real-economy side effects of thebanking crisis will be felt in the final months of2008 and the first two quarters of 2009.

The main mechanism for the slowdown inboth developing and high-income countrieswill be through investment, which for 2009 isexpected to decline 3.1 percent in high-incomecountries. In developing countries, investmentgrowth is projected to slow sharply to 3.4 per-cent in 2009 from more than 13 percent in2007. Because low-income countries have lessaccess to international capital markets, theslowdown will affect them mainly through in-direct mechanisms, including slower globalgrowth, lower commodity prices, slackeningremittance receipts, and partial scaleback inaid flows.

Overall global GDP growth is projected todecline to 0.9 percent in 2009, with develop-ing economies expanding by 4.5 percent—wellbelow the 7.9 percent growth rate recorded in2007. International trade should deceleratesharply, with global export volumes decliningfor the first time since 1982. As a result, bothcommodity prices and inflation are projectedto ease, with oil prices averaging about $75 abarrel in 2009 and food prices projected to

G L O B A L E C O N O M I C P R O S P E C T S 2 0 0 9

2

10363_Pg1-14:10363_Pg1-14 11/29/08 6:56 AM Page 2

decline by about 23 percent compared withtheir average for 2008.

This financial crisis and the expectedabrupt slowing of global growth come at amoment when developing countries consid-ered as a whole are more vulnerable than theyhave been in the recent past. Higher commod-ity prices have raised the current accountdeficits of many oil-importing countries toworrisome levels (they exceed 10 percent ofGDP in about one-third of developing coun-tries), and after having increased substantially,the international reserves of oil-exportingdeveloping countries are now declining as ashare of their imports. Moreover, inflation ishigh, and fiscal positions have deterioratedboth for cyclical reasons and because govern-ment spending has increased to alleviate theburden of higher commodity prices.

Thus, even in the baseline scenario, wherethe rapid equity declines of September andOctober are assumed to end and where creditbegins to thaw as recent policy actions im-prove financial market confidence, a numberof developing countries are likely to be sub-jected to substantial strains, possibly includingbank failures and currency crises. In these veryuncertain circumstances, policy makers mustplace a premium on reducing the likelihoodof domestic turmoil, by reacting swiftly andforcefully to emerging difficulties, including, ifnecessary, seeking assistance from the Interna-tional Monetary Fund (IMF).

Uncertainty continues to cloud the outlookWhile this sober outlook represents a likelyoutcome, a wide range of outcomes remainspossible. The financial turmoil could intensifyfurther, sparking a prolonged credit crunchand global recession. A milder downturn isalso possible, if credit conditions do not dete-riorate as much as anticipated in the baseline.

At the time of this writing, the possibilitythat the situation in high-income countries willdeteriorate substantially cannot be ruled out.Should credit markets fail to respond to the ro-bust policy interventions taken so far, the con-sequences for developing countries could be

O V E R V I E W

3

very serious. Global financing conditionswould deteriorate rapidly, and apparentlysound domestic financial sectors could findthemselves unable to borrow or unwilling tolend—both in international and domestic mar-kets. Such a scenario would be characterizedby a long and profound recession in high-income countries and substantial disruptionand turmoil, including bank failures and cur-rency crises, in a wide range of developingcountries. Sharply negative growth in a numberof developing countries and all of the attendantrepercussions, including increased poverty andunemployment, would be inevitable.

Although a receding concern, high inflationin developing countries, remains a problem, es-pecially if the financial turmoil is resolved rela-tively quickly. While global growth would stillslow in 2009 under such a scenario, the sub-stantial policy stimulus that has been introducedcould cause growth in both developing and de-veloped countries to surge in 2010, reigniting in-flationary pressures and forcing a subsequenttightening of policy and a second bout of slow-ing growth. Policy in countries that currentlyhave large current account deficits and high in-flation needs to be particularly vigilant. Theseeconomies continue to be vulnerable and in-vestors skittish; under these conditions, theircurrencies are likely to remain particularly sen-sitive to changing market perceptions.

The commodity market boom has come toan endThe sharp rise in commodity prices over thepast five years, like the earlier booms of the lastcentury, was associated with a period of strongeconomic growth (partly fueled by relativelyloose fiscal and monetary policy) and a periodof global uncertainty, and it has generatedsignificant inflationary pressures. This mostrecent boom has been the most marked of thepast century in its magnitude, duration, andthe number of commodity groups whose priceshave increased (figure O.1).

The strength and duration of the boommainly reflected the resilience of GDP growthbetween 2003 and 2008.

10363_Pg1-14:10363_Pg1-14 11/29/08 6:56 AM Page 3

In the oil and metals sector, the supply pres-sures that built up over the past five years andwhich drove prices to record heights stemmedmainly from slow-growing supply capacity.That slow growing supply capacity resultedbecause for much of the 1990s rising demandin the rest of the world was met by the slowreabsorption of idle capacity created follow-ing the 1980 oil shock and the collapse ofdemand in the former Soviet bloc when theseformerly communist countries began to allo-cate resources according to market signals. Asa result of this idle capacity, prices remainedlow in the oil and metals sectors and firms didnot have the economic incentives to increaseproductive capacity.

Furthermore, because of low prices and be-cause incremental demand was being met by thiscapacity, investment in the oil and metals indus-tries plummeted, and the sectors that suppliedthe inputs necessary for exploration and ex-ploitation atrophied. That in turn created a mis-match between the underlying rate of growth ofsupply capacity and demand.When the spare ca-pacity was exhausted in the early 2000s, supplywas no longer able to keep pace with strength-ening demand, and prices began to rise.

The story in agricultural markets is different.Food-based demand for agricultural crops hasbeen relatively stable. However, diversion offood crops toward biofuel production hasincreased sharply. Between 2003 and 2007,

two-thirds of the global increase in maize pro-duction went to biofuels. Although the initialimpact was confined to the maize market, asfarmers switched land away from wheat andsoybean production to grow maize, the price ofthese commodities also began to rise. Higher oiland fertilizer prices also increased food produc-tion costs, especially in high-income countrieswhere they can account for as much as 30 per-cent of overall costs. This factor, plus biofuel de-mand for grains, has made the price for theseproducts much more sensitive to changes in oilprices. Finally, a series of poor wheat crops inAustralia compounded the situation, drivingdown stocks and contributing to the price rise.

In addition to these fundamental drivers,agricultural prices have been influenced both byincreased investor interest in these commoditiesas an asset class and by government policies, in-cluding the decision by several countries to im-pose export bans. All of these factors are drivenby forward-looking expectations and may haveexacerbated both the upward rise in prices dur-ing 2007–08 and their more recent decline.

Commodity prices are declining inresponse to slower GDP growthLike earlier commodity booms, this one hascome to an end. Prices in all commodity marketshave fallen sharply since July 2008 (figure O.2),


4

50

Jan.2000

Jan.2001

Jan.2002

Jan.2003

Jan.2004

Jan.2005

Jan.2006

Jan.2007

Jan.2008

100

150

200

250

300

Figure O.2 Real commodity prices in localcurrency units increased by between75 and 150 percent but have fallen since

Energy

Food

Metals and minerals

Source: World Bank.

Real local currency commodity price indexes, CPI-deflated(Jan. 2000 5 100)

80

130

180

230

280

330

380

1900

Source: Grilli and Yang (1988) for 1900 to 1947; World Bank for1948 to 2008.

1920 1940 1960 1980 2000

Real non-energy commodity prices, index (1977–79 5100)

1917 (just prior to WW I)

1951 (postwar rebuilding)

2008 (forecast)

1974 (first oil crisis)

Figure O.1 The recent commodity boom wasthe largest and longest of any boomsince 1900

10363_Pg1-14:10363_Pg1-14 11/29/08 6:56 AM Page 4

reflecting slower GDP growth, increased sup-plies and revised expectations. Because com-modity prices reflect forward-lookingexpectations, the sharp slowing of growth thatis expected over the next year has causedprices to decline rapidly even though the un-derlying supply and demand tensions are littlechanged from just a few months ago whenthese prices were close all-time highs.

Some metals prices have already fallen topre-boom levels and the dollar price of manyinternationally traded foods has fallen back totheir 2006 levels. While much weaker GDPgrowth is projected to cause commodity pricesto ease further in the short run, they shouldnevertheless remain higher than they were dur-ing the 1990s. Real food prices are projected todecline by 26 percent between 2008 and 2010,energy prices to fall by 27 percent, and metalsprices to decline by 32 percent.

In the longer term, growth in the demandfor commodities should easeThe strength, breadth (in terms of the number ofcommodities whose prices have increased), andduration of the current commodity boom haveprompted speculation that the global economyis moving into a new era characterized by rela-tive shortage and permanently higher (and evenpermanently rising) commodity prices.

This outcome does not appear likely. Overthe next two decades, slower populationgrowth andweaker (though still strong) incomegrowth are projected to cause trend global GDPgrowth to ease (figure O.3) and, with it, the de-mand for commodities. As discussed later, theextent to which commodity demand does slowand how easily supply is able to keep pace withdemand will very much depend on the policyenvironment, the pace of technological change,and external factors such as climate change.

Moderating demand for metals dependscritically on increased efficiency in ChinaOver the past 50 years, a combination of con-servation measures, technological change, andchanges in the structure of global GDP (ser-vices tend to be less commodity-intensive than

manufactured goods) has reduced the quantityof metals and energy required to produce aunit of GDP by an average of 0.9 and 0.8 per-cent a year respectively (figure O.4). The foodintensity of GDP has also declined as an in-creasing share of the world’s population hasreached income levels where per person de-mand for basic food commodities is stable.

Beginning in the middle 1990s, the declinein metals intensities began to reverse. That re-versal is explained almost entirely by increas-ing metal intensities in China, which began

O V E R V I E W

5

0

1

2

3

4

5

6

Growth of GDP, annual average (percent)

Figure O.3 Slower population growth shouldresult in weaker GDP and commodity demand

1990s 2000s 2015–30

Contributionto GDPgrowth frompopulationgrowth

Contributionto GDPgrowthfrom percapitaincomes

Source: World Bank LINKAGES model.

Hig

h-in

com

eco

untr

ies Dev

elop

ing

coun

trie

s

Figure O.4 Technological progress hasreduced the quantity of commodities usedper unit of GDP

0.701971 1977 1983 1989 1995 2001

0.75

0.80

0.85

0.90

0.95

1.00

1.05

1.10

Commodity intensity of demand index (1971 5 1.00)

Source: World Bank calculations, using data from the WorldBureau of Metal Statistics, the IEA, and the FAO.

Energy Metals

Metals (excluding China) Food

10363_Pg1-14:10363_Pg1-14 11/29/08 6:56 AM Page 5

in 1995 and grew even more sharply at thebeginning of the 2000s. The uptick in metalsintensities was associated with the investment,manufacturing, and export booms in thatcountry. Currently, metal intensity in China isfour times higher than in developed countriesand twice as high as in other developing coun-tries. China’s metal intensities are expectedto stabilize in coming years and then begin tofall as the country’s very high investment ratedeclines and the transitional shift in globalmanufacturing capacity from high-incomecountries to China slows.

Assuming China’s metal intensity stabilizesand then falls in coming years, global demandfor metals—which has outpaced GDP in recentyears—should first realign itself with GDPgrowth over the next few years and then declinefurther during the next decade, reaching about2.7 percent a year in the period 2015–30.

Future energy demand depends onimproving automobile efficiencyDemand in the energy sector will dependcritically on the pace at which energy effi-ciency continues to improve, especially in thetransport sector. Since 1970 conservation ef-forts and technological progress have reducedenergy demand by 56 percent, compared witha no-change scenario (IEA 2006). With some75 percent of future energy demand expectedto come from the transport sector, especiallyfrom developing countries, the pace of futureenergy demand growth (and its composition)will depend heavily on future efficiency gainsin car technology.

Prospects for such improvements are good,if policy continues to be supportive of bothconservation and efficiency measures. Alreadyexisting technologies—available either in ini-tial rollout phases or as prototypes (flex-fueland hybrid cars, plug-in hybrids, and electricand hydrogen-powered vehicles)—could helpto more than double fuel efficiency. An ambi-tious (and successful) policy to speed the de-velopment and diffusion of these technologiescould see the share of these vehicles rise to90 percent in the high-income world and to

75 percent in developing countries by 2050,substantially reducing private transportation’sdependency on liquid fuels.

In the baseline scenario, demand for oil isexpected to continue rising to around 114 mil-lion barrels a day (mb/d) by 2030 (comparedwith 87 mb/d today). Energy demand is pro-jected to grow somewhat more quickly as coal,natural gas, and non-fossil-fuel energy sourcesincrease their share in total energy supply. Theextent to which this shift occurs will dependimportantly on the policy environment. Amore proactive stance toward restraining car-bon emissions could speed the pace at whichalternative energies become economicallyviable and reduce the expected increase inreliance on coal-powered electrical plants.

Over the next 20 years, supplies ofextracted commodities are likelyto remain ampleThe pace at which the growth in supply capac-ity in the oil and metals sectors catches up todemand will depend on how quickly capacity inthe heavy and specialized equipment and laborsupply sectors can be restored. Years of lowprices and weak investment have reducedcapacity in these sectors, and as a result,delivery times and costs of inputs have morethan quadrupled in many instances. High pricesfor these components are speeding the allevia-tion of these constraints. With the expectedslowing of global GDP growth and lower com-modity prices, investment demand has easedand prices for these specialized investmentgoods are expected to fall further. Nevertheless,deliveries are projected to continue trailingdemand for some time, and prices will remainrelatively high for the next several years.

Over the longer run, the price of extractedcommodities should fall—although they arenot expected to fall to their levels in the1990s. Higher prices than in the past will berequired to ensure that firms continue to in-vest in new capacity.

Although the absolute quantity of fossilfuels and metals in the earth’s crust is declin-ing and the quantity that is extracted each


6

10363_Pg1-14:10363_Pg1-14 11/29/08 6:56 AM Page 6

year is rising, there appears little likelihoodthat the world will run out anytime soon. His-torically, proven reserves of both metals andoil have tended to rise even more rapidly thanproduction, remaining surprisingly constant inthe case of oil at about 40 years of production.In part, that is because measured reserves,rather than being an accurate count of the re-sources remaining in the ground, bear a closerresemblance to the inventory of product thatfirms can readily bring to the market. So longas firms have ample “known reserves” for ex-pected future demand, they have little incen-tive to find more.

As production increases and more knownreserves are brought into service, additionalreserves will likely be discovered. In general,these newer reserves tend to be of lower qual-ity and higher cost than existing ones. How-ever, historically improvements in extractiontechnology have advanced quickly enough tokeep the cost of exploiting new sources stableor even falling, despite increased remotenessand poorer quality. The projected long-termprice of a barrel of oil of $75 (real 2008 dol-lars) is based on the expectation that such aprice will be sufficient to incite additionaloutput from high-cost sources such as theCanadian oil sands.

Even if certain resources do become scarce,ample alternatives exist. For example, if thepace at which new oil reserves are discovereddeclines, the rising price for oil will makealternative sources of energy (including coal,natural gas, nuclear, and renewable alterna-tives) more competitive and induce increasedconservation and technological change. Simu-lations suggest that if oil production fails torise between now and 2030, oil prices mightdouble but most of the energy shortfall wouldbe met by increased coal and natural gas con-sumption—albeit at higher cost.

Food demand will slow with lowerpopulation growth, but biofuels couldexpand crop demand very rapidlyBecause an increasing share of the world’spopulation has reached income levels where

O V E R V I E W

7

demand for most primary food commoditiesno longer rises with income, demand for food isexpected to slow—broadly in line with weakerpopulation growth. However, the potentialrole of biofuel demand for food crops greatlycomplicates the picture. Given today’s tech-nology, maize can be profitably transformedinto ethanol at oil prices in excess of $50 abarrel. Above that price, every percentagepoint increase in the barrel price of oil causesmaize price to rise by 0.9 percent (figure O.5),which means the maize market is effectivelytied to the oil market (this relationship is notstatistically significant when oil is below $50 abarrel). Moreover, because farmers have re-sponded to high maize prices by increasinglygrowing maize in fields where they once grewwheat and soybeans, prices of these (and other)commodities have also become increasinglysensitive to oil prices.

Given that the energy market is muchlarger than the market for maize (if all theworld’s maize were used to produce biofuels,it would only meet 8 percent of energy de-mand), biofuel demand has the potential tochange permanently the nature (and price)of agricultural commodities. The InternationalEnergy Agency (IEA), for example, suggeststhat biofuel demand for grains could increaseby 7.8 percent a year over the next 20 years(compared with 1.2 percent annual increasesfor food demand). If this prognosis is borneout, 40 percent of global grain productioncould be going to biofuels by 2030.

It is probably premature to argue that thenature of these markets is permanentlychanged. On the one hand, technological im-provements are likely to lower the cost ofproducing ethanol from maize (and sugar),which in turn will lower the threshold oilprice above which these food crops becomesensitive to oil prices. However, technologicalchange may also give rise to alternativesources of energy that make ethanol produc-tion from food crops uneconomic. Such alter-natives might include biofuels made from cel-lulose or other nonfood sources, solar power,or hydrogen-based systems. In these cases, the

10363_Pg1-14:10363_Pg1-14 11/29/08 6:56 AM Page 7

new, stronger connection that has been cre-ated between the energy market and the grainmarkets would be broken, and food priceswould likely fall significantly.

Strong productivity growth and unusedcrop land should ensure adequate foodsupply at the global levelFood supplies are unlikely to fall short of de-mand. Over the past 30 years, agriculturalproductivity has improved much faster thandemand; as a result, agricultural output hasincreased rapidly even as the share of agricul-tural workers in total employment has steadilydeclined and prices fallen.

Longer-term prospects are somewhatclouded by the gradual exhaustion of the easyproductivity gains offered by the green revolu-tion. In addition, climate change threatensyields in many developing countries, althoughmost of this effect is not likely to be felt untilafter 2030. Assuming that policies are putin place to expand infrastructure and facili-tate the diffusion of the new technologies(including biotechnologies) that have sustainedagricultural productivity in high-income coun-tries, agricultural output should more than keeppace with food demand over the long term.

However, if developing countries are notsuccessful in combating recent trends for yieldsto decline by increasing investment in ruralagriculture and through the spread and adop-tion of more productive seed varieties andfarming techniques, there is a real risk thatmany countries, notably in Africa (where pop-ulation growth is expected to be faster), willmove from a position of being broadly self-sufficient in food to being net food importers.Most of the shortfall would be met by produc-tion from high-income countries, where pro-ductivity growth has not slowed.

Even if biofuel demand increases substan-tially, enormous potential exists for bringingadditional (albeit lower productivity) land intocultivation. That said, if biofuel-related de-mand for crops is much stronger or productivityperformance disappoints, future food suppliesmay be much more expensive than in the past.


8

Figure O. 5 Oil prices are having a directimpact on food prices

Oil price per barrel versus food price per ton

a. Soybeans vs. Crude Oil Prices

0

100

200

300

400

500

600

700

Crude oil ($/bbl)

Crude oil ($/bbl)

Soybeans ($/ton)

b. Wheat vs. Crude Oil Prices

0

100

0 20 40 60 80 100 120 140

0 20 40 60 80 100 120 140

Crude oil ($/bbl)

0 20 40 60 80 100 120 140

200

300

400

500

Wheat ($/ton)

c. Maize vs. Crude Oil Prices

0

50

100

150

200

250

300

350

Maize ($/ton)

Source: World Bank.

10363_Pg1-14:10363_Pg1-14 11/29/08 6:56 AM Page 8

O V E R V I E W

9

0

10

20

30

40

50

60

70

0

10

20

30

40

50

60

70

Primary exports per capita(left axis)

Primary exports/exports(right axis)

Value of per capita primarycommodities in exports(US$ thousands)

Source: World Bank.

Share of primarycommodities in total

merchandise exports (%)

Figure O.6 On average, poor countries aredependent on commodities but relativelyresource poor

Low-incomecountries

Lower-middle-incomecountries

Upper-middle-incomecountries

High-incomecountries

for resource dependence to generate poorgrowth outcomes. These include:

• The tendency for government spendingin resource-dependent countries to risein booms and fall procyclically duringbusts;

• The tendency for strong revenue inflowsto cause an excessive real appreciationof the currency that hurts the competi-tiveness of the nonresource sectors ofthe economy; and

• The tendency for large commodity-based revenues to foster rent-seekingbehavior, corruption, and even politicalviolence.

Encouragingly, during the course of the re-cent commodity boom, fiscal spending inresource-dependent developing countries hasbeen much more prudent than during earlierbooms. Partly as a result, the currencies ofmost countries have appreciated by less thanin the past. Moreover, corruption among com-modity exporters has improved relative todiversified exporters (figure O.7), suggestingthat perhaps this mechanism for reducing thedevelopment potential of resource wealth hasbeen weakened as well.

Exceptions include newly independentcommodity exporters or states with newlyfound resource wealth. Government spendingin these countries has kept pace with or evenexceeded export revenues, and their currencieshave appreciated much more strongly thanthose with more experience of commoditybooms (figure O.8). With prices now sharplylower, such countries may be encounteringadditional fiscal pressures. In addition, oil ex-porters with relatively low reserves are notsaving significantly more than those with highreserves and, as a result, may be exacerbatingthe competitiveness problems of their non-oilsectors. That, in turn, could be creating afuture problem, because these countries, unlikethose with ample reserves, will have to rely ontheir non-resource-based sectors to generatemost of the growth in coming years.

Simulations suggest that under these unfavor-able circumstances, food crop prices could be asmuch as 30 percent higher than in the baselinescenario.

Commodity-producing countries aremanaging the revenue windfall better thanthey have in the pastHistorically, countries whose economies areheavily dependent on commodities exportshave tended to grow less quickly than thosewith more diverse economies. This tendencymainly reflects low GDP and underdevelop-ment of their nonresource sectors rather thanthe actual quantity of resources held by thesecountries. Indeed, measured by per capitavalue-added from resources, high-incomecountries tend to be more resource rich thandeveloping countries, while their large nonre-source sectors mean they are also less resourcedependent (figure O.6).

Resource dependence need not result inslow growth. But to realize the potential ofresource wealth, governments need to avoidfollowing policies that exacerbate the tendency

10363_Pg1-14:10363_Pg1-14 11/29/08 6:56 AM Page 9

commodities, such as agricultural producers,where the benefits of high prices are less con-centrated. Encouragingly, much of the spend-ing appears to be directed toward investmentgoods, which should contribute to futureproduction potential. In a number of Africancountries, however, investment spending hasbeen financed by heavy bank borrowing,which may pose significant problems as loansbecome due, now that commodity prices havedeclined and access to credit has becomemuch more difficult.

High commodity prices pose challenges forthe poor, especially in consuming nationsFor consuming nations, high commodityprices pose a number of challenges. In the caseof heavily traded commodities such as oil,sharp price hikes can pose serious balance of


10

Figure O.8 Exchange rates, inflation, andgovernment expenditures in new versusestablished oil exporters, 2001–06

% change

Real exchangerate with US$

Percentagechange inCPI, 2008

Change ingovernmentexpenditure/

GDP, 2001–07

Source: World Bank and IMF data.

Note: New producers are defined as countries dependent on oilthat began production after 1985 or were established as a countryafter 1985, including Azerbaijan, Chad, Equatorial Guinea,Kazakhstan, Sudan, and the Republic of Yemen (Turkmenistanlacks data for inflation and the real exchange rate). Theestablished producers include Algeria, Angola, Republic of Congo,Gabon, Islamic Republic of Iran, Libya, Nigeria, Oman, andRepública Bolivariana de Venezuela. The real exchange rate withthe United States (rather than the trade-weighted real exchangerate as in figure 3.5) is reported here to include sufficient countriesfor a useful comparison between the two groups.

a. Real exchange rate with the U.S. dollar, where increaseindicates appreciation. Data for Equatorial Guinea are for2001–04.b. Percentage change in consumer price index in 2008.c. Change in ratio of government expenditure to GDP from2001 to 2007.

210

0

10

20

30

40

50

60

New producers

Established producers

Figure O.7 Primary commodity exporters areexhibiting fewer signs of the behaviorslinked to the “resource curse”

Percentage change in the share of GDP

a. Government expenditures have increased by much lessthan export revenues

26

24

22

0

2

4

6

8

1980s 2000s

Index

21.0

20.8

20.6

20.4

20.2

0.0

1996

Bet

ter

Wor

se

2006

Source: World Bank.

b. The currencies of commodity exporters have appreciatedmodestly

c. Corruption in commodity exporting countrieshas declined

Percentage change in trade-weighted real effectiveexchange rate

Source: IMF data; World Bank staff calculations.

Note: Increase indicates appreciation.

Source: Kaufmann, Kraay, and Mastruzzi 2007; World Bank data.

Change inexports/GDP

Change in governmentexpenditures/GDP

220

215

210

25

0

5

10

Non-oil exporters Oil exporters

1980s boom

Recent boom

Oil and mineral exporters

Agricultural exporters

Diversified exporters

In addition, spending from resource rev-enues in the private sector remains high. Thisis especially true for exporters of non-oil

10363_Pg1-14:10363_Pg1-14 11/29/08 6:56 AM Page 10

payment difficulties and increase the vulnera-bility of net importers. In the case of foodcommodities, which are mainly consumed inthe same country in which they are pro-duced, the issue for most countries is one of atransfer of wealth between producers andconsumers. That said, some countries are sig-nificant net importers of food and have suf-fered significant balance of payment impactsfrom high food prices as well. Both fuel andfood prices have boosted inflation and cut intoreal incomes in developing countries.

In general, economic policy should not re-sist changes in relative prices but should seekto assist adjustment to changing circum-stances. However, the magnitude of thechanges over the past several years has beenunusually large with important implicationsfor inflation, balance of payments, and povertyin developing countries. Moreover, becausehigh food prices can increase malnutritionamong the very poor, resulting in permanentcognitive and physical damage, even a tempo-rary but large hike in food prices demands aprompt and well-targeted policy response.

At the global level, the cost of higher foodand fuel prices to consumers in developingcountries during 2008 is estimated to havebeen about $680 billion. The price increaseshad major macroeconomic effects. High oilprices increased current account deficits in anumber of countries by as much as 5 percentof their GDP. Both food and fuel price in-creases have led to a sharp uptick in inflation.In addition, by increasing costs, the food andfuel increases have increased the number ofpoor and the extent of their poverty. In gen-eral, higher food prices have had a more pro-nounced effect on poverty, because householdsin poor countries spend 50 percent or more oftheir income on food and only 10 percent onfuel. Moreover, for very poor households, foodtends to claim an even higher share in expen-ditures, and fuel a much lower share. Finally,the poverty impacts are likely to be more sig-nificant because the demand for food is moreinelastic than household demand for fuels, be-cause the former can be replaced by biomass.

Not all foods prices have risen by as much asthe prices for rice, maize, and wheat, however.Moreover, during 2007 and the first half of2008, the dollar was depreciating so that localcurrency prices rose by less than the dollarprices. As a result, the real-local-currency in-crease in the price of food actually consumed indeveloping countries was much less than the54 percent increase observed in internationallytraded and dollar-denominated food prices(table O.1). Moreover, not all food consumed inpoor countries is traded and the share of non-traded foods in total consumption varies acrossregions. In Africa, for example, real food pricesrose by an average of 8.3 percent, comparedwith 19.8 percent in the Middle East, which re-lies much more heavily on imported foods.

Overall, the rise in food prices between 2005and the beginning of 2008 is estimated to haveincreased the share of the population of EastAsia, the Middle East, and South Asia livingin extreme poverty by 1 or more percentagepoints. Impacts in Africa were less pronouncedbecause food prices rose by less on average and

O V E R V I E W

11

Table O.1 Food price hikes and consump-tion shares vary by region

FoodPrice share among

Region shock the poor

(percent)Rural populationEast Asia and Pacific 12.4 71.5Europe and Central Asia �0.2 63.4Latin America and the Caribbean 6.9 51.2Middle East and North Africa 25.9 64.5South Asia 5.0 65.3Sub-Saharan Africa 9.6 68.0

Developing world 6.7 66.1

Urban PopulationEast Asia and Pacific 13.8 67.5Europe and Central Asia �0.5 57.9Latin America and the Caribbean 1.6 44.1Middle East and North Africa 12.5 57.1South Asia 4.8 64.4Sub-Saharan Africa 4.9 53.0


Source: World Bank.Note: Price shocks differ between the rural and urban popu-lations because of differing degrees of urbanization amongcountries included in the aggregates.

10363_Pg1-14:10363_Pg1-14 11/29/08 6:56 AM Page 11

because a much larger share of the populationlives in rural areas. In general, rural dwellershave been less seriously affected because, in ad-dition to being consumers, many are producersand benefit from higher revenues. The impacton the urban poor was much higher, increasingthe incidence of poverty by more than 1.5 per-centage points in East Asia, the Middle East,South Asia, and Sub-Saharan Africa (table O.2).Overall the number of extremely poor is esti-mated to have increased by between 130 and155 million, and the poverty deficit (the annualcost of lifting the incomes of all of the poor to

the poverty line) increased by $38 billion, or 0.5percent of developing-country GDP.

For the very poor, reducing consumptionfrom already very low levels, even for a shortperiod, can have important long-term conse-quences. Already, higher food prices during2008 may have increased the number ofchildren suffering permanent cognitive andphysical injury caused by malnutrition by44 million. It is therefore critical that coun-tries react to higher food prices by increasingthe assistance they make available to thosemost at risk.

Most countries have reacted to the hike infood and fuel prices by some combination ofincreased government spending on existingsocial safety net programs, be they subsidies,conditional transfer systems, or food distribu-tion schemes. Others have responded by seek-ing to hold domestic prices down by reducingtaxes or instituting restrictions on exports.

Such programs have been relatively expen-sive, increasing government expenditures byas much as 2–4 percent of GDP. Moreover, inmany cases poor targeting means that much ofthis spending does not benefit those most inneed. And, by interfering with market prices,these programs often impede adjustment, re-ducing producers’ incentives to increase out-put and consumers’ incentives to conserve. Assuch they likely exacerbated the extent of pricerises and extended their duration.

Going forward, policy makers need torestructure their support so that it is bettertargeted on the very poor. Doing so will help en-sure that the next time food (or energy) pricesspike, assistance programs will be both moreaffordable and more effective at delivering assis-tance to those most in need. Of the optionsavailable, targeted cash transfers tend to succeedbest because they have relatively low adminis-trative requirements and minimize the diversionof benefits toward less needy population groups.Unfortunately, these programs may also excludethe many poor who are either unable or unwill-ing to meet the conditions attached to the pro-gram, which are designed to dissuade all but themost needy from participating. In-kind


12

Table O.2 Higher food prices haveincreased both the incidence and severityof poverty worldwideJanuary 2005–December 2007

Initial levels: Change in:

Poverty Income Poverty IncomeRegion headcount gap ratio headcount gap ratio

(percent) (percentage points)Urban populationEast Asia and Pacific 13.2 20.3 6.3 2.7Europe and

Central Asia 2.5 8.7 0.0 0.2Latin America

and the Caribbean 3.7 37.6 0.1 �0.7Middle East

and North Africa 2.7 17.8 2.4 5.7South Asia 32.3 25.0 2.0 0.5Sub-Saharan Africa 34.1 38.1 1.7 0.3

Developing world 15.3 27.1 2.9 0.5

Rural populationEast Asia and Pacific 31.9 23.2 4.9 0.7Europe and

Central Asia 8.2 6.6 0.0 0.0Latin America

and the Caribbean 18.6 43.9 0.1 0.1Middle East

and North Africa 15.4 22.9 0.7 0.9South Asia 43.3 24.0 0.8 0.3Sub-Saharan Africa 54.9 41.5 0.3 0.0


Source: World Bank, using the Global Income DistributionDynamics model.Note: The per capita poverty line equals 1.25 international2005 dollars a day. The ratio of food in total consumptionamong the poor is computed as described in De Hoyos andLessem (2008). East Asia excludes China, and the MiddleEast comprises Jordan, Morocco, and the Republic of Yemen.The income gap ratio expresses, as a percent of the povertyline, how much the income of the average poor person islower than the poverty line.

10363_Pg1-14:10363_Pg1-14 11/29/08 6:56 AM Page 12

programs, such as school feeding and the distri-bution of fortified weaning food for toddlers,can be effective, especially in fiscally constrainedcountries. Subsidies, even targeted ones, tend tobe much less efficient, with as little as one-fifthof the money spent benefiting the poor. Publicworks programs rarely provide sufficient cover-age to meaningfully target poor families. What-ever policies are adopted, it is critical that theoffsetting income support be clearly presentedas temporary and include phaseout strategies toavoid creating an unnecessary longer-term fiscalburden.

The role for international policyUltimately, given the scope of the costs in-volved, neither individual governments nor in-ternational agencies are in a position to offsetthe costs of higher food and fuel prices en-tirely. However, well-targeted programs aremuch more affordable. For the poorest coun-tries, these too may be beyond reach fiscally,and in these cases, the international commu-nity has a role.

Steps so far have concentrated on reallocat-ing existing funds toward those most in needand on strengthening both the financial andinfrastructural capacity of emergency food aidagencies such as the World Food Programme(WFP). Further steps that might be consideredinclude providing the WFP with a more stablesource of financing and affording it a line ofcredit so that it is able to act quickly in in-stances where food prices are unusually high.

Policy makers might also examineprospects for improving the coordinated man-agement of grain reserves so that they can bemore easily brought to the aid of those inneed. Steps might include the constructionof storage facilities in strategic parts of theworld and the creation of a managementsystem perhaps along the lines of that usedby the IEA for oil. Individual food-importingand -exporting nations may wish to explorethe use of market-based future contracts as an

alternative to building stocks and restrictingexports. Such contracts can reduce both priceand quantity uncertainty by providing forguaranteed delivery of fixed quantities ofgrains at fixed prices. They can even be writ-ten conditionally, providing an option to sellor buy that can be exercised depending onmarket conditions.

Trade reform will necessarily form part ofthe solution as well. Steps are required tosanction effectively countries that use exportrestrictions as a mechanism to control domes-tic prices. Not only do such restrictionsinterfere with the domestic supply response,they also tend to exacerbate the price hikesand shortages in the rest of the global econ-omy. Although a successful conclusion to theWorld Trade Organization’s Doha Round ofmultilateral trade negotiations might result inhigher prices in the short run, it would likelyprove beneficial to developing countries byimproving the competitiveness of their agri-cultural sectors and reducing their reliance onimported food.

ReferencesDe Hoyos, R., and R. Lessem. 2008. “Food Shares

in Consumption: New Evidence Using EngelCurves for Developing Countries.” World Bank,Washington, DC.

Grilli, Enzo R., and Maw Cheng Yang. 1988. “PrimaryCommodity Prices, Manufactured Good Prices,and the Terms of Trade of Developing Countries:What the Long Run Shows.” World Bank Eco-nomic Review 2: 1–47.

IEA (International Energy Agency). 2006. EnergyTechnology Perspectives: Scenarios & Strategiesto 2050. Paris: OECD/IEA.

Kaufmann, D., A. Kraay, and M. Mastruzzi. 2007.Governance Matters VI: Aggregate and Individ-ual Governance Indicators for 1996–2006. WorldBank Policy Research Working Paper 4280,Washington, DC.

World Bank. 2007. At Loggerheads: AgriculturalExpansion, Poverty Reduction, and Environmentin the Tropical Forests. Washington, DC: WorldBank.

O V E R V I E W

13

10363_Pg1-14:10363_Pg1-14 11/29/08 6:56 AM Page 13

10363_Pg1-14:10363_Pg1-14 11/29/08 6:56 AM Page 14

15

Prospects for the Global Economy1

The stresses in the financial markets of theUnited States that first emerged in the

summer of 2007 transformed themselves into afull-blown global financial crisis in the fall of2008: credit markets froze; stock marketscrashed; and a sequence of insolvencies threat-ened the entire international financial system.Massive liquidity injections by central banksand a variety of stopgap measures by govern-ments proved inadequate to contain the crisisat first.

The initially hesitant policy response hasbecome increasingly robust. The United Statesgovernment introduced a $700 billion rescuepackage and has taken equity positions in ninemajor banks and several large regional banks.Various debt and deposit guarantees have alsobeen introduced. At the same time, Europeangovernments have announced plans for equityinjections and purchases of bank assets worthsome $460 billion, along with up to almost$2 trillion in guarantees of bank debt. At thetime of this writing, November 20, 2008, mar-kets remain volatile despite the forcefulness ofthese measures and signs that credit condi-tions are improving somewhat in high-incomecountries. Both private-sector and sovereigninterest rate spreads for developing countrieshave spiked even higher, and a growing list ofcountries have been forced to seek assistancefrom the International Monetary Fund (IMF).

During the initial phases of this financialcrisis in 2007, the effects of the financial tur-moil on developing countries were relatively

modest. However, as the crisis intensified in2008 and especially since mid-September, riskaversion (the absence of which had been thehallmark of the preceding boom) has in-creased, and capital flows to developing coun-tries have seized up. As a result, the currenciesof a wide range of developing countries depre-ciated sharply, and developing-market equityprices have given up almost all of their gainssince the beginning of 2008. Initial public eq-uity offerings have disappeared, and risk pre-miums have increased to more than 700 basispoints on sovereign bonds and to more than1,000 basis points on the debt of developing-country firms. Very recent data on bank lend-ing and foreign direct investment inflows arenot available, but indications are that these in-flows have also declined, but less dramatically.

Virtually no country, developing or high-income, has escaped the impact of the wideningcrisis, although those countries with strongerfundamentals going into the crisis have beenless affected. The deterioration in financingconditions has been most severe in countrieswith large current account deficits, and inthose that showed signs of overheating andunsustainably rapid credit creation before thefinancial crisis intensified. Of the 20 develop-ing countries whose economies have reactedmost sharply to the deterioration in conditions(as measured by exchange rate depreciation,increase in spreads, and equity market de-clines), 6 come from Europe and Central Asia,and 8 from Latin America and the Caribbean.

10363_Pg15_50:10363_Pg15_50 11/29/08 7:05 AM Page 15

Much tighter credit conditions will seeinvestment and GDP growth slow sharplyIn this climate, growth prospects for bothhigh-income and developing countries have de-teriorated substantially, and the possibility of aserious global recession cannot be ruled out.

Even if the waves of panic that have inun-dated credit and equity markets across theworld are soon brought under control, thecrisis is likely to cause a sharp slowdown inactivity stemming from the deleveraging infinancial markets that has already occurredand that is expected to continue. In the baselineforecast presented in this chapter, much tightercredit conditions, weaker capital inflows tomiddle-income countries, and a sharp reduc-tion in global import demand are expected tobe the main factors driving the slowdown indeveloping countries. Import demand is pro-jected to decline by 3.4 percent in high-incomecountries during 2009, while net private debtand equity flows to developing countries areprojected to decline from $1 trillion in 2007to about $530 billion in 2009, or from 7.7 to3 percent of developing-country GDP. As a re-sult, investment growth in developing countriesis projected to slow dramatically, rising only3.5 percent in middle-income countries, com-pared with a 13.2 percent increase in 2007.

A pronounced recession is believed to havebegun in mid-2008 in Europe, Japan, and mostrecently, the United States. This recession is pro-jected to extend into 2009, yielding a decline inhigh-income country GDP of 0.1 percent thatyear (table 1.1). In developing countries,growth is projected to slow to 4.5 percent in2009, down from 7.9 and 6.3 percent in 2007and 2008. Overall, global GDP is projected toexpand only 0.9 percent in 2009 (figure 1.1)—below the rate recorded in 2001 and 1991 andindeed, the weakest since records becameavailable beginning in 1970.

Because low-income countries have less ac-cess to international capital markets, the slow-down will affect them mainly through indirectmechanisms, including reduced demand fortheir exports, lower commodity prices, andreduced remittance inflows. International trade

is projected to decelerate sharply, with globalexport volumes falling by 2.1 percent in2009—the first time they have declined since1982 and eclipsing the 1.9 percent falloff thatoccurred in 1975. Export opportunities fordeveloping countries will fade rapidly becauseof the recession in high-income countries andbecause export credits are drying up and ex-port insurance has become more expensive.

Slower growth in high-income countries isestimated to have reduced remittance flowsinto developing countries from 2 to 1.8 per-cent of recipient country GDP between 2007and 2008. At the country level, the extent offurther slowdown will depend critically on ex-change rate developments, with recent swingsin bilateral exchange rates dwarfing the ex-pected changes in remittances denominated inhost-country currencies.

The global growth recession is projected tocause both commodity prices and inflation toease further, with oil prices averaging about$75 a barrel (bbl) in 2009, and food and metalprices projected to decline by about 23 and26 percent, respectively, compared with theiraverage levels in 2008. Nevertheless, com-modity prices will remain well above the verylow levels of the 1990s.

Lower commodity prices should reduce theburden on some segments of the poor (notablyurban dwellers), whose purchasing power hasdeclined because of high food and fuel prices(see chapter 3). Lower prices should also helpdampen headline inflation. Indeed, the rapidrise of food and energy prices over the courseof 2007 and the first half of 2008, coupled withtight capacity in many countries (followingyears of very fast growth fueled by ample liq-uidity) caused headline and core inflation topick up throughout the world. Headline infla-tion increased by 5 percentage points or more inmost developing countries, and more than halfof developing countries had an inflation rate inexcess of 10 percent by the middle of 2008.

This financial crisis and the expected abruptslowing of global growth comes at a momentwhen developing countries considered as awhole are more vulnerable than they have been


16

10363_Pg15_50:10363_Pg15_50 11/29/08 7:05 AM Page 16

P R O S P E C T S F O R T H E G L O B A L E C O N O M Y

17

Table 1.1 The global outlook in summary(percentage change from previous year, except for interest rates and oil prices)

Indicator 2006 2007 2008* 2009† 2010†

Global conditionsWorld trade volume 9.8 7.5 6.2 �2.1 6.0Consumer prices

G-7 countriesa,b 2.2 1.7 3.3 1.6 1.8United States 3.3 2.6 4.5 2.5 2.8

Commodity prices (US$)Non-oil commodities 29.1 17.0 22.4 �23.2 �4.3

Oil price (US$ per barrel)c 64.3 71.1 101.2 74.5 75.8Oil price (percent change) 20.4 10.6 42.3 �26.4 1.8

Manufactures unit export valued 1.6 5.5 9.0 2.1 1.3Interest rates

$ LIBOR, 6-month (percent) 5.2 5.3 3.3 1.9 2.5€ EURIBOR, 6-month (percent) 3.1 4.3 4.9 3.8 4.2

Real GDP growthe

World 4.0 3.7 2.5 0.9 3.0Memo item: World (PPP weights)f 5.0 4.9 3.6 1.9 3.9High-income countries 3.0 2.6 1.3 �0.1 2.0

OECD countries 2.9 2.4 1.2 �0.3 1.9Euro Area 2.9 2.6 1.1 �0.6 1.6Japan 2.4 2.1 0.5 �0.1 1.5United States 2.8 2.0 1.4 �0.5 2.0Non-OECD countries 5.5 5.6 4.3 3.1 5.3

Developing countries 7.7 7.9 6.3 4.5 6.1East Asia and the Pacific 10.1 10.5 8.5 6.7 7.8

China 11.6 11.9 9.4 7.5 8.5Indonesia 5.5 6.3 6.0 4.4 6.0Thailand 5.1 4.8 4.6 3.6 5.0

Europe and Central Asia 7.5 7.1 5.3 2.7 5.0Poland 6.2 6.6 5.4 4.0 4.7Russian Federation 7.4 8.1 6.0 3.0 5.0Turkey 6.9 4.6 3.0 1.7 4.9

Latin America and the Caribbean 5.6 5.7 4.4 2.1 4.0Argentina 8.5 8.7 6.6 1.5 4.0Brazil 3.8 5.4 5.2 2.8 4.6Mexico 4.9 3.2 2.0 1.1 3.1

Middle East and North Africa 5.3 5.8 5.8 3.9 5.2Algeria 1.8 3.1 4.9 3.8 5.4Egypt, Arab Rep. of 6.8 7.1 7.2 4.5 6.0Iran, Islamic Rep. of 5.9 7.8 5.6 3.5 4.2

South Asia 9.0 8.4 6.3 5.4 7.2Bangladesh 6.6 6.4 6.2 5.7 6.2India 9.7 9.0 6.3 5.8 7.7Pakistan 6.2 6.0 6.0 3.0 4.5

Sub-Saharan Africa 5.9 6.3 5.4 4.6 5.8Kenya 6.1 7.1 3.3 3.7 5.9Nigeria 5.2 6.5 6.3 5.8 6.2South Africa 5.4 5.1 3.4 2.8 4.4

Memo itemsDeveloping countries

excluding transition countries 7.8 7.9 6.3 4.6 6.2excluding China and India 6.0 6.1 5.0 2.9 4.7

Source: World Bank.Note: PPP � purchasing power parity; * � estimate; † � forecast.a. Canada, France, Germany, Italy, Japan, the United Kingdom, and the United States.b. In local currency, aggregated using 2000 GDP weights.c. Simple average of Dubai, Brent, and West Texas Intermediate.d. Unit value index of manufactured exports from major economies, expressed in U.S. dollars.e. GDP in 2000 constant U.S. dollars, 2000 prices, and market exchange rates.f. GDP measured at 2000 PPP weights.

10363_Pg15_50:10363_Pg15_50 11/29/08 7:05 AM Page 17

in the recent past. Higher commodity priceshave widened current account deficits of manyoil-importing countries to worrisome levels(they exceed 10 percent of GDP in about one-third of developing countries), and after havingincreased substantially, the international re-serves of oil-exporting developing countriesare now declining as a share of their imports.Moreover, inflation is high, and fiscal positionshave deteriorated both for cyclical reasons andbecause government spending has increased toalleviate some of the burden of higher com-modity prices.

Although the global recession is likely to beprotracted, some elements of an eventual re-covery can already be discerned. These includeearly movement toward stabilization in thehousing sector in the United States; continuedprogress on debt workouts and a strengthen-ing of balance sheets among both banks andhouseholds; a gradual easing of credit condi-tions as government rescue packages take holdand investors begin to return to heavily dis-counted equity markets; increases in real in-comes (stemming from lower food and fuelprices) among individuals with relatively highmarginal propensities to consume; and in-creased space for fiscal and monetary policiesas inflationary pressures ease and governmentoutlays on food and fuel subsidies decline intandem.

Prudent and vigilant policies are keyas uncertainty continues to cloud theoutlookAlthough this sober outlook represents alikely outcome, the situation remains unsta-ble, and a wide range of outcomes are possi-ble, including a scenario where the reboundof growth in 2010 is weaker, held back bycontinuing banking sector restructuring, andnegative wealth effects resulting from lowerhousing and stock market prices.

An even sharper recession is also possible.If the freeze in credit markets does not thaw asanticipated in the baseline, the consequencesfor developing countries could be cata-strophic. Financing conditions would deterio-rate rapidly, and apparently sound domesticfinancial sectors could find themselves unableto borrow or unwilling to lend—in both inter-national and domestic markets. Such a sce-nario would be characterized by a long andprofound recession in high-income countriesand substantial disruption and turmoil, in-cluding bank failures and currency crises, in awide range of developing countries. Sharplynegative growth in a number of developingcountries with all of the attendant repercus-sions, including increased poverty and unem-ployment, would be inevitable.

Although it is a receding concern, high infla-tion in developing countries remains a problem,especially if the impact from the current crisison developing-country investment demand isless pronounced, and the stimulus provided byvarious rescue and fiscal packages in high-income and developing countries feeds a rapidexpansion in demand. Under such a scenario,global growth would still slow in 2009, whichwould tend to dampen inflationary pressuresinitially, but growth could be expected to snapback much more sharply in 2010. Countriesthat now have large current account deficitsand high inflation could suffer from a renewedoverheating of their economies. Policies wouldhave to be very prudent in these circumstances,because the currencies of these countries arelikely to remain sensitive to changing marketperceptions and increased risk aversion.


18

22

0

2

4

6

8

1980 1984 1988 1992 1996 2000 2004 2008

Source: World Bank.

Real GDP, percentage change

Figure 1.1 GDP growth

Asiancrisis

Dot-comcrisis Forecast

Developing countries, excluding China and India

Developing countries

High-income countries

10363_Pg15_50:10363_Pg15_50 11/29/08 7:05 AM Page 18

The challenge for policy makers is not onlyto prevent an escalation of the crisis and to mit-igate the downturn but also to ensure a goodstarting position once the rebound sets in. Fordeveloping countries, this means respondingrapidly and forcefully to signs of weakness indomestic banking sectors, including resortingto international assistance where necessary. Italso means pursuing a prudent countercyclicalpolicy, relying on automatic stabilizers, socialsafety nets, and infrastructure investment thataddresses bottlenecks that have become bind-ing constraints on long-term sustainablegrowth in many countries.

In the current circumstances of heightenedrisk aversion and investor skittishness, policymakers need to be especially wary of takingon excessive levels of debt or creating the con-ditions for an inflationary bubble by reactingtoo aggressively to the global slowdown.Although it is important for policy makers toreact quickly to emerging problems, it is alsoessential that steps and conditions attached toassistance be well focused on overcomingsome of the fundamental sources of weak-ness. Otherwise there is a risk that govern-ments lose the support of markets and tax-payers in their efforts to limit the extent ofnear-term disruptions.

Financial markets

The deterioration in financialconditions accelerated markedlyin September 2008

The protracted turmoil that has plaguedglobal financial and credit markets since

mid-2007 escalated in September 2008, withthe sudden collapse of major financial insti-tutions, first in the United States and subse-quently in Europe. The crisis has spreadrapidly to emerging markets and has raisedfears of systemic risk to the internationalfinancial system. Growing concerns aboutcounterparty risk have disrupted credit mar-kets, especially the interbank and commercialpaper markets.

Earlier in the year major banks were stillable to attract new equity investors in an effortto rebuild their capital bases, which wereeroded by significant losses stemming fromlarge write-downs on mortgage-backed securi-ties and other assets. However, investor confi-dence was shaken by the March collapse ofBear Sterns, the seventh largest securities firmin the world in total assets. In September andOctober, authorities in the United States andEurope had to respond with extraordinarysteps, including large injections of liquidity; co-ordinated reductions in policy interest rates;the takeover of major financial institutions; en-hancements in deposit guarantees; and plans topurchase impaired financial assets (such as theU.S. Troubled Asset Repurchase Program, orTARP), to take equity positions in commercialbanks, and to intervene in the commercialpaper markets (box 1.1).

The turmoil has had a dramaticimpact on emerging market assetsAlthough financial institutions in developingcountries are believed to have limited directexposure to U.S. subprime assets and relatedsecurities, the financial turmoil has affectedvirtually all emerging-market economies ashigh-income-country banks and investmentfunds withdrew from emerging markets andconverted a broad range of risky assets intomore liquid holdings. The rapid increase inrisk aversion has also led to a forceful unwind-ing of the carry trade. The sell-off in risky as-sets carried a dramatic impact on equity prices,bond spreads, and currencies in virtually allemerging-market economies and has also con-tributed to tighter domestic credit conditions inlarger countries, including India, the RussianFederation, and Brazil, and smaller countries,including Thailand and the Philippines. Thesedevelopments were reinforced as local in-vestors also moved out of equity markets, andmore generally, out of investments denomi-nated in local currencies.

Countries with large current accountdeficits, and therefore most dependent on for-eign capital, were hit hardest by the substantial


19

10363_Pg15_50:10363_Pg15_50 11/29/08 7:05 AM Page 19


20

Financial stress escalated in the United States andEurope over the course of 2008, beginning with

the takeover of Bear Stearns by JP Morgan in March,and culminating in September when several majorinstitutions came under severe distress.

Week of September 7The U.S. government seized control of Fannie

Mae and Freddie Mac, institutions that own or guar-antee about one-half of all mortgage assets in theUnited States.

Week of September 14The U.S. investment bank Lehman Brothers filed

for bankruptcy and Merrill Lynch was taken over bythe Bank of America for $50 billion.

The U.S. government seized control of AmericanInternational Group Inc., providing an $85 billionemergency loan and taking a 79.9 percent equitystake in the firm.

Britain’s largest mortgage lender, HBOS, agreed tobe purchased by Lloyds TSB in an $18.9 billion deal.

The Russian government pledged to provide$120 billion to support financial markets and banks(the amount was increased by $50 billion onOctober 7).

U.S. Treasury Secretary Henry Paulson introducedthe Troubled Asset Relief Program, a key element ofwhich enables the government to buy up to $700 bil-lion of mortgage-backed securities. An amendedversion was signed into law on October 4th.

Week of September 21Goldman-Sachs and Morgan Stanley became

bank holding companies.The U.K. government nationalized the mortgage

bank Bradford and Bingley (a loan portfolio of$90 billion).

Week of September 28Washington Mutual became the largest bank

failure in U.S. history, with assets valued at$328 billion.

The Belgian, Dutch, and Luxembourg govern-ments each took a 49.9 percent equity stake in theoperations of the banking and insurance companyFortis within their respective borders, each injecting$16.4 billion in capital. One week later, the Dutchgovernment took full control of the company’soperations in the Netherlands. Fortis’ operations in

Box 1.1 Chronology of recent developments in thefinancial crisis

the BENELUX countries were later sold to theFrench commercial bank, BNP Paribas.

The German government, together with commer-cial banks and federal regulators, provided $50 bil-lion in credit guarantees to Hypo Real Estate.

Citigroup agreed to buy the banking operationsof Wachovia.

France, Belgium, and Luxembourg injected$9.2 billion into the French-Belgian bank Dexia.

The Icelandic government took a 75 percent equitystake in Glitnir, the country’s third-largest bank.

The Swedish central bank announced that itwould lend up to $700 million to the Swedish unitof the Icelandic bank Kaupthing.

Ireland announced unlimited guarantees on retail,commercial, and interbank bank deposits. Similarmeasures were adopted in Austria, Denmark,Germany, Greece, Iceland, Italy, and Portugal.Sweden, the United Kingdom, and the United Statesraised limits on deposit guarantees. On October 3,European finance ministers agreed to raise the mini-mum guarantee on bank deposits to €50,000 acrossall EU member states.

Week of October 5The Icelandic government loaned $683 million to

Kaupthing, and seized control of Landsbanki, andsought a $5.5 billion loan from Russia.

The Spanish government established a $40 to$68 billion emergency fund to purchase assets heldby Spanish banks.

The U.S. Federal Reserve intervened in thecommercial paper market for the first time since theGreat Depression.

The British government made available $87 bil-lion in emergency loans to the banking system andoffered to purchase capital in eight of the largestbanks. The package includes guarantees of £250 mil-lion for new debt and the same for liquidityprovisions.

The central banks of the United States, the EuroZone, Canada, Sweden, and Switzerland each cuttheir benchmark rates by half a percentage point inan unprecedented coordinated effort. Separately,China’s central bank lowered its key one-year lend-ing rate by 27 basis points, the second reduction inthree weeks.

10363_Pg15_50:10363_Pg15_50 11/29/08 7:05 AM Page 20

tightening of credit conditions in internationalmarkets. One-third of developing countries arerunning current account deficits in excess of10 percent of GDP, many of which may beforced to restrict domestic demand severely ascapital inflows dry up. During 2007, for ex-ample, several countries were the recipients ofvigorous increases in private debt flows thatfueled credit growth to the domestic privatesector and intensified inflation pressures. Ayear later private debt flows to the bankingsector declined dramatically in a number ofcases: by $13.2 billion in Kazakhstan, $6.6 bil-lion in Russia, $3.7 billion in South Africa,

$3.1 billion in Turkey, and $2.1 billion inUkraine (comparing January through Septem-ber 2008 with the same period in 2007).

All middle-income countries, even withcurrent account surplus positions, have cometo be substantially affected by the financialcrisis. A Revealed Vulnerability Index indi-cates the extent to which financial conditionsfor developing countries have deterioratedsince September 15, 2008 (upon the failure ofLehman Brothers). The vulnerability indexaverages the standardized depreciation of cur-rencies, domestic equity market losses, and in-creases in risk premiums, as well as the decline


21

The Belarusian authorities requested financial as-sistance from the IMF under a program that couldbe supported by a Stand-By Arrangement.

The Pakistani authorities requested discussionswith the IMF on an economic program supported byfinancial assistance from the IMF.

Week of October 26IMF staff agreed with the Hungarian and Ukrain-

ian authorities’ economic programs supporting loansof $15.7 and $16.7 billion, respectively.

The European Union stood ready to provide aloan of $8.1 billion to Hungary and the World Bankagreed to provide $1.3 billion.

The IMF announced the Short-Term LiquidityFacility designed to channel funds quickly to emerg-ing markets that have a strong track record, but thatneed rapid help during the current financial crisis toget them through temporary liquidity problems.

Week of November 9The Leaders of the Group of Twenty agreed to a

plan of action to restore global growth and achieveneeded reforms of the world’s financial systems.

IMF staff and Pakistani authorities reached agree-ment on an economic program supported by a $7.6billion loan. The Executive Board of the IMF wasexpected to discuss the program shortly under theIMF’s Emergency Financing Mechanism procedures.

Week of November 16IMF staff and Serbian authorities agreed on an

economic program supported by a $0.5 billion loan.

The Icelandic government placed Glitnir into re-ceivership, seized control of Kaupthing Bank, andabandoned its attempt to peg the krona at 131 pereuro, established one day earlier, after it touched 340against the euro.

California, the most populous U.S. state, askedfederal authorities for a $7 billion emergency loan asit was unable to obtain financing in the wake of thebankruptcy of Lehman Brothers.

The British government announced a $685 billionplan to restore confidence in financial institutions,which included insuring up to $438 billion in newdebt issued by banks, along with providing as muchas $88 billion in equity capital.

The National Bank of Ukraine seized control ofProminvestbank, the country’s sixth-largest bank.

Week of October 12European governments announced financing

packages totaling over $2.5 trillion. The packages in-clude recapitalizing the banking sectors, credit guar-antees on interbank lending, and direct loans.

The British government injected $60 billion in eq-uity capital into the country’s three largest banks.

The United States announced that it would com-mit $250 billion of the $700 billion rescue packageto recapitalize the banking sector.

Week of October 19The IMF agreed with Iceland on an economic

recovery program supported by a two-year loanof $2.1 billion.

10363_Pg15_50:10363_Pg15_50 11/29/08 7:05 AM Page 21

in gross capital flows to a country over thepreceding 12 months. The index shows thatvirtually all middle-income countries are expe-riencing financial stress.

Emerging-market equity prices—as cap-tured by the Morgan-Stanley CompositeIndex (MSCI)—tumbled by over 60 percent(in dollar terms) from their peak of October2007, bringing prices back to levels attained atthe beginning of 2005 (figure 1.2).1 The mas-sive correction in equity prices was wide-spread across emerging market economies,with the largest declines found in a number ofEuropean and Central Asian economies—Ukraine (80 percent), Romania (75 percent),Bulgaria (75 percent), and the Russian Feder-ation (73 percent). Other large emerging mar-ket economies, including Brazil, China, andIndia, experienced corrections of over 60 per-cent. Despite the declines of the past year, eq-uity prices in emerging markets remain abovethose in mature markets from a longer-termperspective, albeit characterized by muchhigher volatility.

The selloff in emerging market assets trig-gered a marked depreciation of exchange ratesin a large number of countries, reversing muchof the appreciation of the past two years. Forexample, the Brazilian real dropped by 40 per-cent against the dollar (20 percent against theeuro) from early August to mid-November2008, but the currency stands only 8 percentbelow its January 2007 dollar value. The

South African rand depreciated by nearly 60percent against the dollar (38 percent againstthe euro) from late October 2007 to mid-November 2008. Similarly, the Turkish liradepreciated by over 40 percent against thedollar (20 percent vis-à-vis the euro) from lateOctober 2007 to mid-November 2008, butthe currency has appreciated by 6 percent inreal effective terms between January 2007 andOctober 2008.

The banking crisis that erupted in September2008 is restricting credit to developing coun-tries, and in particular to the least creditworthyborrowers. Sovereign bond spreads widened toa peak of 1,100 basis points in late October2008 from 330 points in late August—wellabove the record 150 basis points registered inJune 2007—before recovering to just over 700basis points by mid-November. At that time,sovereign bond spreads exceeded the “dis-tressed debt” threshold of 1,000 basis points in14 of 38 emerging market economies currentlypart of JPMorgan’s EMBI Global Index.Corporate bond spreads jumped by still morethan sovereign spreads (figure 1.3). Spreads onrisky non-investment grade (BB-rated) emerg-ing market corporate bonds widened to 1,750basis points in mid-November, up more than1,450 points since mid-2007.


22

50

100

150

200

250

Source: Morgan-Stanley, Standard & Poor’s.

MSCI equity price indexes (January 2005 5100)

Figure 1.2 Emerging market equities are hithard as turbulence evolves to crisis

Euro Area

United States Oct. 2007

Emerging markets

Jan.2005

Jan.2006

Jan.2007

Jan.2008

0Jan.2007

May2007

Sep.2007

Jan.2008

May2008

Sep.2008

200

400

600

CorporateCEMBI composite

SovereignEMBIG composite

800

1000

1200

Source: JPMorgan-Chase.

Note: CEMBI 5 Corporate Emerging Markets Bond Index; EMBIG 5 Emerging Markets Bond Index Global.

Spreads in basis points

Figure 1.3 Emerging-market bond spreadswiden, especially for corporates

10363_Pg15_50:10363_Pg15_50 11/29/08 7:05 AM Page 22

The rise in spreads was only partially offsetby a 1.3 percentage point decline in the bench-mark yield on 10-year U.S. Treasury notes be-tween June 2007 and November 2008, so theyield on dollar-denominated sovereign bondsissued by emerging markets reached 10.5 per-cent, the highest in four years.

Private capital flows expected tocontinue declineEven before the intensification of the financialcrisis, tighter credit conditions were curtailinggross private debt and equity flows to devel-oping countries (figure 1.4). Cross-border syn-dicated bank loan commitments declined to$315 billion over the 12-months endingOctober 2008, down from a record $400 bil-lion a year earlier (but still above levelsrecorded over 2005–06). Bond issuance by de-veloping countries decreased to $72 billionover the year to October, down from a recordhigh of just over $170 billion at mid-2007.Corporate bond issuance declined sharply,with large falloffs in South Africa (by $15.6billion), India ($12.8 billion), and Russia($9.4 billion). Indeed, non-investment-gradebonds by corporations located in emergingmarkets accounted for about 40 percent oftotal issuance from January to October 2008,compared with 60 percent over 2005–06. Andequity issuance plummeted along with fallingequity prices to total $90 billion in the period,

down from a record high of $200 billionachieved at the start of 2008.

Even assuming a restoration of market con-fidence and a thawing of credit markets andcapital movements, overall credit conditionsfor emerging markets will remain substan-tially tighter than in the recent past. As a con-sequence, net private debt and equity flows todeveloping countries are anticipated to declinefrom the record-high $1.03 trillion (7.6 per-cent of developing-country GDP) set in 2007to about $530 billion (3 percent of GDP) in2009. Although net foreign direct investmentshould be moderately more resistant to thedownturn (as in past episodes), tighter creditmay cause high-income firms to reduce theirforeign direct investment by much more thanthey did during earlier financial crises, whichwere centered in developing countries.

Tightening of credit sharply reducesdomestic growth prospectsBefore the financial turmoil developed into acrisis of global proportions, developing coun-tries were affected mainly by slowing demand inhigh-income countries through the export chan-nel. Many developing countries had shownstrong resilience in the face of the gradually de-teriorating external environment, because theireconomies were supported by strong investmentgrowth and shielded by large amounts of inter-national reserves. This situation has changeddramatically since September 2008.

Unlike gradual adjustments in markets forgoods and services, adjustments in financialmarkets come fast and suddenly, and theyoften tend to “overshoot.” More importantly,the escalation of the crisis directly affects theengine for domestic growth in many develop-ing economies, because obtaining finance forcapital spending has become abruptly moredifficult. Moreover, the crisis is placing strongpressure on foreign exchange reserves and, atthe same time, can reveal dangerous currencymismatches in private sector balance sheets.

During the global boom of the past fiveyears, local banks and private companieswhose local currencies were appreciating


23

0

100

Jan.2004

Jan.2005

200

300

400 Bank loancommitments

Source: Dealogic Analytics.

US$ billions

Figure 1.4 Private debt and equity flowsdecline by a third in 2008

Bondissuance

Equityissuance

Jan.2006

Jan.2007

Jan.2008

10363_Pg15_50:10363_Pg15_50 11/29/08 7:05 AM Page 23

found it attractive to borrow abroad indollars. With the sudden turnaround in cur-rency movements, the currency mismatch onprivate sector balance sheets has likely led tosubstantial losses across firms and banks andeven for households. To the extent that thisdevelopment results in loan defaults, it maystrain domestic banking systems and placepressure on banks to find alternative sourcesof funding at a time when global financingconditions have deteriorated markedly. Stressinduced by currency movements thus carriesthe potential to further degrade prospects forinvestment spending in developing countries.

Outlook for high-incomeOECD countries

The intensification of the financial crisis inthe United States, and its widening to

major European countries in the autumn of2008, is expected to exact a significant toll oneconomic activity across the high-incomecountries belonging to the Organisation forEconomic Co-operation and Development(OECD). Even if confidence in global creditmarkets is restored quickly, reductions in thefinance available for firms and consumers,coupled with a slowdown in developing-country import demand, have set the stage fora recession in the United States, Europe, andJapan beginning in the second half of 2008 andlasting into 2009.

A movement to joint recession across keyOECD countriesThrough the first quarter of 2008, the slow-down among the OECD countries was fairlymoderate (although industrial productionstagnated in the quarter): exports benefitedfrom strong import demand from developingcountries and from oil exporters, while fallingimports served to boost the contribution ofnet trade to GDP growth. But GDP fell to0.3 percent growth (at an annual rate) in thesecond quarter for the key advancedeconomies, down from 2.4 percent in 2007;and as GDP growth moved to decline across

the United States, Europe, and Japan duringthe third quarter, OECD growth dropped to�0.6 percent.

Industrial production slipped to negativeground across all major OECD economies inboth the second and third quarters of the year,as fading overseas demand combined with alack of domestic orders tied to sluggish condi-tions in housing and autos in a number ofcountries (table 1.2, first and second panels).Growth of export volumes dropped from 14.6percent in 2007 to 2.5 percent during the thirdquarter (saar), as intra-OECD trade (especiallywithin Europe) softened at the same time asdeveloping-country demand slowed.

During the second quarter, conditions inEurope and Japan deteriorated sharply.Japan’s GDP declined at a steep 3.7 percentseasonally adjusted annualized rate (saar), andEuro Area GDP fell 0.7 percent (figure 1.5 andtable 1.2, first panel). A falloff in householdspending tied to the effects of much higher in-flation, a decline in investment, and a dra-matic shift in the growth contribution of tradeall contributed to the turnaround. In Europe,increased sluggishness in export markets andthe long bout of euro appreciation pressuredexports and imports into negative territory inthe second quarter, with contributions to over-all growth slipping to nil from 1.4 points in thefinal quarter of 2007.

In the United States, conversely, GDPpicked up 2.8 percent (saar) in the secondquarter, as fiscal stimulus and looser monetarypolicy boosted consumption spending. More-over, the pace of decline in residential invest-ment slackened, while contributions fromtrade—still benefiting from the weak dollar—increased to a large 1.6 percentage points ofgrowth (figure 1.6). U.S. domestic demand hasbeen depressed since the final quarter of 2006,as a rise in domestic savings helped to unwindthe global imbalances that were of suchconcern a couple of years ago. As financialdislocations heightened in the third quarter,including unprecedented declines in equity mar-kets in Europe, Japan, and the United States,consumer spending came under increasing


24

10363_Pg15_50:10363_Pg15_50 11/29/08 7:05 AM Page 24


25

Figure 1.5 Change in GDP in the UnitedStates, Europe, and Japan

GDP growth (percentage change)

24

22

0

2

4

6

Source: World Bank, national statistical agencies.

JapanGermanyEuro AreaUnited States

Q4,

2007

Q1,

2008 Q

2,20

08Q

3,20

08Table 1.2 High-income OECD countries: growth and related indicators

Seasonally adjustedGrowth year-over-year annualized growth Growth year-over-year

Indicator, country 2006 2007 Q108 Q208 Q308 H108 Q308

GDP growth (percent)High-income OECD 2.9 2.4 1.5 0.3 �0.6 1.9 0.5

United States 2.8 2.0 0.9 2.8 �0.5 2.3 0.7Japan 2.4 2.1 2.5 �3.7 �0.4 1.0 0.0Euro Area 3.0 2.6 2.6 �0.7 �0.8 1.7 0.3

Germany 3.2 2.6 5.7 �1.7 �2.1 2.2 0.8France 2.4 2.1 1.6 �1.1 0.6 1.6 0.6

United Kingdom 2.8 3.0 1.1 0.0 �2.0 1.9 0.3

Industrial productionHigh-income OECD 2.9 2.4 0.1 �3.2 �3.8 1.4 �2.2

United States 2.2 1.7 0.4 �3.2 �5.9 1.0 �4.5Japan 4.1 2.9 �1.7 �3.5 �4.3 1.8 �1.9Euro Area 3.4 2.8 1.3 �4.5 �2.6 1.2 �1.8

Germany 5.9 6.1 4.9 �3.2 �4.7 4.2 �2.3France 0.9 1.2 0.1 �6.1 �2.6 0.7 �1.9United Kingdom 0.7 0.4 �1.7 �2.9 �5.8 �0.2 �3.0

Consumer pricesa

High-income OECD 2.3 2.0 3.1 3.2 4.0 3.2 4.0United States 3.2 2.9 4.3 4.0 4.9 4.2 4.9Japan 0.3 0.1 0.7 1.2 2.1 1.0 2.1Euro Area 1.9 2.3 3.3 3.5 3.9 3.4 3.9

Germany 1.7 2.3 2.8 3.1 2.9 3.0 2.9France 1.7 1.5 2.8 3.2 3.0 3.0 3.0

United Kingdom 2.3 2.3 2.2 2.4 5.0 2.3 5.0

Export volumesHigh-income OECD 13.9 14.6 9.2 2.8 2.5 6.7 4.8

United States 10.5 6.9 5.3 14.7 21.9 9.8 11.2Japan 8.1 5.9 4.0 2.1 �12.9 7.1 �1.6Germany 13.0 6.2 11.0 0.4 �5.9 6.4 3.9France 9.9 3.7 32.6 �14.6 �0.1 5.1 4.2United Kingdom 11.7 �10.4 0.7 �1.3 4.2 �0.8 0.2

Source: National statistical agencies through Haver Analytics and Thomson/Datastream.Note: CPI inflation for high-income OECD countries is GDP weighted.a. Year-over-year growth rates.

0

2

4

6

Figure 1.6 The contribution of U.S. domesticdemand to GDP growth

Source: U.S. Department of Commerce; World Bank calculations.

Percentage points (4-quarter moving average)

Q12000

Q12001

Q12002

Q12003

Q12004

Q12005

Q12006

Q12007

Q12008

Contribution ofdomestic demand

U.S. GDP

10363_Pg15_50:10363_Pg15_50 11/29/08 7:05 AM Page 25

pressure (figure 1.7).2 And with export per-formance for OECD economies fading on theback of sputtering global demand, the EuroArea and Japan fell into technical recession inthe quarter, while growth in the United Statesreverted to decline.

Financial crisis places outlook underexceptional uncertaintyGiven the dramatic developments over Sep-tember to November 2008, the depth of thecoming recession is difficult to gauge. Shouldcredit markets remain frozen and asset pricescontinue to fall, then the decline in outputover the next year could be extreme. However,the extraordinary measures now being takenby fiscal and monetary authorities are ex-pected to eventually restore confidence so thatbanks will no longer hoard cash, and busi-nesses can obtain the finance essential for nor-mal operations. Nevertheless, the outlook forOECD countries remains grim. A commonelement is a falloff in domestic demand—increasingly deep in business capital spending—no longer offset by support from net tradebecause of a coincident marked slowdown ingrowth among the developing countries.

With U.S. private consumption dropping anunprecedented 3.1 percent in the quarter, the

decline in GDP would have been much moresevere save for the contribution of net trade.Notwithstanding that GDP is projected to de-cline more sharply in the fourth quarter of2008, growth for the year is expected to regis-ter 1.4 percent, because of the strong contribu-tions of trade in the first half of the year. Anabrupt decline in investment and a 1 percentfalloff in consumer spending are expected tocause GDP to fall in both the first and secondquarters of 2009, with a shallow recovery be-ginning in the second half of the year. GDP isprojected to decline by 0.5 percent for all of2009 but to recover to a still below-par 2.0 per-cent in 2010 (figure 1.8).

Financial conditions in the Euro Area arenow also perilous. After having fallen 0.7 per-cent in the second quarter (saar), GDPdropped 0.8 percent during the third quarterand is expected to register modest declines incoming quarters before picking up steam to-ward the end of 2009. Growth is expected toregister a weak 1.1 percent increase for 2008as a whole and a 0.6 percent decline in 2009,before strengthening in 2010 to a still below-trend advance of 1.6 percent. The depth of re-cession in Europe should be comparable tothat in the United States, in part because cor-porate finance in Europe is more reliant on thebanking sector but also because lower com-modity prices will dampen import demand in


26

Figure 1.7 U.S. household wealth fallssharply in the last quarters

22,000

21,500

21,000

2500

0

500

1,000

1,500

2,000

Source: Federal Reserve.

US$ billions

Q2,

2007

Q3,

2007

Q4,

2007

Q1,

2008

Q2,

2008

Totalassets

Householdreal estate

Financialassets

Equities,mutual funds

21

0

2001 2002 2003 2004 2005 2006 2007 2008 2009 2010

1

2

3

4

Source: National statistical agencies, World Bank.

Real GDP growth (percentage change)

Figure 1.8 GDP to decline across theOECD

United States

Euro Area

Japan

ForecastDot-comrecession

10363_Pg15_50:10363_Pg15_50 11/29/08 7:05 AM Page 26

the Middle East and North Africa, a regionthat has been an important origin of exportdemand for Europe. The rapid decline in out-put, plus the drop in commodity prices shouldalleviate inflationary pressures in Europe. Butthe recent depreciation of the euro measuredagainst the dollar may continue, as investors’continue to perceive U.S. government securitiesas safe-haven assets.

Aside from sharp equity market declines,Japan’s financial markets have been less af-fected (through mid-November 2008) byfallout from the global financial and bankingcrisis. Nevertheless, the macroeconomic land-scape is surprisingly similar to that in theUnited States and Europe. Household spend-ing has retrenched, higher inflation has com-pressed purchasing power, and consumer sen-timent has plummeted to 17-year lows. TheTankan survey of business investment inten-tions has been marked down steeply, in linewith a downward shift in expectations forJapan’s export prospects. Slumping import de-mand in emerging Asia has been amplified bya tightening of policy interest rates in countriessuch as India, Indonesia, the Philippines, andThailand to stem inflation pressures. Duringthe third quarter, Japan’s GDP dropped by0.4 percent (saar) and for the remainder of2008, creeping spillover from the credit crunchand falling exports will cause Japan’s GDP torecede further, coming to register 0.5 percentgrowth for the year. Recession in Japan is ex-pected to be less pronounced than elsewhere,with output declining by 0.1 percent in 2009before picking up to 1.6 percent in 2010 asglobal investment demand revives and stimu-lates Japanese exports.

Outlook for the developingcountries(A deeper discussion of developments in eachof the six developing regions may be found inthe Regional Appendix in this book.)

Even before international credit channelsfroze, there were increasing signs of slow-

ing economic activity in developing countries.

Slowing growth in the high-income economies,falling equity markets, and reduced interna-tional capital flows cut into investment in de-veloping countries, while a sharp accelerationin inflation linked to the surge in commodityprices constrained consumer spending.

Industrial production (outside of China)had been robust but began to slow in mid-2008. In the rest of East Asia, the downwardshift in production growth has been sharp,dropping from 20 percent in January (saar) toa decline of 5 percent by May 2008, before re-covery to nil by September (figure 1.9).

More dramatic has been the steep recentfalloff in China’s industrial production growth,from 20 percent in July to a decline of 0.2 per-cent in September (saar). Softening exportgrowth, together with tightening microman-agement of inventories—given uncertain salesprospects—have underpinned this develop-ment. This, in turn, has carried aggregate pro-duction growth in East Asia to zero as of thethird quarter.

Output growth also faded in India and else-where in South Asia, while production in Hun-gary, Poland, Turkey, and the Baltic states beganto decline more recently. Output dynamics havealso faltered in Latin America, as production inChile, Colombia, and Mexico have dropped tonegative ground, while that in Brazil has slowed


27

210

25

Jan.2006

Jun.2006

Nov.2006

Apr.2007

Sep.2007

Feb.2008

Jul.2008

0

5

10

15

20

25

30

Source: World Bank data.

Industrial production (percentage change, saar)

Figure 1.9 East Asian countries show steepfalloff in output growth

East Asia, excluding China

ChinaEast Asia & Pacific

10363_Pg15_50:10363_Pg15_50 11/29/08 7:05 AM Page 27

sharply (figure 1.10 and table 1.3, secondpanel). Imports across developing regions arealso showing signs of easing, reflecting a soften-ing of domestic demand. And export volumesfrom emerging markets displayed fading mo-mentum over the first three quarters of 2008,notably in Latin America, as import demand inthe OECD countries declined sharply. OverallGDP growth for the developing countriesslowed from the 7.9 percent advance recordedin 2007 to 5.3 percent (annualized) during thesecond quarter of 2008 (see table 1.3 and asso-ciated notes).

Financial turmoil likely to curbinvestmentFixed investment has been a powerful drivingforce for growth across developing countriesover the last decade, particularly in East Asiaand the Pacific and in Europe and CentralAsia, increasing its contribution to overallgrowth to almost 4 percentage points in recentyears, and well outstripping contributionsfrom trade (figure 1.11). But the intensificationof the credit crisis in the United States has se-verely constrained finance to developing coun-tries, with ominous implications for growth

prospects. The effects of the global financialcrisis on developing countries will differ by re-gion and by the ability of individual countriesto offset adverse effects on domestic bankingsectors and the broader financial market.

Emerging-market corporate borrowers al-ready are seeing a sizable widening of spreadsand are increasingly being shut out of interna-tional bond markets (see discussion above). Apullback in syndicated bank lending is emerg-ing (as commercial banks and other financialinstitutions in the high-income countries shoreup balance sheets by limiting new lending orby calling in existing lines of credit), and ini-tial public equity offerings from key emergingmarkets have dried up. Even before the freez-ing of credit flows that has accompanied thebanking crisis, overall capital inflows to devel-oping economies were down 35 percent overthe first nine months of 2008 from the sameperiod a year earlier.

The slowdown is likely to be morepronounced in 2009Looking forward, recent adverse trends areanticipated to intensify, driven by an especiallysharp decline in investment growth in devel-oping countries, weaker exports as import de-mand from high-income economies declines,and lingering and in some cases still-escalating


28

210

0

25

5

10

15

20


Figure 1.10 Output growth in Latin America, South Asia, and Europe and Central Asia isfading


Jan.2006

Jun.2006

Nov.2006

Apr.2007

Sep.2007

Feb.2008

Jul.2008

South Asia

Latin America& Caribbean Europe &

Central Asia

21

0

1992 1994 1996 1998 2000 2002 2004 2006 2008a

1

2

3

4

Contribution to real GDP growth (percentage points)

Figure 1.11 Investment was the driving force for growth in developing countries


a. Projected.

Investment

Net exports

10363_Pg15_50:10363_Pg15_50 11/29/08 7:05 AM Page 28

inflation. Developing-country GDP growth isprojected to decline to 4.5 percent in 2009,more than 3 percentage points below the av-erage of the past five years (figure 1.12). Noregion or country is likely to escape thisgrowth recession. Recovery in 2010 to a 6.1percent advance is predicated upon a relativelyquick improvement in financial and growthconditions among the high-income countries, aprospect currently subject to a high degree ofuncertainty.

Growth outcomes for 2009 are anticipatedto vary significantly across developing coun-tries and regions, depending on their relianceon external flows and bank lending to financeinvestment, trade links to deeply affectedhigh-income countries, direct and indirect ex-posures to the subprime mortgage crisis, andthe degree of participation of foreign banks inthe domestic financial sector. Moreover, policyresponses to the crisis will play a large role inshaping the near-term economic outlook.


29

Table 1.3 Developing regions: growth and related indicators

Growth Seasonally adjusted Growthyear-over-year annualized growth year-over-year

Indicator 2006 2007 Q108 Q208 Q308 H108 Latest

GDP growth (percent)Developing countries 7.7 7.9 7.5 5.3 — 6.9 —

East Asia and the Pacific 10.1 10.5 9.4 9.2 — 9.0 —South Asia 9.0 8.4 11.3 2.9 — 8.3 —Europe and Central Asia 7.5 7.1 8.8 �1.2 — 5.9 —Latin America and the Caribbean 5.6 5.7 4.4 5.4 — 5.3 —Middle East and North Africa 5.3 5.8 4.0 3.8 — 4.0 —Sub-Saharan Africa 5.9 6.3 4.9 5.2 — 4.5 —

Industrial productionDeveloping countries 8.8 9.7 10.6 9.6 �2.4 9.2 5.0

East Asia and the Pacific 13.0 15.0 18.3 17.1 0.2 14.7 10.2South Asia 10.6 9.1 9.5 1.1 �4.1 6.1 1.0Europe and Central Asia 7.6 6.9 5.9 3.3 �9.8 5.1 0.3Latin America and the Caribbean 4.3 4.3 0.2 0.4 3.6 3.4 2.2Middle East and North Africa �0.8 �0.5 8.6 3.5 �0.2 3.6 3.5Sub-Saharan Africa 3.9 5.8 �2.1 15.6 �9.3 6.5 4.7

Consumer pricesa

Developing countries 6.2 6.1 8.6 10.4 9.9 9.5 9.9East Asia and the Pacific 5.1 5.3 7.7 9.5 8.2 8.6 8.2South Asia 7.6 7.6 10.1 11.0 22.0 10.6 22.0Europe and Central Asia 5.6 8.0 11.0 11.3 11.0 11.2 11.0Latin America and the Caribbean 5.6 6.5 8.8 9.7 7.5 9.3 7.5Middle East and North Africa 5.1 7.2 11.2 10.8 12.7 11.0 12.7Sub-Saharan Africa 6.2 6.0 8.2 10.4 11.3 9.3 11.3

Export volumesa

Developing countries 13.9 14.6 14.0 13.8 — 14.0 —East Asia and the Pacific 19.2 18.7 16.2 17.0 17.1 16.2 19.2South Asia 4.8 9.0 13.8 9.4 12.8 13.8 10.5Europe & Central Asia 11.8 11.8 14.3 10.3 — 14.3 —Latin America and the Caribbean 6.9 4.5 0.7 �2.5 �11.4 �0.7 �5.4

Source: National Statistical Agencies through Haver Analytics.Note: Growth rates at annual or annualized rates, unless otherwise indicated. Consumer prices for regions are medians. Quarterly2008 growth for developing regions is based on data available for key economies. No data on export volumes for South Asia andSub-Saharan Africa are available. East Asia and Pacific: China, Indonesia, Malaysia, the Philippines, and Thailand. South Asia:India. Europe and Central Asia: Czech Republic, Hungary, Poland, Russian Federation, and Turkey. Latin America and theCaribbean: Argentina, Brazil, Chile, Colombia, and Mexico. Middle East and North Africa: Arab Republic of Egypt. Sub-Saharan Africa: Nigeria, and South Africa.a. Quarterly data, year-over-year growth.

10363_Pg15_50:10363_Pg15_50 11/29/08 7:05 AM Page 29

Inflation has been rising but is nowset to declineThe projected global growth recession andmuch lower commodity prices should helpease the surge in inflation observed over 2007and 2008. Headline consumer price inflationamong the OECD countries increased from amodest 2 percent in 2007 to a 4 percent year-over-year pace in the third quarter of 2008,led by increases in the United States (4.9 per-cent), the Euro Area (3.9 percent), and theUnited Kingdom (5 percent). Although thepeak in commodity prices appears to havepassed, the momentum of headline inflationduring August picked up to 8.5 percent in theUnited States before easing in September,while falling to 2.3 percent in Japan and to3.6 percent in Europe (figure 1.13).

Consumer price inflation accelerated muchmore quickly in developing countries than inthe advanced economies (figure 1.14, andtable 1.3, third panel). In the majority of de-veloping economies, most of the increase inheadline inflation was attributable to the directeffects of higher commodity prices; increasesin core inflation (which excludes food andenergy) were limited. Indeed, inflation indeveloping countries has remained relativelylow over the past five years of rapid growth,

despite substantial increases in oil and metalsprices since 2003 (box 1.2). However, thesharp rise in food and fuel prices in the firsthalf of 2008 pushed median inflation in thedeveloping world to 12 percent by July 2008,and more than 30 countries were facingdouble-digit inflation rates.

In a welcome development, median infla-tion has since retreated to below 10 percent inSeptember, as falling commodity prices haveimproved CPI developments across a widerange of developing countries.


30

0

3

6

9

12

Median inflation rates (%)

Figure 1.14 Inflation in emerging markets surged on higher food and energy prices


Jan.2000

Jan.2001

Jan.2002

Jan.2003

Jan.2004

Jan.2005

Jan.2006

Jan.2007

Jan.2008

High-income OECD countries

Developing countries

0

1

2001 2002 2003 2004 2005 2006 2007 2008 2009 2010

2

3

4

5

6

7

8

Real GDP (percentage change)

Figure 1.12 Developing-country GDP growth is expected to fall below 5 percent in 2009

Source: World Bank.

Forecast

23

0

3

6

9

Percentage change (saar)

Figure 1.13 Headline inflation is easingacross industrial countries


Note: HICP 5 Harmonized index of consumer prices.

a. Year-over-year.

Jan.2006

Jan.2007

Jan.2008

Japan

United States

Euro Areaa (HICP)

10363_Pg15_50:10363_Pg15_50 11/29/08 7:05 AM Page 30


31

Strong growth in developing countries during the1960s coincided with a low-inflation environ-

ment, while high inflation during the subsequent twodecades coincided with low average growth (box fig-ure 1.2a). The recent sharp pickup in GDP growthoccurred in an environment of low and stable infla-tion. Although the causality is always difficult to un-tangle, the potential adverse impact of high inflationon the ability to interpret price signals, on disciplinedfiscal management, and on savings and investmentare well understood.

These negative effects underscore the importanceof preventing an acceleration of inflation beyond thedirect impact of increased commodity prices.

The relatively low inflation of the recent past canbe clearly illustrated at an aggregate level with amodel that explains median inflation in the developingworld as stemming from commodity price increasesin local currencies (measured as a median) andpersistence (inflation tends to depend on pastinflation as a result of indexation and as inflationaryexpectations are adjusted). In the model, CPItdenotes consumer price inflation in developingcountries, Pt denotes annual commodity prices, bothin (median) local currencies for the 1962–2008period. Numbers in parentheses denote t-ratios, and� denotes the first-difference operator:

Box 1.2 Commodity prices and inflation indeveloping countries

�log(CPIt)�0.01�0.09 �log(CPIt�1)�0.79 �log(Pt),(1.39) (5.09) (11.60)

adjusted R2 � 0.78.

Box figure 1.2b shows a dynamic simulation ofthe equation since the 1960s.

Inflationary pressures during the 1970s are wellexplained by a combination of commodity price in-creases and persistence. During the 1990s, however,many developing countries experienced high inflationunrelated to commodity prices and therefore not ex-plained by the estimated equation. These increases ininflation were caused more by loose policy reactionsto debt crises, especially in Latin America, and thetransition toward market economies in Europe. Inthe recent period, inflation has actually remainedlower than the model would predict, largely as a re-sult of institutional reforms, which made monetarypolicy more independent in many countries and in-flation targeting more prevalent.

The literature provides a range of explanationsfor the relatively low inflation of recent years.Dominant in the debate is the role of inflation expec-tations, which have been brought down sharply byinstitutional reforms. Increased competition in globalmarkets, making it more difficult to pass throughincreases in the higher costs of production, is anotherexplanation. Moreover, the additional low-cost

01965 1970 1975 1980 1985 1990 1995 20052000 2010

4

8

12

16

Percentage change(5-yr moving average)

Percentage change(5-yr moving average)

Box figure 1.2a Inverse long-term correlationbetween inflation and growth

Source: World Bank.

Developing-country realGDP growth (right axis)

Developing-countrymedian inflation (left axis)

2

0

4

6

8

0.00

Box figure 1.2b How do international commodity prices explain inflation indeveloping countries?

1961 1966 1971 1976 1981 1986 1991 20011996 2006

0.05

0.10

0.15

0.20

Source: World Bank.

Dynamicsimulation

Actual consumption index

Log scale

10363_Pg15_50:10363_Pg15_50 11/29/08 7:05 AM Page 31

Many countries that experienced thesharpest run-up in inflation have been sub-jected to a dangerous combination of escalatingfood and energy prices and generally tight con-ditions in domestic markets, caused by a rapidincrease in credit creation. Of the 24 countrieswhere inflation picked up by more than 5 per-centage points within the past year, 10 are inEurope and Central Asia, where inflation wasspurred by very strong capital inflows orbooming commodity revenues. Other countriessubject to strong domestic inflationary pres-sures include Bolivia, Chile, the Philippines,República Bolivariana de Venezuela, andVietnam. Some of the biggest jumps in head-line inflation were seen in Sub-Saharan Africancountries, but that is largely because foodrepresents more than 50 percent of consump-tion in many African countries.

The ramp-up of commodity prices over2006–08 and the associated acceleration ofinflation have posed difficult policy challenges.The continued practice of official subsidization

of fuels and basic foodstuffs across manycountries in the Middle East and North Africa,and East and South Asia is contributing towider fiscal deficits, narrowing the room pol-icy makers have for maneuver in other areas,including targeted income support, investmentin the Millennium Development Goals, andcountercyclical fiscal policy. Many monetaryauthorities have faced a trade-off betweensupporting growth and dampening inflationand inflation expectations. Brazil, Indonesia,Mexico, the Philippines, South Africa, andThailand have raised policy rates by 25 basispoints or more. The recent fall in commodityprices and the global slowdown are likely toease this difficulty over time, and indeed, ashift toward monetary accommodation is nowunder way to mitigate a portion of the growth-dampening effects of the financial crisis.

Regional outlooksGDP in East Asia and the Pacific increased by8.5 percent in 2008, down from 10.5 percent in


32

clines in commodity prices, headline inflation willease gradually, even if core inflation moves up mod-erately. But the probability of higher inflation is cer-tainly not negligible.

supply from developing countries, notably China,in global markets carried deflationary effects. Andthe share of commodities in world production andtrade has declined over time, as a result of which theimpact of commodity price increases is now substan-tially less than during the 1970s (box figure 1.2c).

On the other hand, many of the factors that havekept inflation low over the last five years may havelost a degree of efficacy. Indeed, with the recent risein commodity prices, the share of commodities in theglobal economy, and with that, their effects on thegeneral price level, is increasing rapidly. Many fast-growing developing countries have reached capacityconstraints in infrastructure, energy, and other inputsto production, and the increase of low-cost goods inglobal markets is waning. With broader inflationrates rising, it becomes easier to pass through costincreases, and, importantly, low inflation expecta-tions might be revised upward quite quickly. Theseare serious challenges to be faced to prevent areemergence of a high-inflation environment. In thebaseline forecast, we assume that as a result of thesharp global growth slowdown and the recent de-

01963 1968 1973 1978 1983 1988 1993 1998 2003

10

5

15

20

25

30

Nominal shares (%)

Source: World Bank.

Box figure 1.2c Commodities are a declining share in global merchandise trade

Oil

Ores and metals

Agriculture

10363_Pg15_50:10363_Pg15_50 11/29/08 7:05 AM Page 32

2007. (Excluding China, growth in the regionfell to 5.3 percent, from 6.2 percent in 2007.)Of importance to the slowing pace of growth,China’s GDP growth in the third quarter easedto 9 percent, from 10.6 percent in the firstquarter, on a slump in investment and exports.Rising oil and food prices boosted median in-flation in the region to 9 percent in 2008, com-pared with 5 percent over the preceding2 years. The deterioration in the outlook forJapan and the United States reduced exportgrowth, which for East Asian countries outsideof China fell from 10.5 percent in 2006 to4 percent in 2008 (figure 1.15). In turn, manu-facturing output fell from 5 percent growth in2007 to a decline of the same magnitude in2008. And gross capital flows fell by a third, to$64 billion over the year to September 2008.

Prospects for 2009 and 2010 have dimmedwith the deterioration in the external environ-ment. The global banking crisis has had littledirect effect on the region, but several countriesare more vulnerable to spillovers in the form ofhigher corporate spreads, reduced capitalflows, and plummeting domestic equity mar-kets. Private sector investment in particularstands at risk. Although China is well cush-ioned against coming shocks, several countriesremain exposed to a steep downturn in world

trade and more difficult financing conditions.While the decline in oil and food prices willsupport external positions and provide somerelief on the inflation front, reduced investmentspending is expected to contribute to a substan-tial slowdown in regional growth to 6.7 percentin 2009. A gradual recovery in key foreign mar-kets will offer fresh impetus for exports andproduction and will help return growth in theregion to a 7.8 percent pace by 2010.

In Europe and Central Asia, output is likelyto increase by 5.3 percent in 2008, down from7.1 percent in 2007, though growth held up re-markably well during the first half of the year.First-half GDP growth in Russia (7.8 percent),Poland (6 percent), Turkey (5.8 percent), andRomania (8.8 percent) were grounded in strongdomestic demand, along with higher oil pricesand fiscal revenues for the region’s hydrocar-bon exporters. But in 2009 deteriorating exter-nal positions and new risks from the globalbanking crisis are likely to depress prospects forvulnerable countries, and the downside risksare substantial. Sovereign spreads jumped inOctober 2008, notably for the Russian Federa-tion and Turkey, and especially for Ukraine andKazakhstan (figure 1.16).


33

0

300

150

450

600

750

900


Figure 1.16 Sovereign bond spreads widen across Europe and Central Asia


Jul.2007

Oct.2007

Jan.2008

Apr.2008

Oct.2008

Jul.2008

Bulgaria

Poland

Turkey

Hungary

Russian Federation

26

23

0

3

6

9

12

Percentage change

Figure 1.15 Key developments in 2008 for East Asia and the Pacific (excluding China)

Source: World Bank.

2006 20

07 2008

Exportgrowth

Industrialproduction

Medianinflation

Medianpolicy rate(interest

rate)

Termsof trade

GDPgrowth

10363_Pg15_50:10363_Pg15_50 11/29/08 7:05 AM Page 33

While most countries have maintained for-ward momentum, a divergence in growth per-formance has emerged: Latvia is in recession,the Romanian economy is overheating, andthe Kyrgyz Republic, Tajikistan, andMoldova, which are receiving World Banksupport, are facing the impact of rising foodprices. The situation in Russia shifted dramat-ically from concerns about domestic overheat-ing to fears of financial crisis, as equity pricesgyrated with the turmoil in global markets,and oil prices fell.

Most countries have experienced stronggrowth in domestic demand, but net trade hasremained a drag on growth. At the same time,rapid credit expansion and wage escalationhas made the region more vulnerable to deteri-oration in external financing conditions. Themedium-term outlook points to a sharp declinein regional growth to 2.7 percent in 2009, dri-ven by a falloff in investment tied to difficultfinancing conditions and a marked weakeningin export market demand. Growth is projectedto firm to 5.0 percent by 2010, as credit marketsstabilize, inflation pressures ease, and growthin external markets resumes, paving the wayfor a revival in spending and exports.

Latin America and the Caribbean countrieshave enjoyed four years of robust growth,while current account surpluses, accumulationof reserves, and improved policies have servedto restrain core inflation rates, improve thequality of banking systems, and build up sub-stantial buffers against financial contagion.However, in 2008 headline inflation jumpedin response to higher oil and food prices, andpolicy makers in countries such as Brazil andChile raised interest rates. Gross capital in-flows to the region compressed by 45 percentbetween January and September 2008, com-pared with the same period in 2007. The deteri-oration in external demand and in internationalfinancial markets, combined with the recentfalloff in commodity prices, reduced GDPgrowth to 4.4 percent in 2008 from 5.7 percentin 2007.

The global slowdown and a shortfall in cap-ital flows present substantial risks to sustained

growth, pressuring private sector investment inparticular. As commodity prices continue toweaken, some major exporters, Argentina ofnote, will likely see current account surplus po-sitions shift to deficit. For other countries, in-cluding Brazil and Mexico, the depth of reces-sion in the United States and Europe will turnexports to negative growth, while contractionin imports should lead to a return of surplusposition (figure 1.17).

GDP growth in the region is expected todrop to 2.1 percent in 2009 before recoupingto 4 percent by 2010. Country-specific eventscould also pose a challenge: conditions in sev-eral Andean countries have tended toward lessstability; República Bolivariana de Venezuelahas seen another wave of nationalizations, andits growth is expected to fall from 8.4 per-cent in 2007 to 3.2 percent by 2010; andArgentina’s GDP is expected to slow from8.7 percent in 2007 to 4 percent by 2010, witha 1.5 percent growth trough in 2009.

The developing countries of the MiddleEast and North Africa region offer a good ex-ample of the diversity of effects stemming fromthe ramp-up in global fuel and food prices—atboth extremes of the spectrum. In oil-export-ing economies, a rise in oil and natural gasrevenues to $200 billion supported 5.8 per-cent growth in 2008, down from 6.4 percent


34

24

22

23

21

0

Brazil Argentina Mexico

1

2

3

4

Current account balance as a share of GDP (%)

Figure 1.17 In Latin America and the Caribbean, current accounts of largest economies diverge

Source: World Bank.

2006

2007

2008

2009

2010

10363_Pg15_50:10363_Pg15_50 11/29/08 7:05 AM Page 34

in 2007. Slower domestic demand growth inthe Islamic Republic of Iran (where GDPgrowth declined from 7.8 percent to 5.6 per-cent) was a key factor in this development.Outside Iran, growth among oil exportersstepped up to 5.9 percent.

In the more-diversified economies that arehighly dependent on oil and food imports, ex-ports slowed in 2008 as growth turned slug-gish among key trading partners in Europeand the United States. But strong recovery inMorocco following drought in 2007 and con-tinued solid performance in Tunisia and Jordanboosted growth to 5.7 percent from 3.8 per-cent. For the region as a whole, these develop-ments yielded a flat profile of activity at 5.8 in2008 (figure 1.18).

The region’s oil exporters will face the chal-lenge of diminished revenues in 2009. Theglobal oil price is anticipated to fall from itsJuly 2008 peak ($145/bbl) to below $80 in2009. Growth in oil-exporting countries isprojected to slow to 3.9 percent in 2009. Al-though their economies are unlikely to be se-verely affected by developments in financialmarkets, several countries stand more exposedto spillover effects from these developments,including Lebanon, Jordan, and the Arab

Republic of Egypt. Moreover, oil-exportingmembers of the Gulf Cooperation Council arevulnerable to losses on international invest-ment positions, potentially prompting a hiatusin the recent buildup of foreign direct invest-ment within the region. For the region’s oil-importing economies, lower energy prices willreduce the import bill and provide somebreathing space on the inflation front follow-ing the surge in prices in the first half of 2008.Growth for the region should recover to a5.2 percent pace by 2010 as external demandrecovers, and declines in hydrocarbon pricesgive way to a period of greater stability.

GDP growth in South Asia eased to an esti-mated 6.3 percent in 2008, from 8.4 percent in2007 and from a 25-year high of 9 percentduring 2006. High food and fuel prices,tighter international credit conditions, andweaker foreign demand have led to worseningexternal accounts and contributed to a slow-down in investment growth. Deterioration intrade balances, however, has been offset inlarge part by large remittance inflows, partic-ularly for Bangladesh, Nepal, and Sri Lanka,where remittances represent 8 percent ofGDP or more. Policy makers responded tohigh commodity prices and rising inflationpressures by partially adjusting domestic fuelprices, cutting development spending, and(initially) tightening monetary policy.

The global financial crisis is placing furtherdownward pressure on growth. Lower capitalinflows (down 40 percent, over January-September 2008 compared with the sameperiod in 2007) and harder credit terms willreduce private investment, while reduced remit-tance inflows will add to pressures on growth.Food and fuel price subsidies have pushedfiscal outlays higher, reversing recent progressin fiscal consolidation. And increasing deficitsare narrowing the scope to provide supportfor other urgent public programs, includingthe region’s overburdened infrastructure. Theslowdown in growth is most apparent in Indiaand Pakistan, where industrial production fellsharply, and the momentum of production forSouth Asia has recently declined from a peak


35

2

0

4

6

8

30

20

50

90

100

70

80

60

40

110

GDP growth (%)

Figure 1.18 Oil revenues, recovery from drought underpin growth in the Middle East and North Africa in 2008


2000 2001 2002 2003 2004 2005 2006 2007 2008

Oil prices (right axis)

Oil exporters GDP (left axis)

Diversified economies GDP (left axis)

World Bank average oil price ($/bbl)

10363_Pg15_50:10363_Pg15_50 11/29/08 7:05 AM Page 35

of 12 percent in April 2008 to a decline of2 percent in August (saar) (figure 1.19).

South Asia’s GDP is expected to drop to5.4 percent in 2009 but to recoup to 7.2 per-cent by 2010. Firming growth will be sup-ported by improving external demand andlower commodity prices.

Growth in Sub-Saharan Africa, outside ofSouth Africa, increased to a remarkable 7 per-cent in 2007, the highest in some 35 years, asoutcomes for both oil-importing and oil-exporting countries were robust. Growthamong oil exporters increased to 8.2 percentin 2007, exceeding 5.5 percent gains for afifth year running; growth in oil-importingeconomies breached a 25-year record, gaining5.4 percent. GDP advances have become morebroad-based and less volatile in recent years,especially among oil importers. And a notableand encouraging feature of Africa’s recent per-formance is the sustained contribution of in-vestment to overall GDP growth.

In 2008, activity outside of South Africa re-mained strong at 6.6 percent as GDP gainsamong oil-producing countries eased moder-ately to 7.8 percent, joining the larger group ofoil-importing countries where GDP gainsslowed to 4.2 percent in the year. Investmentcontinues to provide an underpinning for GDPgrowth, while net exports are contributing to

growth despite a sharp slowdown in globaltrade (figure 1.20).

The region’s growth is expected to decline to4.6 percent in 2009, before firming to 5.8 per-cent by 2010 as a result of recovery in exter-nal demand. Excluding South Africa, growthis anticipated to ease to 5.7 percent in 2009and to accelerate to 6.6 percent in 2010. Butthis scenario is subject to significant downsiderisks. Should the global slowdown prove muchdeeper than anticipated, fostering a sharp fallin world trade growth, the contribution of netexports to African GDP growth will diminish.And many countries in the region have be-come more vulnerable to terms of tradeshocks, as high fuel and food prices have ledto a deterioration in external positions overthe past years. Higher food and fuel priceshave also widened the poverty gap, raising therisk of possible social unrest.

World trade

World trade volumes are expected to con-tract in 2009 for the first time since

1982 (figure 1.21). This decline is driven firstand foremost by a sharp drop in demand, asthe global financial crisis imposes a rare simul-taneous recession in high-income countries anda sharp slowdown across the developing


36

401976 1981 1986 1991 1996 2001 2006

50

60

70

80

90

Primary commodities as a share of total exports (%)

Source: UN Conference on Trade and Development, Comtrade.

Sub-Saharan Africa

Sub-Saharan African oil-importing countries

Figure 1.20 In Sub-Saharan Africa, primary commodity exports increased as prices surged

25

0

5

10

15

20


Figure 1.19 South Asian production slips inthe last months

Source: World Bank.

Jan.2006

Jan.2007

Jan.2008

South AsiaDevelopingcountries

High-income OECD countries

10363_Pg15_50:10363_Pg15_50 11/29/08 7:05 AM Page 36

world. The global credit crunch is likely to af-fect private investment especially, which is themost cyclical and most internationally tradedcomponent of GDP. At the same time, thecredit crunch is restricting export finance. Al-ready there is anecdotal evidence that commer-cial bank trade credits are drying up and thatexport receipts are becoming more difficult toinsure. Similarly, exporting firms may cut backon shipments if their access to credit lines islimited. Finally, the crisis has been associatedwith sharp, unpredictable swings in exchangerates, which also will hamper trade.

Signs of an economic slowdown have beenvisible for some time (growth in U.S. importdemand was already falling in 2005), but theyhave been building at a much more gradualpace than occurred, for example, during thesharp correction in 2001, when importgrowth dropped from plus 15 percent tominus 5 percent within a year (figure 1.22).The current gradual decline in demand growthmeans that some exporting countries have hadtime to shift to higher growth markets in de-veloping countries. For example, while theshare of the United States in India’s exportsfell from 17.1 percent in 2004 to 15.3 percentin 2007, China’s share in India’s imports rosefrom 5.5 percent to 8.4 percent.

In addition, several countries (many inLatin America) that experienced a slowing ofexports because of low U.S. demand growthbenefited from higher commodity prices.Moreover, the impact on the volume of worldtrade was mitigated by the strong intra-regional growth of trade in East Asia, largelydriven by China’s continuing integration intoglobal markets. China’s import and exportgrowth continued to exceed 20 percent overthe past two years; while outside of China, ex-port growth remained robust (figure 1.23).


37

0

1981 1984 1987 1990 1993 1996 1999 2002 2005 2008

6

23

3

9

12

Growth of global trade volumes (percentage change)


Figure 1.21 World trade is expected to decline in 2009 for the first time since 1982

0

Jan.1999

Jul.2000

Jan.2002

Jul.2003

Jan.2005

Jul.2006

Jan.2008

25

5

10

15

10

0

20

30

40

Import volumes(percentage change)

Import volumes(percentage change)


Figure 1.22 Decline in high-income importgrowth affects developing-country exports

U.S. imports(left axis)

Latin America &Caribbean exports(left axis)

China exports (right axis)

210

25

0

5

10

15

20

25

30

35

Export volumes of developing countriesa (percentage change)

Figure 1.23 Developing-country exportshave been strong, even outside China


a. Excluding China.

Jan.1992

Jan.1996

Jan.2000

Jan.2004

Jan.2008

Annual (12-month moving average)

Momentum (annualized 3 month over 3 month)

10363_Pg15_50:10363_Pg15_50 11/29/08 7:05 AM Page 37

These mitigating and compensating factorsare unlikely to be active in 2009. The slow-down in high-income import demand has ac-celerated, few growth centers are left to whichexports can be redirected, and commodityprices are falling. Thus developing countriesare set to experience a sharp, but likely tem-porary, fall in their export revenues. Thosecountries with insufficient reserves to sustainimport growth will need to rely on some com-bination of exchange rate depreciation andslower growth to restrain imports.

Remittance flows to developing countries,which reached an estimated $283 billion in2008 (Ratha, Mohapatra, and Xu 2008),began easing in the second half of the year andare projected to slow sharply in 2009. Mi-grant earnings in host-country currency termsare anticipated to be compressed by the reces-sion in the industrial economies, lower rev-enues in high-income oil-exporting countries,and slower growth in many developing coun-tries that are destinations for migrants.

While the baseline projection is for remit-tance flows to developing countries to declineas a share of recipient-country GDP from 1.8to 1.6 percent in 2009, the extent of the de-cline at the country level will depend criticallyon exchange rate developments. Recentswings in bilateral exchange rates have out-weighed the expected change in remittancesdenominated in host-country currencies. Fu-ture exchange rate movements will also playan important role.

Global current account balances areexpected to show substantial shiftsThe global recession and attendant decline inworld trade and in commodity prices will havea dramatic impact on current account balances.Surplus positions in Japan and the Euro Areashould increase to $240 billion and $180 bil-lion, respectively, during 2009 as commodityprices fall and trade volumes compress. Despitethe improvement in the U.S. terms of trade, itscurrent account shortfall is expected to deterio-rate from $770 billion in 2008 (5.4 percent ofGDP) to $830 billion or 5.8 percent of GDP in

2009. The sharp downturn in world trade hitsthe United States especially hard: export vol-umes are expected to drop 2.6 percent in 2009,while imports contract 1.1 percent. Overall, thehigh-income OECD countries’ current accountdeficit is projected to narrow by $185 billion to$375 billion in the year.

The lower industrial-country deficit has itscounterpart in lower surpluses in high-incomeoil exporters and in developing countries. Thecurrent account surplus in developing countriesis expected to fall from a peak $500 billion, or3.7 percent of GDP, in 2007 to $333 billion, or2 percent of GDP in 2009 (figure 1.24). WhileChina’s (and thus East Asia’s) surplus is antici-pated to increase, in other regions, surpluses areexpected to narrow, or deficits to widen.

Some appreciation of the impact of theglobal recession on current account balancescan be gained by looking at countries groupedby their primary commodity trade (here weexclude China, because the country’s massivecurrent surplus—nearly $400 billion in2008—masks underlying developments acrosssmaller countries). The expected 26 percentfall in the price of oil and 23 percent fall innon-oil commodity prices (see the Commodity


38

2200201020082006200420022000

2100

0

100

200

300

400

500

600

Current account balance (US$ billions)


Figure 1.24 Developing-country current account surpluses to wane after 2008

All developingcountries

East Asia & Pacific

Middle East & North Africa

South Asia

Europe & Central Asia

Latin America & Caribbean

10363_Pg15_50:10363_Pg15_50 11/29/08 7:05 AM Page 38

Markets section below) will shift the terms oftrade in favor of oil- and food-importingcountries, following a string of at least fiveyears of substantial losses. For developing-country oil exporters, the fall in global de-mand will reduce both oil prices and exportvolumes (which are expected to shift from5.1 percent growth in 2008 to 0.3 percent in2009), and their current account surplus willfall from a peak of 6.4 percent of GDP in 2008to 1.4 percent in 2009 (figure 1.25). By con-trast, the current account deficit of developing-country importers of both oil and food almosthalves, from a peak of 8.1 percent of GDP in2008 to 4.3 percent by 2009, because of lowercommodity prices and a sharp slowdown ofimports from 8.1 percent growth in 2008 to1.5 percent in 2009. The impact of these globaldevelopments should begin to dissipate in2010 as oil prices stabilize, the prices of nonen-ergy commodities decline by only 4.3 percent,and world trade begins to recover.

The financial crisis has induced major fluc-tuations in exchange rates during the autumnof 2008. Almost every currency in the worldhas depreciated against the dollar and the yen,reflecting a “flight to quality” into U.S. Trea-sury securities, the unwinding of yen-basedcarry trades, and the deleveraging of banks,

firms, and investors. No developing-countrycurrency has appreciated against the dollar bymore than 0.5 percent during this period. Onaverage, currencies of developing countrieshave fallen by 15 percent against the dollar, buthigh-income-country currencies (save Japan’s),have also depreciated (figure 1.26). In generalthe competitiveness of the United States, Japan,China (whose currency has held steady versusthe dollar), and those countries whose curren-cies have been pegged to the dollar will havebeen reduced; competitiveness for countrieswhose currencies have depreciated will be im-proved in these markets. As a result, the strongimpetus that net exports have provided forU.S. growth is likely to be attenuated; at thesame time, net exports are likely to support thegrowth recovery of many developing countries.

Commodity markets

Commodity prices—which have been trend-ing higher since 2003—continued the ro-

bust rise that began in 2007 into the first halfof 2008. As of mid-November, prices havesince fallen sharply, giving up most of theirgains of the first half of the year. The abruptdecline reflects a classic response of commodi-ties to slowing global growth at the end of aboom (for more on this subject, see chapter 2),a decline that has been amplified and acceler-ated by the financial crisis. In the summer of


39

Figure 1.25 Current account balances forcommodity-exporting and -importingdeveloping-country groups (excludingChina)

210

28

22

24

26

0

2

4

6

8


Oil and foodimporters

2007

2008

2009

2010

Oil exporters

Current account balance as share of GDP

Oil importers /food exporters

Oil importers

0

2

4

6

8

10

12

14

Source: World Bank.

% depreciation since Sept.15 –Oct. 30

High-income

countries

EastAsia &Pacific

Europe &

CentralAsia

LatinAmerica

&Caribbean

MiddleEast &North Africa

SouthAsia

Sub-Saharan

Africa

Figure 1.26 Almost all currencies havedepreciated against the dollar

10363_Pg15_50:10363_Pg15_50 11/29/08 7:05 AM Page 39

2008, energy prices were 80 percent higher indollar terms than a year earlier, while nonen-ergy prices were 35 percent higher (figure 1.27and table 1.4).

Almost all the advance in nonenergy com-modity prices during 2008 came from grains(up 60 percent), fats and oils (up 34 percent),and fertilizers (up 140 percent). Metals prices,which increased rapidly between 2003 and2008, picked up just 8 percent over the first sixmonths of 2008. Almost all commodity pricespeaked in early or mid-2008, and most have de-clined sharply since then. Crude oil pricesdropped from $143/bbl in early July to less than$50/bbl in mid-November. The price dropstemmed from weaker realized demand across

OECD member countries, appreciation of thedollar, and concerns about demand prospects inthe wake of financial turmoil (figure 1.28). Thesharp decline in crude oil prices has also been asignificant contributor to declines in other com-modities, because thesemarkets are increasinglylinked through production costs and throughthe development of biofuels (see chapter 2).

Falloff in demand in high-incomecountries drives decline in oil pricesOil demand in the OECD countries has beendeclining for three years, with most of the re-duction in the United States, which was af-fected by slowing economic activity and theconsumption-dampening effects of higher oiland gasoline prices. U.S. oil demand fell 5.6 per-cent over the first 10months of 2008 (year-over-year). Gasoline consumption dropped 3 per-cent as consumers reduced the number of milesdriven and began switching to more energy-efficient vehicles. Oil demand also slowed inEurope. Overall, OECD demand is expected tofall by more than 2.2 percent during 2008 andby less than 2 percent in 2009. Demand in de-veloping countries and emerging markets hascontinued to grow by about 4 percent, with de-mand strongest in Asia and theMiddle East, thelatter fueled until recently by strong economicgrowth and in some countries fuel subsidies andthus low consumer prices (box 1.3).


40

100Jan.2005

Jan.2006

Jan.2007

Jan.2008

125

150

175

200

225

250

275

300

325

100

150

200

250

300

350

400

450

500


Figure 1.27 Commodity prices surged beforeretreating in the second half of 2008

Commodity price indexes(2000 5 100, current US$)

Commodity price indexes(2000 5 100, current US$)

Non-energy prices(left axis)

Energy prices(right axis)

Source: Datastream and World Bank.

World Bank average crude oil price ($/bbl)

Figure 1.28 Crude oil prices correct sharplyafter unprecedented run-up

25

50

100

75

125

150

Jan.2004

Jan.2005

Jan.2006

Jan.2007

Jan.2008

Table 1.4 Forecast of commodity prices

Percent change 2000–05 2006 2007 2008 2009f 2010f

Energy 13.5 17.3 10.8 45.1 �25.0 0.9Oil 13.6 20.4 10.6 42.3 �26.4 1.8Natural gas 10.4 33.9 1.0 57.2 �10.8 �4.2Coal 12.7 3.1 33.9 97.8 �23.1 �10.0

Nonenergy 8.3 29.1 17.0 22.4 �23.2 �4.3Agriculture 6.0 12.7 20.0 28.4 �20.9 �1.3

Foods 6.0 10.0 25.6 35.2 �23.4 �0.3Grains 4.8 18.4 26.1 50.9 �27.7 2.6

Raw materials 5.0 22.7 9.0 13.0 �14.9 �2.7Metals andminerals 12.3 56.9 12.0 5.0 �25.5 �5.5

Copper 15.2 82.7 5.9 �0.6 �32.2 �4.2

Source: World Bank data.f. Forecast.

10363_Pg15_50:10363_Pg15_50 11/29/08 7:05 AM Page 40


41

Deterioration in external balances has become asignificant problem for many developing coun-

tries. Overall, in 2007, current account deficits ex-ceeded 10 percent of GDP in about one-third of de-veloping countries, up from about one-quarter in2006. Twelve countries ran deficits in excess of 20percent of GDP in 2007. In part, these deficits reflectthe impact of higher oil prices on oil-importing coun-tries: oil balances account for more than half of thecurrent account deficit in one of every two develop-ing countries. Excluding the massive rise in China’ssurplus, the current account deficit of oil-importingcountries has increased significantly during the risein prices, from close to zero in 2002–03 to about$130 billion in 2007, an amount equal to 2.2 per-cent of GDP (box figure 1.3a). In contrast, the cur-rent account surplus of oil-exporting countries im-proved from 2 percent of GDP to more than 7percent in 2005–06, though it declined to below 5percent in 2007.

The rise in oil prices has contributed to, but doesnot fully explain, this disparity in current accountbalances. The rise in oil-importing countries’ deficitshas been less than the increase in their net oil bal-ance, while the increase in oil-exporting countries’surplus has been well below the improvement intheir oil balance (box figure 1.3b). As should beexpected, many countries have made compensating

Box 1.3 Impact of commodity prices on externalbalances and capital flows

adjustments in trade and domestic absorption to ac-commodate the rise in oil prices. And there is littlecorrelation between the size of countries’ net oil bal-ance and the size of current account balances. Sev-eral oil-importing countries continue to run sizablecurrent account surpluses (exceeding 10 percent ofGDP in four countries), whereas several oil-exportingcountries are running sizable current account deficits(notably Kazakhstan and Sudan). Some countries(Botswana, Nepal, Paraguay, Swaziland, and Thai-land) managed to run current account surpluses eventhough their deficits on the oil component of thetrade balance exceeded 5 percent of GDP.

How have countries been able to finance theirlarge external imbalances?The financing of increased current account deficitshas not come principally from higher portfolio flowsor reserve drawdowns, but from foreign direct in-vestment and aid. Much of the surge in private debtand equity flows to developing countries over thepast few years has gone to countries with sizablecurrent account surpluses. For example, private debtflows to the 11 countries with current accountsurpluses in 2005–07 accounted for half of the totalto all developing countries (in 2007, Russia ran acurrent account surplus equal to 6 percent of itsGDP and yet received $125 billion in private debt

2200

2100

0

20012000 2002 2003 2004 2005 2006 2007

100

200

300

600

500

400

US$ billions

Box figure 1.3a Oil balances in developingcountries, 2000–07

Oil exporters

China

Other oil importers

Source: World Bank.

25

0

20012000 2002 2003 2004 2005 2006 2007

10

15

5

GDP (%)

Box figure 1.3b Oil balances as a share ofGDP in developing countries, 2000–07

Oil exporters

China

Other oil importers

Source: World Bank.

10363_Pg15_50:10363_Pg15_50 11/29/08 7:05 AM Page 41

OPEC producers—reluctant to raise outputsignificantly during the run-up in prices dur-ing the first half of 2008—increased produc-tion at midyear, mainly through Saudi Arabia,which unilaterally agreed in June to lift outputby 0.5 million barrels a day (mb/d). Iraq’s out-put breeched 2.5 mb/d for the first time since2001 as attacks on infrastructure subsidedsomewhat. But increases elsewhere in the Gulfwere partly offset by declines in Nigeria,where civil strife continued to cause largesupply disruptions.

As discussed in chapter 2, the non-OPECsupply response has been disappointing. Pro-duction has been plagued by several factorsduring the current decade, notably rising costsand taxes, and the ongoing depletion of agingfields. Despite these constraints, non-OPECsupplies are beginning to increase in a numberof regions and are projected to rise in thesecond half of 2008 and over 2009–10.

As a result of weakening demand and ex-pected supply increases, oil prices are antici-pated to average $75/bbl in nominal termsduring 2008–10, implying a cumulative realdecline of more than 30 percent.

Prices for many metals are fallingon the back of weaker demandSeveral metals prices have plummeted in recentmonths because of slowing global growth andimproving supply prospects. Nickel prices havefallen more than three-quarters from their 2007peak, partly because of difficulties in the auto-mobile and construction sectors. Prices havefallen below the marginal costs of high-costproducers, and some plants are being closedand new projects delayed or reconsidered. Zincprices have fallen almost as much as nickel inpercentage terms. Lead prices have fallen morethan 60 percent on improving supply prospects.Prices for these metals are expected to declinefurther as new capacity comes online.

Copper is among the few metals whoseprice remained elevated during the first half of2008, despite weak demand; numerous supplydisruptions tied to strikes in Latin Americaand delays bringing on new capacity kept cop-per prices high. However, prices plunged inthe second half of the year in the wake of thefinancial crisis and the weakening global eco-nomic environment. China’s import demandhas been weak in 2008, and the slowdown in


42

foreign direct investment covered over half of thecurrent account deficit, and in 4 countries net officialdevelopment assistance (ODA) disbursements ex-ceeded 10 percent of GDP (box table).

flows). And to date few countries have had to drawdown their ample foreign reserve holdings. Instead,of the 13 countries with the largest current accountdeficits (and where data are available), in 8 countries

Developing countries with large external imbalances, 2007(percent)

Country Current account Oil balance Non-oil commodity balance Commodity balance FDI inflows Net ODA disbursements

Burundi �37.6 �11.6 �0.4 �12.0 0.0 46.0Seychelles �32.7 �36.5 0.7 �35.8 18.8 1.8Togo �21.9 �8.5 4.6 �4.0 3.5 3.6Nicaragua �21.5 �13.1 6.5 �6.6 0.1 13.8Latvia �20.8 �3.2 1.2 �2.0 7.3 0.0Georgia �20.0 �5.5 �1.8 �7.3 16.9 4.7Fiji �18.8 �11.7 2.9 �8.8 11.9 1.8Malawi �18.6 �4.3 11.6 7.3 — 21.3

Source: World Bank.Note: — � not available. For Burundi, Fiji, Malawi, Seychelles, and Togo, data are for 2006.

10363_Pg15_50:10363_Pg15_50 11/29/08 7:05 AM Page 42

global housing construction more broadlycontributed to diminishing demand. Pricesnonetheless remain well above productioncosts and are expected to continue to declinesignificantly through 2010 as new capacitycomes online.

Aluminum—the one major metal whoseprice has not surged during the current cyclebecause of the growth of capacity in China—became more expensive recently because ofstill-strong global demand and increasingcosts of electricity, a major input to the pro-duction of aluminum. The outlook for alu-minum prices depends critically on the pace ofinvestment in new capacity (especially inChina and the Middle East), as well as on thelevel of energy costs and deregulation ofpower markets. Even if new capacity is con-centrated in areas with stranded, low-cost en-ergy sources, such as the Middle East, there islimited downside potential for prices, becausealuminum has been fluctuating near the upperportion of the cost curve.

Taken together, the index of metals andminerals prices is projected to fall 25 percentin 2009 and an additional 5 percent in 2010compared to 2008.

Prices of agricultural commodities arefalling sharply from peaksPrices for food traded internationally increasedalmost 60 percent during the first half of 2008in dollar terms, with basic staples such as grainsand oilseeds showing the largest increases.Wheat prices more than doubled, from $200 aton to $440 between March 2007 and March2008, while rice prices almost tripled in the fourmonths ending April 2008 (figure 1.29). Soy-bean and palm oil prices increased 44 percentfrom 2007 to 2008. Prices have since declinedsharply. Wheat prices, for example, fell to lessthan $240 a ton in November. Since their peakin April 2008, grain prices have declined bymore than 30 percent. The spike in rice prices inApril and May 2008, on concerns regarding theadequacy of global food supplies and exportbans, appears to have subsided, with pricesfalling from nearly $1,000 a ton to $550 a ton

in November. Export bans that had been inplace were either eliminated by many countriesor partly circumvented through bilateralagreements. For example, Egypt, which hadaccumulated 7 million tons of rice during theperiod of its export ban, is expected to curb theintervention soon. Similarly, India has allowedshipments of non-basmati rice to four Africannations.

Oilseed prices also have fallen sharply. Palmoil prices averaged less than $480 a ton inNovember, down from $1,250 a ton in March2008. Similar declines took place in most edi-ble oils (soybean oil dropped from $1,475 aton to $835, and coconut oil from $1,470 aton to $705 over the period). The weakeningof edible oil prices reflects not only slowingeconomic growth but also improved supplies,and perhaps mounting pressure in the Euro-pean Union (EU) to scale back biofuel man-dates—most of the EU’s biofuel production isbiodiesel, whose feedstock is rapeseed oil, aclose substitute for palm and soybean oils.

Rubber prices began easing in July andAugust 2008, an unsurprising development be-cause they track crude oil prices closely (syn-thetic rubber is made from petroleum). Signsof weakening prices have also been evident inbeverages, with cocoa averaging a little over$2 a kilogram in November, down from $3.00in June 2008. Other agricultural commodities,especially raw materials and some foods such


43

50Jan.2005

100

150

200

250

350

300

450

400

1,000

900

800

700

600

500

400

300

200

Commodity prices(current US$)

Commodity prices(current US$)

Figure 1.29 Grains prices show sharpdeclines from recent peaks


Jan.2006

Jan.2007

Jan.2008

Wheat (left axis)

Rice (right axis)

Maize (left axis)

10363_Pg15_50:10363_Pg15_50 11/29/08 7:05 AM Page 43

as bananas and sugar, have experiencedsmaller declines, because their price increaseswere not as sharp and they are less closelylinked to energy prices.

Fertilizer prices experienced the largest in-crease among all commodity groups in 2008,with the index up 116 percent between Januaryand August 2008; prices were driven up by thecombination of strong demand growth (in re-sponse to high crop prices), limited surplusproduction capacity, higher production costsrelated to high energy prices, and an exporttax imposed by China to protect domesticsupplies. Phosphate prices (DAP), for exam-ple, increased by almost 110 percent betweenJanuary and August 2008 while urea pricesdoubled in just four months (December 2007to April 2008). The decline in crude oil andgrain prices, along with weak demand, how-ever, is now being reflected in fertilizer prices.Urea, for example, declined to $250/ton (atwo-year low) while DAP averaged below$650/ton as of November 20.

Current crop prospects are favorable.Grain production is projected to increaseabout 4 percent in the current crop year, andoilseed production is anticipated to rise bytwice as much. Although this production in-crease will allow some rebuilding of stocks,continued growth of demand for biofuelsshould keep pressure on inventories. Maizeused for ethanol in the United States is ex-pected to increase to 33 percent of the crop in2008, accounting for nearly all of the increasein global maize consumption and causingglobal maize stocks to fall. In contrast, largeincreases in wheat and oilseed productionshould allow some rebuilding of stocks.Stocks will remain low by historic standards,however, and prices will remain vulnerable tosupply disruptions or demand surges.

Overall, grain prices are projected to de-cline about 28 percent in 2009 and to recoup3 percent in 2010. Fats and oil prices are an-ticipated to fall by 27 percent in 2009 and an-other 5 percent in 2010. And beverages areprojected to decline 18 percent and 4 percent,respectively. Despite these developments, food

prices are expected to remain much higherthan during the 1990s and more than 60 per-cent higher than their levels in 2003.

In the baseline, the very tight credit condi-tions observed in November are projected todissipate during the first quarter of 2009—which together with a strong crop this seasonshould ensure that prices do not rise sharply inthe medium term. However, if farmers in high-and middle-income countries are unable to getfinancing for seed and fertilizer purchases forplantings for next season, plantings may belower than expected, which could cause agri-cultural prices to rebound during the 2009/10crop year. Farmers in key agricultural producingand exporting countries, including Australia,Argentina, Brazil, the United States, and theEuropean Union, rely on short-term financingfor inputs (e.g., fertilizer) and longer-termfinancing for the purchase of machinery. Theshort-term financing is typically guaranteed byplacing land as collateral and to a lesser extentby hedging in futures markets for a minimumprice guarantee (the latter mostly in the UnitedStates). The credit crunch combined with de-clining commodity prices has made banksreluctant to lend. The situation may worsen ifland prices begin to decline—there are alreadysigns that land prices are falling in some EUcountries—or if credit conditions do not beginto thaw. At the same time, farmers appear tohave lost faith in hedging instruments.

Commodity price declines carry significantimplications for the terms of tradeThe decline of commodity prices anticipatedfor 2009 will drive sharp changes in develop-ing countries’ terms of trade. Some 30 coun-tries are expected to gain more than 1.5 per-cent of GDP from the decline in oil prices(figure 1.30). Of these, Cyprus, Guyana,Jamaica, Jordan, the Kyrgyz Republic,Moldova, Nicaragua, the Seychelles, andTajikistan stand to gain more than 2.5 percentof GDP. And the fall in food prices will help toease both external and fiscal positions (as thecost of food subsidies declines) for many ofthe world’s poorest countries, including Benin,


44

10363_Pg15_50:10363_Pg15_50 11/29/08 7:05 AM Page 44

Eritrea, Ghana, Guinea, Haiti, Madagascar,Niger, Senegal, and Togo.

At the same time, oil-exporting countrieswill experience large terms of trade losses, withAngola, Azerbaijan, the Republic of Congo,Equatorial Guinea, Gabon, the Islamic Repub-lic of Iran, Kuwait, Libya, Nigeria, and SaudiArabia incurring first-round income losses inexcess of 10 percent of GDP. Weaker metalsprices are anticipated to reduce incomes bymore than 2 percent of GDP in Chile, Mauri-tania, Mongolia, Papua New Guinea, Suri-name, and Zimbabwe. Countries that relyheavily on grains exports are likely to be hithard. Exporters like Argentina (maize, soy-beans, wheat), Bolivia (soybeans), The Gambia(groundnuts), Guinea-Bissau (groundnuts),Guyana (rice), and Paraguay (soybeans) willexperience losses ranging from 1.6 percent to9 percent of GDP, though for some the impactwill be softened by falling oil prices.

Taking into account the effects of commod-ity price declines on both import and exportprices, more than half of the countries in asample of 162 economies are expected to seean increase in the terms of trade, of which 24will experience gains in excess of 1.5 percentof GDP. About a quarter of the countries, in-cluding most oil producers, are seen to incur

first-round income losses in excess of 1.5 per-cent of GDP (figure 1.31).

Key risks and uncertainties

The freezing of credit markets, collapse ofstock markets, large shifts in exchange

rates and commodities prices, and unprece-dented policy reactions have combined to cre-ate an extremely uncertain environment formarket participants and forecasters alike. Sev-eral possible outcomes for the global economyremain plausible at this juncture—even assum-ing that a catastrophic meltdown of markets isavoided. Global GDP growth could reasonablybe expected to be as strong as 1.4 percent in2009 and as weak as 0.4 percent, comparedwith the baseline projection of 0.9 percentgrowth presented in this chapter.

The confidence interval around projectionsfor 2010 is even wider. Instead of the typicalcyclical rebound envisaged in the baseline,output could remain subdued as consolidationin the banking sector acts as a persistent dragon growth, and credit growth remains almoststagnant for several years (see Hebling 2005;IMF 2008). Alternatively, the crisis may haveless pronounced direct effects on the real econ-omy, in which case the aggressive monetaryloosening and large-scale fiscal stimulus thatthe crisis has provoked could lead to a sharperrebound in 2010. Such a scenario runs the risk


45

Source: World Bank.

Terms of trade shock (as share of GDP)

Figure 1.31 First-round income impact of lower commodity prices will be positive inmore than half of developing countries

–25

–20

–15

–10

–5

0

5

2823 22 21 0 1

26

24

22

0

2

4

6

Source: World Bank.

Grains price shock (as share of GDP)

Vulnerable countries

Argentina

KazakhstanVenezuela,R.B. de

Seychalles

Jordan

Figure 1.30 Most vulnerable countries willbenefit from the decline in grains andoil prices

Oil price shock (as share of GDP)

10363_Pg15_50:10363_Pg15_50 11/29/08 7:05 AM Page 45

of reaccelerating inflation, which would likelyneed to be followed by a tightening of bothmonetary and fiscal policies.

In such an environment, policy makers inboth developing and high-income countriesmust be prepared to weather a worst-case sce-nario of even lower growth, including the pos-sibility of a decline in world GDP for the firsttime in the postwar period, as well as a finan-cial meltdown that could lead to a sudden stopof credit flows to all but the most creditwor-thy borrowers. Whatever the eventual out-come, the environment over the next twoyears will be radically different from thatwhich was expected only a few months ago,and policies will need to adapt.

Understandably (and correctly) under cur-rent circumstances, with the world economyconfronted with systemic financial risks,short-term attention is focused on dealingwith the immediate crisis, minimizing risks,and reacting to rapidly evolving develop-ments. Major risks concern the possibility ofbalance of payments and currency crises in in-dividual countries—a real risk at this stage forat least some developing countries. A collapseof the domestic banking system in select de-veloping countries is also a tenable possibility.In the case of Russia, where the economy isflush with petrodollars, the authorities look tobe in a position to rescue domestic financialinstitutions. Other countries are less wellpositioned and may have to draw upon inter-national assistance, a development that shouldbe undertaken quickly if necessary. The longerthe global stress lasts, the more currencies maycome under pressure. The increase in corpo-rate spreads still exceeds the increase in sover-eign spreads by a large margin. In all casespreventing a financial crisis in one countryfrom infecting a broader group of countrieswould be difficult. Therefore, instead of ex-ploring the details of a potential crisis, it isparamount to avoid a crisis altogetherthrough coordinated international action.

From a longer-term perspective, concernsare of a very different nature. The question is:How will developing countries emerge from

the current downturn, and will they retain theunderlying strength, confidence, and strongmacroeconomic fundamentals that under-pinned the record growth of the past fiveyears? The danger for all countries is that tooaggressive an effort to combat what looks tobe an inevitable slowdown may prove toocostly and undermine the strong fundamentalsthat had earlier underpinned growth. Coun-tries need to react quickly and forcefully tosigns of weakness in their financial sectors, in-cluding by liquidity injections and recapitaliz-ing banks where necessary.

Care must also be taken, however, to avoidthe possible entrenchment of inflationarypressures by ensuring that more general ef-forts to provide support to banking systemsare highly targeted and efficient, and that anynecessary liquidity injections are reabsorbedonce growth revives. The long-term costs ofsuch policies could be substantial even if theyhelp to lighten the coming recession. Policymakers must ensure that the steps taken areclear and coherent. So far, the worst has beenavoided by huge government interventions. Ifthe market comes to view such interventionsas ineffective, because they are poorly under-stood or seen as not addressing the most criti-cal problems, then the policies likely will beineffective. In this case, global economic diffi-culties could become very serious indeed.

Long-term prospects andpoverty forecast

Despite the current financial turmoil andsharp slowdown in growth anticipated for

2009, longer-term prospects for developingcountries have changed only modestly com-pared with last year’s forecast. In part prospectsare little changed because a slowdown had al-ready been anticipated, albeit to a much lesserdegree. The primary reason, however, lies in thelong-term supply potential of developing coun-tries, which should allow output to recoup thelost production induced by the coming growthrecession during the first five years of the nextdecade.


46

10363_Pg15_50:10363_Pg15_50 11/29/08 7:05 AM Page 46

Per capita GDP in developing countriesover the period 2010–15 is expected to expandat a relatively rapid annual pace of 4.6 per-cent, much faster than the 2.1 percent pace ofthe 1990s and the 0.6 percent average of the1980s, replicating the average performance ofthis decade. Improvements in macroeconomicpolicies (lower inflation, relaxation of traderestrictions, more flexible exchange rateregimes, and lower fiscal deficits) have com-bined with structural reforms (privatizationand regulatory initiatives) to reduce uncertaintyand generally improve incentives for invest-ment. Projected future growth rates are higherthan in the 1990s (and much more so than inthe 1980s) in every developing region exceptEast Asia and the Pacific, where growth is ex-pected to decline somewhat because of an agingpopulation.

Rapid growth should enable developingcountries, as a group, to achieve the MillenniumDevelopment Goal of halving poverty by 2015.The poverty forecast for 2015 is 15.5 percent,well below the target of 20.9 percent—half ofthe revised 1990 level as explained in moredetail below. The East Asia and Pacific regionhas clearly surpassed its individual target, andSouth Asia is on target. The main concern re-mains Sub-Saharan Africa. Although the inci-dence of poverty in the region has been de-clining over the past decade, at about 37.1percent in 2015, the share of people living inextreme poverty will remain well above theregion’s target of 29 percent (table 1.5).

This year’s poverty forecast is consistentwith the World Bank’s revised poverty esti-mates for developing countries. The newpoverty estimates largely result from a revisionof purchasing power parities (PPP) by using anew International Comparison Project surveyof prices paid by households. The 2005 surveyimproved on the 1993 data and methods usedto prepare previous estimates. The new pricedata reveal that the cost of living is higher inlow- and middle-income countries than hadbeen suggested by past surveys. Other factorsinfluencing the changes to the poverty esti-mates include revisions to national accounts

and the incorporation of new and more recenthousehold surveys (see box 1.4 and Chen andRavallion 2008 for more detail).

The new poverty estimates provide a signif-icantly different picture of global poverty—back to 1990 and for the most recent year, 2005(figure 1.32). Global poverty in 1990, the bench-mark year for the Millennium DevelopmentGoals, is now estimated to have been 41.7 per-cent of the developing-country population


47

Table 1.5 Poverty in developing countriesby region, selected years

Region or country 1990 2005 2015

Number of people living on less than $1.25/day (millions)East Asia and the Pacific 873.3 316.2 137.6

China 683.2 207.7 84.3Europe and Central Asia 9.1 17.3 9.8Latin America and the Caribbean 49.6 45.1 30.6Middle East and North Africa 9.7 11.0 8.8South Asia 579.2 595.6 403.9

India 435.5 455.8 313.2Sub-Saharan Africa 297.5 388.4 356.4Total 1,818.5 1,373.5 947.2

Number of people living on less than $2.00/day (millions)East Asia and the Pacific 1,273.7 728.7 438.0

China 960.8 473.7 260.9Europe and Central Asia 31.9 41.9 26.7Latin America and the Caribbean 86.3 91.3 72.4Middle East and North Africa 44.4 51.5 33.3South Asia 926.0 1,091.5 959.5

India 701.6 827.7 714.5Sub-Saharan Africa 393.6 556.7 585.0Total 2,755.9 2,561.5 2,115.0

Percentage of the population living on less than $1.25/dayEast Asia and the Pacific 54.7 16.8 6.8


India 51.3 41.6 25.4Sub-Saharan Africa 57.6 50.9 37.1Total 41.7 25.2 15.5

Percentage of the population living on less than $2.00/dayEast Asia and the Pacific 79.8 38.7 21.6


India 82.6 75.6 57.9Sub-Saharan Africa 76.2 73.0 60.8Total 63.2 47.0 34.6

Source: World Bank.

(compared with the previous estimate of28.7 percent using the old prices and guide-lines). This implies that the target for thepoverty MDG is 20.9 percent, rather than theprevious 14.4 percent. The revisions had a sig-nificant affect on all regions, except LatinAmerica and the Caribbean, which saw onlyminor adjustments. The case of China is

illustrative. The headcount index for 1990jumped from 33 percent to 60.2 percent. Thisdramatic change was attributable mainly tothe poor price comparison basis for the ear-lier estimate rather than to any underlyingchange in China itself.

The combination of a new estimate of meanconsumption and a new poverty line also


48

Surveys of prices are obviously critical in determin-ing the cost of the common basket of goods and

services in each country that is used to definepoverty. The price surveys determine the purchasingpower parity (PPP) exchange rate used in translatingdomestic prices into international dollars. Comparedwith the measure provided by market exchangerates, these PPP exchange rates provide a more accu-rate measure of the affordability of nontraded goodsin the poverty basket (because prices of nontradedgoods vary enormously across countries at differentlevels of development). Previously, the PPP exchangerates were based on a 1993 survey of prices that cov-ered relatively few countries and used a weak surveymethodology. In 2005, the World Bank, in partner-ship with other international organizations and na-tional statistical offices, concluded a new price sur-vey that covers 146 countries—of which 101 aredeveloping—and between 600 and 800 products(World Bank 2008). The survey includes China forthe first time and updates the earlier survey of India,which dated from 1985. The new price survey hashad two impacts on measured poverty.

First, it has revealed that prices paid by the poorin developing countries are higher than thought previ-ously, thereby reducing estimates of mean per capitaconsumption (or income) based on a common unit,that is, international dollars. In China, for example,average per capita consumption in 2005 was esti-mated to be about $2,400 at the old PPP exchangerate but only about $1,400 at the new PPP exchangerate. These newer price levels imply quite naturallythat households can afford fewer goods and thatmany more are living on less than a $1 a day.

Second, in light of the new price survey, the defin-ition of the “international” poverty line has been

Box 1.4 The impact of the new price survey onpoverty estimates

reevaluated. The international poverty line is meantto capture a notion of extreme poverty. As such, it iscalculated as the average poverty line of the poorestcountries. Using the new PPP estimates, the newpoverty line for extreme poverty is now measured at$1.25 (in 2005 international dollars, and representsthe average of the poverty lines of the fifteen poorestcountries for which there is data). This new povertyline is about 14 percent lower than the old interna-tional poverty line, which in 2005 dollars is $1.45.This reflects the higher PPPs for the poorest coun-tries implied by the 2005 survey data. To capture abroader notion of poverty, the World Bank’s povertyforecast has also presented statistics relevant to theso-called $2/day poverty line that was double the$1/day poverty line and reflected the average of thenational poverty lines of the middle-income coun-tries. The new $2/day poverty line, measured in 2005prices, represents the median of all of the nationalpoverty lines in the available surveys.

The increased estimates of prices in many coun-tries and a lowering of the international poverty linehave changed the picture of poverty globally—overtime and in 2005. As reported in Chen and Raval-lion (2008), these two effects partially offset eachother. The revisions in the PPPs alone, with nochange in the poverty line, would have raised thepoverty estimate in 2005 from the previous 17 per-cent to 32 percent. The reduction in the poverty lineto $1.25 a day lowers this estimate to 25 percent.The net effect is to raise the poverty estimate for2005 by 8 percentage points. The upward revision inthe poverty level does not imply that the rate ofpoverty reduction, say between 1990 and 2005, hasnot been as rapid as previously reported.

10363_Pg15_50:10363_Pg15_50 11/29/08 7:05 AM Page 48

implies a change in the starting value of thegrowth-to-poverty elasticity. Even if theshape of the income distribution is broadlythe same as in earlier income surveys (as is thecase for many countries), the fact that thepoverty line intersects the distribution at a dif-ferent spot means that the impact of a givenincrease in per capita incomes has changed.Nevertheless, the rate of improvement in theheadcount poverty rate between 1990 and2005 has not changed that much using thenew estimates.3 This year’s forecast reports anannual decline in global poverty between1990 and 2005 of some 3.3 percent, which isvery close to the earlier estimated annual de-cline of 3.2 percent. However, the higherpoverty level means that 25.2 percent of thedeveloping world’s population was living onless than $1.25 a day in 2005, compared withlast year’s estimate of 18.1 percent for 2004.As before, much of decline in global povertybetween 1990 and 2005 results from in-creased incomes in China, where the level ofextreme poverty fell from over 60 percent in1990 to less than 16 percent in 2005.

It should be noted that the impact on thepoverty forecast of the recent rise in food andenergy prices is not fully reflected in these pro-jections, which largely reflect neutral changesin per capita incomes.4 As discussed in chap-ter 3, the increase in food prices betweenJanuary 2007 and January 2008 is likely tohave increased global poverty by between130 million and 155 million people, or by1.3–1.5 percentage points. With prices nowdeclining but not expected to return to theirearlier levels, at least some of this deteriora-tion is likely to be permanent.

Notes1. Prices are as of November 20, 2008.2. Total assets of U.S. households began to decline

in the fourth quarter of 2007, as real estate valuesdropped by $185 billion and financial assets fell by$200 billion. By the second quarter of 2009, thecumulative decline in household assets amounted to$2.4 trillion, the equivalent of 16 percent of GDP.

3. It is difficult to make an exact comparison be-cause last year’s forecast was benchmarked to 2004,not 2005 as is this year’s forecast. As well, there havebeen (slight) revisions to historical national income andproduct accounts.

4. Because of the inherent delays in processinghousehold surveys, the current forecast reflects surveysthat were taken in 2005—before the rapid increase incommodity prices in 2007 and the first half of 2008.

ReferencesChen, Shaohua, and Martin Ravallion. 2008. “The De-

veloping World Is Poorer Than We Thought, ButNo Less Successful in the Fight against Poverty.”World Bank Policy Research Working Paper, No.4703, August. World Bank, Washington, DC.

Helbling, Thomas, andMarco Terrones. 2003. “Real andFinancial Effects of Bursting Asset Price Bubbles.”In IMF World Economic Outlook, April 2003.

IMF. 2008. “Financial Stress and Economic Downturns.”in IMF World Economic Outlook, October 2008.

Ratha, Dilip, Sanket Mohapatra, and Zhimei Xu. 2008.“Outlook for Remittance Flows 2008–2010.”Migration and Development Brief 8. World Bank,Washington, DC.

World Bank. 2008. Global Purchasing Power Paritiesand Real Expenditures: 2005 International Com-parison Program. Washington, DC: World Bank.


49

0

East A

sia

&Pac

ific China

South

AsiaIn

dia

Europ

e&

Centra

l Asia

Midd

leEas

t &Nor

thAfri

ca

Sub-S

ahar

anAfri

ca

Latin

Amer

ica&

Caribb

ean

Wor

ldto

tals

Source: World Bank data and staff calculations.

Note: The comparison reflects both revisions necessitated by thechange in purchasing power parities and the new internationalpoverty line, which was $1.45 per day in 2005 dollars under theold methodology and has been revised down to $1.25 per day in2005 dollars with the new methodology.

70

60New

Old

50

40

30

20

10

Headcount poverty index in 1990

Figure 1.32 Revised poverty estimatesfollowing from new price survey

10363_Pg15_50:10363_Pg15_50 11/29/08 7:05 AM Page 49

10363_Pg15_50:10363_Pg15_50 11/29/08 7:05 AM Page 50

51

The Commodity Boom:Longer-Term Prospects

2

The enduring importance of commodities tothe world economy and their volatility has

been driven home with the rise, and recentdecline, of prices for energy, metals, and food.Before they began to fall in the second halfof 2008, the real prices of energy and metalsmore than doubled over the past five years,while the real price of internationally tradedfood commodities increased 75 percent (seechapter 1 for more detail on the most recentdevelopments in commodity markets).

This chapter reviews the main characteris-tics of this most recent boom in commoditymarkets and examines the structure and behav-ior of both their demand and supply, with aview to better understanding prospects over themedium to long term. The discussion does notinclude forests or fisheries, given their com-plexity and the greater importance of issuesrelated to the public commons than for oil,metals, minerals, and agricultural products.

Several important insights emerge from thischapter that are likely to drive developmentsover the next several decades.

The magnitude and duration of the com-modity price boom are unprecedented.

• The upswing of the current boom lastedfive years. Average commodity pricesdoubled in U.S. dollar terms (in partboosted by dollar depreciation), makingthis boom longer and stronger than anyboom in the 20th century.

• Like earlier booms, this one ended whena slowdown in global growth eased de-mand pressures. The unusual strengthand duration of this boom reflect theunusual resilience, until now, of globalgrowth, particularly in developingcountries.

For oils and metals, an extended period oflow or falling prices created the conditions forthe boom and help explain the weak supplyresponse.

• Low prices throughout much of the1980s and 1990s reflected periods ofrelatively weak growth and abundantspare capacity. Idle capacity arose, bothbecause energy demand declined in thewake of high oil prices in the 1970s andearly 1980s and because demand for oiland metals in the former Soviet Union(FSU) fell sharply when altered eco-nomic incentives caused these countriesto radically increase the efficiency withwhich commodities are used.

• During the 1990s, much of the risingdemand for oils and metals was met bythe relatively easy rehabilitation of thisalready-existing capacity. This helped tokeep global commodity prices low anddeterred investment in new supplycapacity, thus depressing activity in thesectors supplying inputs to commodityexploration and exploitation.

10363_Pg51_94:10363_Pg51_94 11/29/08 7:07 AM Page 51

• As a result, a mismatch developed be-tween the trend growth of demand andthe trend growth of supply capacity.This mismatch became apparent in theearly to mid-2000s, when spare capacitywas exhausted and demand began tooutstrip supply, pushing up oil andmetals prices.

• Metals prices also were boosted by strongdemand growth, linked to unusually highand rising metal intensities in China.Going forward, the intensity of metals de-mand in China should decline as invest-ment rates fall and market mechanismsprovoke an increase in efficiency similarto that observed in the FSU.

The supply response in oil and metals is ex-pected to remain sluggish over the next fewyears, but new discoveries and technologicalprogress are likely to boost supply over thelong run.

• Ongoing shortages in the sectors thatprovide exploration and exploitationservices, and the long lags between ini-tial investments and the coming on-stream of new production, suggest thatsupply conditions may remain relativelytight in the oil and metals sectors andthat prices, although declining, are un-likely fall to their 1990s levels.

• Despite rising production levels, knownreserves of most metals and oil have re-mained fairly constant because of newdiscoveries and improvements in extrac-tion technology.

• Although oil prices are likely to fallbelow existing levels during the currentdownturn, they are expected to riseduring the recovery and stabilize ataround $75 a barrel in real terms becausenew supplies—for example, from off-shore oil fields and Canadian tar sands—have higher production costs, and a ma-jority of known reserves are located inregions that are politically unstable ornot open to outside investors.

• Given continued technological progressand appropriate policies, high oil priceswill prompt and use development of al-ternative energy sources (including re-newables) and greater efforts at conser-vation, raising energy supplies andsignificantly reducing the demand for oil.

• For metals, slower growth in commodity-intensive developing countries (as popula-tion growth slows and income levels catchup with the West), the easing of China’sinvestment surge, a rise in the share oftotal output held by the less-commodity-intensive service sector, and substitutionaway from expensive materials shouldslow demand over the long term, facilitat-ing a decline in prices.

The extension of the boom to agriculturalmarkets mainly reflects the rising demand forbiofuels and high energy prices.

• Higher energy and fertilizer prices raisedproduction costs in agriculture, and thecombination of high oil prices andbiofuel subsidies and mandates boosteddemand for some food crops. Poor har-vests in Australia also contributed to adecline in grain stocks.

• Demand growth for food and feed in de-veloping countries (such as China andIndia) has not accelerated and was not amajor contributor to the rise in foodprices.

• Real-side speculation (the decision to holdon to physical stocks in anticipation offurther price increases) and financial in-vestments along with policy reactionssuch as the imposition of export bans, alsocontributed to the rapid increase in grainand oilseed prices during 2007 and 2008.

The prospects for growth in the supply ofagricultural commodities at the global level aregood, while demand growth is likely to slow.

• Historically, agricultural productivity hasincreased more quickly than population,


52

10363_Pg51_94:10363_Pg51_94 11/29/08 7:07 AM Page 52

allowing food production to keep pacewith growing demand, even as the shareof the population working in agriculturedeclined. Over the next 20 years, assum-ing sufficient investment is forthcomingin developing and high-income countries,the spread of more-intensive productiontechniques coupled with improved vari-eties that are emerging from recent ad-vances in biotechnology, should allowglobal productivity gains on par with his-torical trends despite some productivitylosses caused by climate change.

• Considerable potential remains forbringing new (albeit somewhat less pro-ductive) land under cultivation in LatinAmerica, Africa, and the FSU countries.

• The demand for agricultural commodi-ties will slow as population growthslows and as incomes in developingcountries continue to rise (at higher in-comes, the incremental rise in demandfor agricultural commodities sparked byfurther increases in income is relativelysmall).

• Robust supply growth and slowing de-mand are expected to reduce agricul-tural prices in the long run. Increaseddemand for biofuels, however, will ex-tend the period of high prices unlesspolicies change or energy prices fallmore rapidly than expected.

• A more-rapid-than-expected warming ofthe planet could reduce agricultural pro-ductivity sharply, leading to rising foodprices.

While global supply prospects are good, un-less policy responds forcefully, food produc-tion in many developing countries may fallshort of output gainsYield gains associated with the green revolu-tion are waning in many countries. Productiv-ity levels in much of Africa and Europe andCentral Asia are also declining; they are onlyone half those of best-practice developingcountries, even after having controlled for dif-ferences in climate and soil.

Unless large-scale agricultural investmentand knowledge creation and dissemination arestepped up, food production in many of thesecountries will not keep pace with demand. Asa result, these countries will become increas-ingly dependent on imported food.

Simulations suggest that if productivitygrowth in developing countries disappoints,global food prices will be higher, and manydeveloping countries—especially those withrapidly growing populations—will be forcedto import more-expensive food from high-income countries, where productivity growthshows fewer signs of waning.

The remainder of this chapter explores eachof these themes in more detail. The next sec-tion compares the main characteristics of thecurrent commodity price cycle with earlierones. Then an examination of the long-termdemand and supply sides of commodity mar-kets follows. The chapter then brings the fore-casts for supply and demand in commoditymarkets together to form a base-case scenariofor prices, along with some alternative scenar-ios. While a wide range of future outcomes forsupply, demand, and prices are possible, thesimulations support a highest-probability out-come where today’s high prices should inducesufficient additional supply to keep commodityprices well below their recent highs over themedium to long term—although they are notexpected to descend as low as they were in the1990s. A final section concludes.

Characteristics of the currentcommodity price boom

Booms and busts are relatively commonoccurrences in commodity markets (box 2.1

and figure 2.1). Like its predecessors, thisepisode of high prices has occurred during aperiod of strong global growth and heightenedgeopolitical uncertainty, and it generated sig-nificant inflationary pressures (see chapter 1).1

However, this commodity boom was differ-ent in important ways as well. It was amongthe most marked of the past century in its

T H E C O M M O D I T Y B O O M : L O N G E R - T E R M P R O S P E C T S

53

10363_Pg51_94:10363_Pg51_94 11/29/08 7:07 AM Page 53


54

It is in the nature of commodities for their prices toshow pronounced cyclical behavior. Indeed, some

of the most influential early insights about the roleof expectations in pricing behavior derived from ob-servations of how the interaction between prices andquantities in agricultural markets tended to generateprice cycles (Kaldor 1934).

Prices in commodity markets tend to exhibit cyclicalbehavior because supply decisions (how much to plant,how many mine shafts to dig) must be made by marketparticipants well before the final sale price of the com-modity is known. Because producers in the market areuncertain about future demand and the productiondecisions of other producers, the tendency in the aggre-gate is for the independent production decisions toovercompensate for short-term imbalances betweendemand and supply and therefore for commodity pricesto be volatile. The longer the lag between the produc-tion or investment decision of producers and the actualincrease in output, the longer the cycle in prices.

Individual commodities differ in the extent towhich they exhibit cyclical behavior and in the mech-anisms underlying the cycles. The output of industrialcommodities tends to be most volatile, mainly be-cause their demand tends to fluctuate with the busi-ness cycle and (in the case of crude oil) to be subjectto policy-related supply shocks (box figure 2.1a).

While prices of all commodities are sensitive tospare capacity, the duration of booms and busts inthe metals, minerals, and the oil sectors tends to belonger than in agricultural markets because of thelonger lags between investing in new capacity andthe eventual increase in supply.

Their revenues also tend to be more volatile thanrevenues in agricultural commodities becausechanges in production mainly reflect demand shocks.As a result, both prices and quantities move intandem, rising during periods of high demand anddeclining in periods of low demand. In contrast,demand for agricultural products tends to be morestable, and volatility tends to stem from supplyshocks. As a result, price movements tend to reducerevenue volatility among agricultural commoditiesbecause prices tend to move in the opposite directionof supply shocks—rising when supply is low andfalling when supply is ample. Thus, for example,copper, lead, and zinc have much higher price andrevenue volatility than maize, soybeans, and wheat,but the differences in output volatility are much lessmarked (box figure 2.1b).

Box 2.1 Commodity price cycles

The current boom in agricultural prices is differ-ent in this regard, because it reflects a demand shockrather than a supply shock, meaning that prices haverisen even as overall production (including that des-tined for biofuels) has increased.

Phosphate

Source: World Bank.

3 5 7 9

Percent

11 13 15

TinIron Ore

PetroleumAluminum

CoffeeCocoa

SoybeansCotton

LeadMaize

Palm oilWheat

CopperRiceTea

LogsZinc

RubberSugar

Bananas

Box figure 2.1a Volatility of productionaround trend, 1960–2007

Global revenue volatility (%)

Contribution of:

Source: World Bank.

Volatility decomposition of select commodities (global),1986–2006

Box figure 2.1b Volatility decomposition ofglobal revenue for selected commodities,1986–2006

220

0

20

40

60

Coffee

Coppe

r

Lead Zinc Oil Tin

Cocoa

Cotto

nRice Te

aCor

n

Soybe

ans

Whe

at

Global price volatility (%)

Global output volatility (%)

Residual (%)

10363_Pg51_94:10363_Pg51_94 11/29/08 7:07 AM Page 54


55

80

130

180

230

280

330

380

1900 1910

Source: Grilli and Yang (1988) for 1900 to 1947; World Bank for 1948 to 2008.

1920 1930 1940 1950 1960 1970 1980 1990 2000

Real non-energy commodity prices, index (1977–79 5100)

1917 (just prior to WW I)

1951 (postwar rebuilding)

2008 (forecast)

1974 (first oil crisis)

Figure 2.1 The recent commodity boom was the largest and longest of any boom since 1900

magnitude, duration, and the number of com-modity groups whose prices have increased.

The size of the price increases areunprecedentedThe magnitude of commodity price increasesduring the current boom is without precedent.

The real U.S. dollar price of commodities hasincreased by some 109 percent since 2003, or130 percent since the earlier cyclical low in1999. By contrast, the increase in earlier majorbooms never exceeded 60 percent (table 2.1).

The unusual amplitude of the price in-creases during this boom partly reflects the

Table 2.1 Principal characteristics of major commodity booms

Common features 1915–17 1950–57 1973–74 2003–08

Rapid global real growth — 4.8 4.0 3.5(average annual percent)

Major conflict and geopolitical World War I Korean War Yom Kippur War, Iraq conflictuncertainty Vietnam War

Inflation Widespread Limited Widespread Limited secondround effects

Period of significant World War I Postwar rebuilding Not a period of Rapid buildup ofinfrastructure investment in Europe and Japan significant infrastructure in China

investment

Centered in which major Metals, Metals, agriculture Oil, agriculture Oil, metals,commodity groups agriculture agriculture

Initial rise observed in prices of Metals, Metals Oil Oilagriculture

Preceded by extended period No World War II destroyed Low prices and a Extended periodof low prices or investment much capacity supply shock of low prices

Percent increase in prices 34 47 59 131(previous trough to peak)

Years of rising prices prior to peak 4 3 2 5

Years of declining prices prior to trough 4 11 19 —

Source: World Bank.— � Not available.

10363_Pg51_94:10363_Pg51_94 11/29/08 7:07 AM Page 55

fact that the U.S. dollar has been depreciatingduring the same period and most primarycommodity prices are quoted in dollars. Thereal commodity prices in developing countries(local currency prices deflated by local infla-tion), have increased by much less than theirdollar counterparts. The real dollar price ofinternationally traded metals and mineralsrose by 158 percent between 2000 and 2007,but by only 78 percent in developing coun-tries. Similarly, the real dollar price of interna-tionally traded food commodities increased 64percent compared with a much lower 14 per-cent in developing countries (figure 2.2).2

The boom covers a wide range ofcommodities and has lasted much longerthan previous onesThis boom also differs from earlier ones inthe breadth of commodities that have seentheir prices rise sharply. The initial accelera-tion in prices was first visible in the oil mar-ket and was quickly followed by develop-ments in the metals and minerals market. The


56

0

55

110

165

220

Percent change, 2000–07

Figure 2.2 The real local currency price ofcommodities rose much less than the realdollar price

Energy

Source: World Bank.

Note: Real US$ price has been deflated by the unit value ofmanufactures (MUV); real price in developing countries representsa trade-weighted average of local currency price increases deflatedby local consumer prices increases.

Metals Food

Nominal price increase (US$)

Real US$ price increase (MUV deflated)Real price increase in developing countries

real price of agricultural products wasbroadly stable, especially in developing coun-tries, and began to rise sharply only in early2007 (figure 2.3).

This is very different from the 1950s boom,when post–World War II rebuilding (and fearsof shortages) increased metals prices and poorharvests raised agricultural prices, but theprice of oil remained flat. In the 1970s boom,agricultural and oil prices increased, but met-als prices rose initially and then collapsed withthe decline in aggregate demand.

The current price boom is unusually long.The U.S. dollar price of internationally tradedcommodities has been rising for more thanfive years, much longer than the price boomsof the 1950s and 1970s. Only the 1917 boomsaw a sustained increase in commodity pricesover a similarly long period (four years).

Typically a commodity price boom is fol-lowed by a bust as demand reacts to highprices by contracting and supply reacts by ex-panding. For example, the 1970s and 1980sbusts were associated with a sharp slowdownin world output, which eased demand pres-sures at the same time as supply was rebound-ing. Until most recently, the current boom hasbeen marked by a weak supply response (seebelow) and sustained global growth.

50Jan.2000

Jan.2001

Jan.2002

Jan.2003

Jan.2004

Jan.2005

Jan.2006

Jan.2007

Jan.2008

100

150

200

250

300

Real local currency commodity price indexes, CPI-deflated(Jan. 2000 5 100)

Figure 2.3 Oil and metal prices led thisboom, with food prices rising only muchlater

Source: World Bank.

Energy

Food

Metals and minerals

10363_Pg51_94:10363_Pg51_94 11/29/08 7:07 AM Page 56

The roots of the boom incommodity prices

This commodity price boom has been sup-ported by strong growth in global demand,

primarily from developing countries. With thepossible exception of a few metals, however,the strong GDP growth of the past five yearsdoes not by itself account for the magnitude orduration of the current boom (box 2.2). GlobalGDP was actually growing faster in the lead-upto the 1970s boom, with Japan—taking therole China plays today—emerging as a neweconomic power with growth in excess of10 percent (figure 2.4). However, the strengthand duration of this boom owes much to theresilience of developing-country growth, whichcontinued at high levels for much longer thanduring previous episodes of high commodityprices. On the one hand, this reflected the sur-prising facility with which both industrial anddeveloping countries absorbed the initial verylarge hikes in commodity prices—itself a reflec-tion of the very buoyant external conditions,including notably historically low interest rates,weak inflation, and ample liquidity (see WorldBank 2007a, 2008).

Other important factors were also at work.The supply response in extractive industrieshas been muted because of the low prices of

the late 1980s and 1990s, which reduced in-centives to develop new deposits and to investin the physical and human capital required toexpand supply. In agriculture, higher oil andfertilizer prices, along with increased demandfrom biofuels and a reduction in grain stocks,have been more important than fast growthper se.

An extended period of low pricesdepressed investment in new capacityThe influence of low prices was perhaps mostmarked in the oil sector, where following theoil shocks of the 1970s and 1980s, conserva-tion efforts and substitution toward othersources of energy depressed demand for oil andoil prices. Indeed, it took 15 years for world oildemand to regain its 1979 level. Meanwhile, theexpansion of oil production, particularly in theNorth Sea, Mexico, and Alaska eliminatedthe market power that the Organization ofPetroleum-Exporting Countries (OPEC) hadexploited to keep prices high even in the face ofrapidly declining demand. By mid-1986, nom-inal prices had fallen to less than $10 a barreland OPEC’s spare production capacity wasequal to 8.7 mb/d—more than 13 percent ofworld demand at that time.3

Global spare capacity was further aug-mented during the 1990s, when demand for oilin the FSU declined precipitously and more orless permanently. As the prices of primary com-modities were allowed to reflect world pricesand many of the energy- (and metals-) inten-sive industries that had characterized the Sovietera closed or retooled, demand for energy (andmetals) in these countries declined rapidly.Overall, oil demand declined by 40 percent be-tween 1987 (its peak) and 1999—or by 5 mil-lion barrels a day—the equivalent of 7 percentof world demand in 2000.

Initially, oil production in the FSU fell byabout as much, so there was an enormousbuildup of dormant capacity. IncludingOPEC’s surplus capacity of about 5 mb/d,total dormant capacity from these countriesequaled around 10 mb/d in 1995 (figure 2.5).4


57

Annual percent change in global GDP

1970sboom

Source: World Bank.

Figure 2.4 Global growth lasted longer andwas stronger during the recent commodityboom than in earlier ones

2000sboom

01967 1972 1977 1982 1987 1992 1997 2002 2007

1

2

3

4

5

6

7

8

10363_Pg51_94:10363_Pg51_94 11/29/08 7:07 AM Page 57

opment of existing wells declined by morethan 50 percent, from $72 billion in 1980 to$30 billion in 1999 (figure 2.6).5 As a result,demand for the inputs required for oil explo-ration and extraction was weak, and capacityin these supporting industries declined, as didthe number of new engineers trained to findand extract oil.6


58

The growth surge of developing countries between2003 and 2007 contributed to strong demand

for commodities. Had output expanded more slowly,in line with the long-term growth potential of devel-oping countries—estimated to be about 6.4 per-cent—oil demand would have been lower by onlyabout 1 million barrels a day, or just under 1.2 per-cent of world consumption; demand for metalswould have been about 1.5 percent lower anddemand for grains about 1.9 percent lower.

Overall, demand for most commodities at theglobal level rose less quickly than world GDP, andfor most commodities, the contribution ofdeveloping countries to the increase in commoditydemand was in line with their GDP growth (boxfigure). Incremental developing-country demandfor some commodities was much stronger than inhigh-income countries, both because developing-country GDP was growing at a faster rate andbecause relatively commodity-intensive manufactur-ing activities were being transferred from high-income to developing countries in this time period—a factor that by itself should have had no impact onglobal commodity demand.

Indeed, despite the acceleration in world GDP,consumption for most commodities did not riserapidly. Coal and certain metals represent notableexceptions, and here the demand of China has playeda particular role. Between 2003 and 2007, China’sconsumption of aluminum increased by 7.1 milliontons, or 26 percent of world demand. Coal consump-tion increased by 458 million (oil equivalent) tons, or18 percent of global demand. However, China’s pro-duction of these commodities increased by almost asmuch—7.0 million tons in the case of aluminum and

Box 2.2 Developing-country growth and globalcommodity demand in the recent past

421 tons in the case of coal—so its demand surge con-tributed relatively little to overall market tightness. In-deed, by mid-2008, the price of aluminum rose byonly 74 percent compared with its average value inthe 1990s (versus 200 percent for metals and mineralsin general), and coal rose by 392 percent, about thesame as natural gas, but much less than oil.

Importantly, despite rapid gains in developing-country GDP and income growth, grain demand didnot accelerate appreciably for developing countriesconsidered as a whole or for China alone. In fact,Chinese consumption of wheat and rice declined,and China’s contribution to incremental global corndemand was roughly in line with its increase in GDP.

Box figure 2.2 China was the key globalmetals contributor to global demand growth

Percent increase in global commodity demand, 2003–07

2505

Source: World Bank.

10152025303540

Aluminu

mCoa

l

Coppe

r

Real G

DPCor

n

Crude

oil

Whe

atRice

Contributions to total from:

Ethanolproduction

OECD Rest ofthe world

China

The buildup of excess capacity meant thatthe real price of oil during the 1990s remainedlow, at $16 a barrel, equivalent to half theprice experienced during 1985. It also meantthat there was little incentive to invest in new,higher-cost oil fields. Overall spending bymajor American multinational oil companieson exploration for new wells and the devel-

10363_Pg51_94:10363_Pg51_94 11/29/08 7:07 AM Page 58


59

As the transition continued, oil-producingfirms in the region were able to rehabilitateexisting capacity relatively easily and to reorientexpanding output to Western markets, wheredemand continued to rise. Between 1995 and2005, world oil demand increased by nearly14 mb/d, with 8 mb/d of that total being metby the dormant capacity in the countries of theFSU and OPEC.7 As a result, underlying ca-pacity grew less than half as fast as demandthroughout the period.

With spare and dormant capacityabsorbed, prices surged in 2004By 2004, the dormant capacity that had beencreated by the decline in demand in the FSUhad been reabsorbed. When demand growth(which had been subdued following the burst-ing of the Internet bubble) regained strength,supply was unable to keep pace, in turn re-sulting in a surge in prices.8

Metal demand also declined sharply as nu-merous heavy industries in the FSU went outof business. Global demand for metals andminerals eased sharply beginning in 1990and only returned to trend rates in 1997 (fig-ure 2.7). As was the case in the oil sector, thepickup in metals prices beginning in 2003 didnot reflect unusually strong demand—exceptfor aluminum—whose price, as it happens,has been relatively stable. Rather, it reflectedlow stock levels and depressed capacity.

Indeed, the strong correlation between theprices of metals and minerals on the one handand oil on the other during 2003–06 isunusual. Historically, the correlation betweenthe prices of these commodities tends to bemuch less pronounced than between oil andagricultural goods (table 2.2) because high oilprices tend to cut into industrial productionand demand for metals, while food demand isrelatively inelastic.0

1981

1983

1985

1987

1989

1993

1995

1997

1999

1991

2003

2005

2001

20

40

60

80

100

0

10

20

30

40

50

60

70

US$ 2006, billions

Crude oil prices (right axis)

Source: Energy Information Agency; World Bank.

Figure 2.6 Real spending by major Americanmultinational oil companies declined by 60percent in the 1980s

Real price per bbl, US$ 2000

Exploration (left axis) Development (left axis)

01980

Source: World Bank, British Petroleum, International EnergyAgency, Petroleum Economics Ltd.

1985 1990 1995 2000 2005

5

10

15

20

Bbl per day, millions

OPECsparecapacity

Former Soviet Uniondormant capacity

Total spare capacity

Figure 2.5 Dormant capacity helped keep oilprices low in the 1990s

24

22

0

2

4

6

8

10

1985

Source: World Bank, International Monetary Fund.

1990 1995 2000 2005

3-year, moving average of the % change in metal demand

Figure 2.7 Global metal demand also fellduring the transition

Aluminium

Copper

Nickel

Zinc

10363_Pg51_94:10363_Pg51_94 11/29/08 7:07 AM Page 59

Increasing prices sparked a boom ininvestment in the oil, metals, and mineralsmarketsGlobal private investment in exploration fornonferrous metals rose from $2 billion in2002 to $7 billion in 2006 and to an estimated$9 billion in 2007. Overall investment in thesector more than doubled between 2001 and2005 in a number of mineral-rich countries in-cluding Canada, Mexico, the Russian Federa-tion, South Africa, and the United States(UNCTAD 2007). At the same time, invest-ment in the oil sector increased dramatically,75 percent in the case of the American multi-national companies (see figure 2.6).

After years of low investment, the abilityof service sectors to deliver inputs to thecommodity-producing firms had atrophied.As a result, the surge in demand for invest-ment goods over the last several years has ex-ceeded capacity by a wide margin and costshave skyrocketed.

In the oil sector, operating costs havemore than doubled, and the cost of inputs toexploration and extraction have increasedsubstantially. For example, the day-rate priceof semisubmersible rigs in the Gulf ofMexico (0–3,000 ft. water depth) increasedfrom $36,000 in 2000 to $325,000 in March2008, a ninefold increase. Similar increaseshave been observed in other items, such as

water jack-up rigs, whose day rates have in-creased fivefold in West Africa.

Such factors have put upward pressure onthe costs of developing new mines and oilfields. Operating costs for marginal producersrose by 25 percent for copper and 28 percentfor aluminum between 2002 and 2005 (IMF2006), and in the case of at least one nickelproject, they rose by 170 percent.9 Highercosts are reported to have increased the cost ofextracting a barrel of crude from Canada’s oilsands to $75, while deepwater offshore pro-jects may cost more than $50 a barrel.

Although higher prices are inducing sub-stantial increases in capacity in the input in-dustry, that capacity will not be in place forseveral years. As a result, some delivery timeshave more than doubled. For example, in themining sector it currently takes 45 months todeliver a grinding mill, compared with a morenormal 20 months; for rope shovels thedelivery time has gone from 9 months to 24.Large haul trucks, normally available within4 months, now take 2 years.10 Other servicesmay take even longer to come into balance;the training of technical personnel such asengineers typically takes many years.

As a consequence, it may take some timebefore the surge in investment now under wayleads to a surge in the delivery of inputs, andeven more time before delivery of inputs


60

Table 2.2 Comovement among major commodity prices, 1960–2007

Commodity Maize Wheat Rice Coffee Cotton Copper Aluminum Iron ore Gold

Wheat 0.91Rice 0.82 0.81Coffee 0.70 0.45 0.63Cotton 0.83 0.80 0.82 0.82Copper 0.75 0.55 0.71 0.35 0.75Aluminum 0.70 0.46 0.63 0.37 0.76 0.41Iron ore 0.72 0.49 0.63 0.36 0.76 0.0 0.34Gold 0.69 0.0 0.65 0.54 0.80 0.0 0.44 0.0Crude oil 0.72 0.55 0.65 0.58 0.81 0.0 0.48 0.0 0.83

Source: World Bank.Note: The numbers are the adjusted R2s of a regression of each price on all other prices (individually), a time trend, and theMUV, both directions. The residual was tested for stationarity (5% level of significance). If cointegration was confirmed in onedirection, the table reports the respective adjusted R2. If cointegration was found in both directions, the higher adjusted R2 isreported. If no cointegration was found, implying that any correlation would, in fact, be a spurious correlation, the result was notreported, and the respective cell shows 0.0 (e.g., gold with wheat or copper).

10363_Pg51_94:10363_Pg51_94 11/29/08 7:07 AM Page 60

translates into actual increases in oil, metal,and mineral production. All of this suggeststhat notwithstanding recent declines, supplywill continue to be relatively scarce for severalmore years and that prices will remain higherthan in the 1990s for some time.

The boom in agricultural prices reflectsboth high costs stemming from oil pricesand increased demand from biofuelsThe rise in the price of agricultural commodi-ties occurred much later than it did for eitheroil or metals and minerals. Dollar prices wererising as early as 2003, but these increasesmainly reflected exchange rate movements.Relative to consumer prices in developingcountries, internationally traded food priceswere broadly stable until 2007, when theprices of internationally traded food commodi-ties (such as maize, wheat, and soybeans) rosevery rapidly (figure 2.8).

The timing of the rise in agricultural pricespoints strongly to the impact of energy markets(box 2.3) First, agriculture production is fairlyenergy intensive. The increase in oil pricesraised the price of fuels to power machineryand irrigation systems; it also raised the price

of fertilizer and other chemicals that are energyintensive to produce. The impact across differ-ent countries is difficult to quantify owing to alack of data. In the United States, fuel, fertil-izer, and chemicals accounted for 34 percent ofmaize production costs and 27 percent ofwheat production costs in 2007 (USDA 2008).Energy, fertilizer, and chemicals would typi-cally make up a smaller share of productioncosts in developing countries, because produc-tion is less intensive. Nevertheless such costscan be significant where intensive techniquesare used. Thus, fertilizer is estimated to haveaccounted for 18 percent of variable costs forirrigated wheat in the Indian Punjab in 2002and for 34 percent of soybean costs in theMato Grosso, Brazil (World Bank 2007b).

Second, high oil prices sparked an increasein biofuel production in the United States andEurope that boosted demand for certaingrains and oilseeds thus contributing to theirrapid price rise in the course of 2007 and early2008 (Mitchell 2008). Overall, two-thirds ofthe increase in world maize production since2004 has gone to meet increased biofuel de-mand in the United States, thereby reducingthe quantity available for food and feed uses.Estimates of the impact of increased demandfor biofuels on the rise in nominal maizeprices range from 70 percent (Lipsky 2008),to 60 percent (Collins 2008), to 47 percent(Rosegrant and others 2008).

The increased demand for crops used forbiofuels contributed to price increases forother food by reducing the land allocated toother crops. For example, in the United Stateshigh prices increased land devoted to maizeproduction by 22 percent in 2007, with mostof the increase at the expense of soybeans, theproduction of which declined by 16 percent.Area planted to rapeseed and sunflowers—used for biodiesel production—increased inEurope and elsewhere at the expense of wheat.Moreover, rising prices for maize, wheat, andsoybeans redirected consumer demand towardother food products, aggravating price pres-sures on other grains. For example, rice pricesrose from $376 a ton in January 2008 to $907


61

50Jan.2000

Jan.2001

Jan.2002

Jan.2003

Jan.2004

Jan.2005

Jan.2006

Jan.2007

Jan.2008

100

150

200

250

300

Domestic real

Nominal US$

Food price indexes

Figure 2.8 Real food prices were broadlystable in developing countries until mid-2007

US$ nominal and domestic CPI-deflated price indexes(Jan. 2000 5 100)

Source: World Bank.

Note: Individual country data deflated by local consumer priceindex and aggregated by country shares in global imports.

10363_Pg51_94:10363_Pg51_94 11/29/08 7:07 AM Page 61

a ton in April, partly in response to the grow-ing concern about the adequacy of global foodsupplies and the 120 percent increase in wheatprices during the previous six months.

While biofuels have contributed to higherfood crop prices, they also represent an op-portunity for profitable production in devel-oping countries (OECD 2007; GTZ 2006).Additional ethanol production need not implyreducing food crops production. Brazil, forexample, is a low-cost producer of ethanolfrom sugarcane and has an estimated 180 mil-lion hectares of pasture that could be used toproduce additional sugarcane for ethanol—without reducing the food sugar crop. ManySub-Saharan African countries, includingAngola, Mozambique, and Tanzania, also

have the potential to produce ethanol prof-itably from sugarcane on land that is not usedfor food crop production. Finally, nonfoodcrops such as jatropha can be used to producebiodiesel in many developing countries.11

In addition to the impact of oil markets,food prices were boosted by a series of poorwheat harvests, notably in Australia.12 Beforethe run-up of prices in 2007, wheat stocks hadfallen to the second lowest level of the past40 years (figure 2.9).

Reported global stocks of corn and rice de-clined before 2007, mainly due to a reduction ofvery large government stocks in China. Becausethese stocks have been greatly underestimatedfor the past 30 years (figure 2.10), current stocklevels are not that different than what the world


62

Crude oil prices affect the prices of other commodi-ties in a number of ways. On the supply side,

crude oil enters the aggregate production function ofmost primary commodities through the use of variousenergy-intensive inputs and, often, transportation overlong distances, an energy-demanding process. Somecommodities, such as aluminum, have to go throughan energy-intensive primary processing stage.

On the demand side, some commodities competedirectly with synthetic products, which are producedfrom crude oil (cotton with man-made fibers, naturalrubber with synthetic rubber). The demand for othercommodities (maize, sugar, rapeseed, and other oils)has increased to produce biofuels. And the price ofenergy commodities such as gas and coal are affectedbecause of their substitutability with crude oil.

Increases in crude oil prices also increase thedisposable income of oil-exporting countries. Becausethese countries are heavy consumers of some com-modities (e.g. tea and gold), and demand for theseproducts is sensitive to incomes, high oil pricessharply increased regional demand for these products.Finally, crude oil price spikes are often associated with

Box 2.3 The historical link between crude oil andother commodity prices

inflationary pressures. As a result, the demand (andhence the price) of precious metals often rise with oilprices, because investors and households view thesemetals as more secure ways for storing wealth.

Crude oil price increases reduce the disposableincomes of consumers, which, in turn, may slowindustrial production. In principle, lower disposableincome should have a negative impact on the con-sumption of food commodities. However, becausethe income elasticity for most food commodities issmall, this effect is limited, and the positive impactof crude oil price increases on the prices of foodcommodities—through increased production andtransportation costs—tends to overshadow thenegative impact of reduced global consumption.

In contrast, the negative effect of high energyprices on industrial production reduces the demandfor metals, thereby putting downward pressure ontheir prices. This tends to offset the positive effectfrom higher production and transportation costs.As a result the correlation between metals and oilprices is much lower than between oil and foodprices (see table 2.2).

10363_Pg51_94:10363_Pg51_94 11/29/08 7:07 AM Page 62

forces that otherwise would have helped toattenuate the rise in prices and shorten the du-ration of the boom. As discussed in chapter 3,although the various subsidies and price con-trols that were in place or were introducedmuted the poverty impact of higher prices,they have also reduced producers’ incentivesto increase output and consumers’ incentivesto substitute less-costly items in their foodbaskets. And export bans limited suppliesavailable on international markets. For exam-ple, India’s ban on rice exports in April 2008was followed by other rice exporters, whichprompted some countries, notably the Philip-pines, to increase rice imports to build upstrategic reserves, thus further boosting inter-national prices.

The activities of financial investors mayhave contributed to price rises as well. Tradi-tionally, hedgers and speculators have beenthe dominant players in futures exchanges,but over the past few years, investment fundshave become important players as well. Suchfunds may have indirectly influenced commod-ity prices. Since 2003 index fund investors,who allocate funds across a basket of com-modity futures, have invested almost $250 bil-lion in U.S. commodity markets, about half ofit in energy commodities (Masters 2008).While such purchases create no real demandfor commodities, they may have influencedprices because these funds are large comparedwith their physical market counterparts andbecause they have expanded rapidly. Their in-fluence on prices is especially likely, if the rapidexpansion of these markets contributed to ex-pectations of rising prices, thereby exacerbat-ing swings, as argued by Soros (2008).

The empirical evidence on whether suchfunds have contributed to the recent pricesurge is mixed. In the nonferrous metals mar-ket (where a similar buildup of financial posi-tions has occurred), Gilbert (2008) found nodirect evidence of the impact of investor activ-ity on the prices of metals but some evidenceof extrapolative price behavior that resulted inprice movements not fully justified by marketfundamentals. He also found strong evidence


63

0

5

1960 1965

10

15

20

25

30

35

40

Stock levels as a % of annual production

Figure 2.9 Most of the decline in global grainstocks reflects lower stocks in China

Source: World Bank.

Global stocks(with newChina data)

Global stocks(excluding China)

Global stocks(using older datafor China)

1970 1975 1980 1985 1990 1995 2000 2005

01960

Source: World Bank.

1966 1972 1978 1984 1990 1996 2002 2008

5

10

15

20

25

30

35

40

45

Period of strongincrease inbiofuel production

Figure 2.10 Outside of China, only wheatstocks are unusually low

CornRice

Wheat

Stock levels as a % of annual production

thought them to be during the early 1990s. It isthus unclear whether market participants tookthe decline in global stocks as a signal of com-ing scarcity or simply a return to stock levelsthat were consistent with relatively low prices adecade ago.

Government policy and investment fundactivity may have exacerbated the increasein commodity pricesThe extent of food price rises during thisboom was probably exacerbated by the ac-tions of governments, which impeded market

10363_Pg51_94:10363_Pg51_94 11/29/08 7:07 AM Page 63

that futures positions of index providers overthe past two years have affected soybean (butnot maize) prices. Similarly, Plastina (2008)concluded that between January 2006 andFebruary 2008, investment fund activitymight have pushed cotton prices 14 percenthigher than they would have been otherwise.On the other hand, two IMF (2006, 2008)studies failed to find evidence that specula-tors have had a systematic influence on com-modity prices. A similar conclusion wasreached by a series of studies undertaken bythe Commodities Futures Trading Commis-sion, the agency that regulates U.S. futures ex-changes (Büyüksahin, Haigh, and Robe 2008;ITF 2008).

Although evidence that financial invest-ments have contributed to the rapid run-up incommodity prices is limited, it seems likelythat real-side speculation (the decision to holdstocks in anticipation of further price increasesor to order more than needed now for thesame reasons) likely contributed to the rapidincrease in prices during 2007 and 2008.13

Long-term demand prospects

The longevity of the current boom and thewide range of commodities that have been

affected have prompted many observers towonder if the global economy is moving into anew era characterized by relative shortage andpermanently higher (and even permanently ris-ing) commodity prices. This section looks atdemand and supply conditions in commoditymarkets over the medium to long term andconcludes that slower population and GDPgrowth, changes in the structure of GDP, andtechnological improvements in productionand use of commodities make this scenariounlikely.

Demand for (and supply of) commoditiesover the past 35 years has been rising steadily.The quantity of energy consumed has increasedby an average of 2.2 percent a year during1970–2005, that of metals and minerals by3.1 percent, and that of food by around 2.2 per-cent. However, demand for these commodities

has grown less quickly than GDP, albeit morequickly than population (figure 2.11).

Expressed another way, the commodity in-tensity of GDP has been declining. For oil andfood, this process has been going on continu-ously since the 1970s. For metals, the sametrend was observed until the mid-1990s whenit began to reverse (figure 2.12)

More generally, growth in the demand forcommodities is influenced by a wide range offactors including several fundamental economicdrivers.

Incomes and population. As per capitaincomes rise, demand for commodities alsotends to increase, but the sensitivity ofdemand to an increase in income differs acrosscommodities and changes as income levelsrise. For example, at low-income levels,demand for grains rises relatively quickly asincome increases, but as per capita incomesreach about $3,000 dollars, the pace at whichgrains demand rises declines, ultimately fallingto close to zero. Thus, a 10 percent increase in


64

Palm oil

0 1 2 3 4

Percent

5 6 7 8 9

SoybeansAluminum

GDPRubberCocoa

TeaLogs

CopperMaize

BananasWheat

RiceSugarCotton

PopulationZinc

PetroleumCoffee

Iron orePhosphate rock

TinLead

Average annual growth rate, 1975–2006

Figure 2.11 Demand for most commodities has grown less rapidly than GDP but more rapidly than population

Source: World Bank.

10363_Pg51_94:10363_Pg51_94 11/29/08 7:07 AM Page 64

incomes is associated with a 6 percent increasein grains demand in low-income countries butalmost no increase in high-income countries(table 2.3). As a result, beyond a certainincome level, grains demand is mainly dictatedby population growth. The sensitivity ofdemand for metals to incomes is much higherbut tends not to change as income levels rise.Energy is the reverse of grains, with thedemand for energy rising more rapidly thanincomes in high-income countries.

The composition of GDP. Commodity de-mand depends on more than just GDP. Thecomposition of demand also plays an impor-tant role. Over time, the commodity intensityof GDP has declined partly because demandhas evolved toward goods and services that

are much less intensive in their use of com-modities. This trend is illustrated in table 2.4,which shows the value per kilogram of a vari-ety of different products. Newer products,such as computers and mobile telephones,have a growing share in world GDP and con-tain very little in the way of commodities(proxied here by their weight).

The same effect can be seen at the sectorallevel. Industrial activity tends to be more com-modity intensive than agricultural activity,which in turn is more commodity intensivethan services. Thus, part of the decliningcommodity intensity of demand over the past35 years reflects the rise of the service sector,which accounted for 50 percent of world GDPin 1971 and 69 percent in 2005—a trend thatis shared by both high-income and developingcountries.

Technological change. Increased efficiencyin the use of commodities in production andconsumption has also contributed signifi-cantly to the dematerialization of economicactivity. Examples include improvements ingas mileage in automobiles and the substitu-tion of artificial for natural fibers in clothing.


65

Figure 2.12 The quantity of most commoditiesused per unit of GDP was declining untilrecently

0.70

1971

1974

1977

1980

1983

1986

1989

1992

1995

1998

2001

2004

0.75

0.80

0.85

0.90

0.95

1.00

1.05

1.10

Commodity intensity of demand, index (1971 5 1.00)

Source: World Bank.

Energy Metals

Metals (excluding China) Food

Table 2.3 Impact of a 10 percent increasein incomes on commodity demand(Percent)

Income group Grains Energy Metals

Low 6.0 4.5 10.1Lower middle 3.3 7.2 10.1Upper middle 1.4 9.2 10.1High 0.0 1.1 10.1

Source: World Bank.

Table 2.4 Modern goods make lessintensive use of commodities(US$)

Good Value per kilogram

Iron ore 0.04Steam coal 0.07Wheat 0.27Crude oil 0.47Standard steel 0.56Newsprint 0.89Supertanker 4.00Motor car 33.00Dishwasher 56.00TV set 133.00Submarine 222.00Large passenger aircraft 1,334.00Laptop computer 2,224.00Mobile telephone 4,448.00Jet fighter 13,344.00Windows 2000 Software, CD Rom 44,480.00Telecom satellite 88,960.00Banking services �

Source: Radetzki 2008a.

10363_Pg51_94:10363_Pg51_94 11/29/08 7:07 AM Page 65

Long-term projections suggest that themain factors driving commodity demandwill slowTo a significant degree, future demand forcommodities will reflect the combined impactof, GDP growth, changes in the compositionof demand, and technological progress(table 2.5).

• Population growth over the next twodecades is expected to slow significantlyfrom 1.2 percent during the 2000s toabout 0.8 percent in the period2015–30, which should help moderatecommodity demand compared with pastdemand.

• Per capita income growth is also pro-jected to slow somewhat for the world asa whole, mainly because incomes in thelargest developing countries are expected

to rise less quickly than they did duringthe 1990s. Nevertheless, developing-country per capita incomes are projectedto triple, rising from $1,550 to $4,650between 2004 and 2030. This meansthat, although global demand for grainsand some metals is likely to decelerate,energy demand is likely to strengthen.

• The composition of GDP is not expectedto continue to move toward services butto stabilize more or less at current levels.This suggests that commodity intensitiesmay decline less rapidly than they havein the past.

• Prospects for technological progress arethe least certain element likely to deter-mine future commodity demand. Shouldpolicy succeed in continuing past gains,then this too should tend to moderatecommodity demand.


66

Table 2.5 Fundamental economic factors drive future commodity demand

Average annual growth rate

Period Per capita income Population GDP Share of services in GDP Share of services in GDP

(percent) (percent)

1990sWorld 1.2 1.5 2.7 0.99 64.6High income 1.8 0.7 2.5 0.89 67.8Low and middle income 2.0 1.6 3.6 1.73 49.8Low income 2.3 2.2 4.5 0.96 44.3Middle income 2.2 1.2 3.5 1.84 50.8

2000sWorld 1.8 1.2 3.1 0.47 68.5High income 1.7 0.7 2.5 0.51 71.8Low and middle income 4.2 1.3 5.6 0.24 53.8Low income 4.1 1.9 6.1 1.50 49.4Middle income 4.6 0.9 5.5 0.04 54.5

2015–30World 1.7 0.8 2.5 �0.41 50.3High income 1.2 0.1 1.3 0.02 59.0Low and middle income 3.9 0.9 4.9 �0.07 35.6Low income 3.8 1.5 5.4 �0.02 44.0Middle income 4.1 0.7 4.8 �0.08 35.0

Change (2015–30 vs. 2000s)World �0.2 �0.4 �0.6 �0.88 �18.3High-income �0.5 �0.7 �1.2 �0.49 �12.8Low and middle income �0.3 �0.4 �0.7 �0.31 �18.2Low income �0.3 �0.4 �0.7 �1.52 �5.4Middle income �0.5 �0.2 �0.7 �0.12 �19.5

Source: World Bank LINKAGES model.

10363_Pg51_94:10363_Pg51_94 11/29/08 7:07 AM Page 66

A number of technologies currently availableas prototypes or in early stages of commercial-ization could help more than double fuel effi-ciency over the next several decades. In 2005,about 8 liters of fuel were needed to drive 100kilometers; by 2050, fewer than 3 liters may beneeded (IEA 2008a). Even more optimistic sce-narios project that by 2050, 90 percent of thevehicles in the high-income world and 75 per-cent in the developing world will be powered byalternative fuels, such as plug-in hybrids (hybridcars with large batteries that can be pluggedinto the main electrical network), electric, andhydrogen-powered cars. Such a shift wouldreduce considerably private transportation’sdependence on liquid fuels. Indeed, prototypeand soon-to-be-released electric and hydrogen-powered cars already exist (box 2.4).

Strong growth in developing countries isexpected to dominate future energydemandAssuming that energy efficiency continues toimprove at about the same rate as in the past,total demand for energy is projected to rise by55 percent between now and 2030, with 80percent of that emanating from fast-growingdeveloping countries (table 2.6). Overall,weaker population growth and technologicalchange are likely to outweigh the impact ofrising developing-country incomes and theirincreased weight in overall demand. Hence therate of growth of energy demand is expectedto ease over time, declining from an average of1.8 percent during the past 15 years to about1.3 percent in the period 2015–30.

In the baseline scenario, climatic and envi-ronmental concerns are expected to contributeto a modest shift away from petroleum prod-ucts toward less carbon-intensive fuel sources,such as natural gas, and renewable fuels, suchas wind, solar, and geothermal. Oil’s share inoverall energy consumption is expected todecline, with demand rising more slowly.Demand growth is projected to fall from1.7 percent a year in 2005–15 to 1.1 per-cent in 2015–30, reaching between 112 and118 million barrels a day by 2030.


67

The remainder of this section discusses inmore detail how these factors and technologi-cal change are expected to play out in individ-ual commodity markets.

Demand prospects for energyRising incomes and technology are expectedto play crucial roles in determining future en-ergy demand. Assuming no improvement inenergy efficiency, given expected increases inincomes and population, demand for energywould rise by more than 120 percent betweennow and 2030, with growth in developingcountries responsible for three-fourths of thatincrease. Assuming the composition of energydemand and supply did not change, that wouldimply that demand for oil would more thandouble, from 82 mb/d in 2007 to 174 mb/din 2030.

Efficiency gains and conservation effortsreduced energy demand by 50 percentover the past 35 yearsOf course, these assumptions are somewhatsimplistic, viewed against the light of recenthistory. Energy efficiency over the past 50years has in fact improved sharply. Since1960, the efficiency of jet transport has morethan tripled (Lee and others 2001) while fuelefficiency in cars has also increased signifi-cantly. Overall, between 1970 and 2004, tech-nological change lowered energy demand 56percent from what it would have been other-wise (IEA 2007). Much of the improvementresulted from substitution and conservationprompted by higher prices. Ongoing techno-logical change and increased efficiency inChina (Lin and others 2006) and the FSUcountries (see earlier discussion) also playedimportant roles.14

Looking forward, similar improvements inenergy efficiency are possible if supported byan appropriate policy mix. Of particular im-portance will be efficiency in the transport sec-tor, which is expected to account for some75 percent of the increase in future oil use,largely because of rising incomes and car own-ership in developing countries (IEA 2007).

10363_Pg51_94:10363_Pg51_94 11/29/08 7:07 AM Page 67

Another important feature of the composi-tion of energy demand is the importance ofcoal, which currently accounts for more thana quarter of global energy consumption. Coalis primarily used by developing countries(62 percent), with China accounting for morethan 40 percent of global consumption. Thebaseline simulations indicate a slight increasein coal’s share, from 25.3 percent in 2005 to27.8 percent in 2015. However, the projectionis subject to two risks: on the upside, if newclean coal technologies (including carbon se-questration) come on board, coal’s share inglobal energy consumption is likely to bemuch higher. However, if such technologies

do not materialize, coal use is likely to be sub-jected to significant environmental regulationthat could significantly reduce its economicattractiveness.

The future path and mix of energydemand will depend on policySimulations suggest that a more aggressivestance toward reducing carbon emissionscould generate a further moderation in energydemand and in fossil-fuel use. For example, a$21 tax per ton of carbon dioxide could be ex-pected to reduce demand for energy by 33 per-cent (see the simulations at the end of thechapter). Because of its high carbon content,


68

Hydrogen and electricity are emerging fuels fortransportation; fully ethanol-powered and flex-

fuel cars are already well-established commercialsuccesses in Brazil and increasingly in the UnitedStates and Europe. Existing hybrid cars offer a50 percent improvement in fuel efficiency for citydriving, while plug-in hybrid cars have the potentialof reducing reliance on gasoline even more.Hydrogen-fuel-cell and all-electric cars could reducethat dependence to zero, but considerable progressneeds to be made in increasing the efficiency ofbattery technology and in the production andconversion of hydrogen into electricity before thesevehicles will be competitive.

Currently, most major car manufacturers haveprototype versions of all such cars. General Motorshas announced its intention to sell commercially assoon as 2010 an extended-range electric vehicle (the“Volt”), which is a battery-powered electric car thatuses a small flex-fuel engine to extend its range forhighway driving. The Volt is expected to be able torun up to 40 miles a day (more than the averagedaily driving distance of 75 percent of Americans) onbatteries alone and 250 miles using its flex-fuel gen-erator. The car is expected to have an EPA rating of100 miles a gallon (Connor 2008), and its operatingcosts could be 0.02 cents a mile or one-sixth the costof a vehicle powered with gasoline at $3.80 a gallon(Padget 2008).

Box 2.4 Alternative fuels for transportationMeanwhile, Honda is already leasing a limited

number of hydrogen-fuel-cell-powered cars to thegeneral public in southern California. While costs ofoperation are similar to gas-powered cars, the carsthemselves are extremely expensive and the leasesbeing offered imply a substantial subsidy. The cost offuel-cell stack systems (the mechanism that convertshydrogen into power and that uses platinum) willhave to decline tenfold before these vehicles becomeeconomically viable.

For both plug-in hybrids and electric cars, themajor stumbling block is the size, weight, and cost ofthe battery required to power them. With currenttechnology, the battery needed to power an electriccar 500 kilometers weighs five times as much as theequivalent amount of gasoline and would cost$50,000. Over the next several decades, technologi-cal progress achieved through the commercializationof hybrid cars is expected to raise battery efficiencyand reduce costs, so that plug-in hybrids will bewidely available by 2020.

Prospects for all-electric cars are less clear, mainlybecause of the time that it takes to recharge batter-ies, a factor that makes them much less attractivethan gas-powered vehicles. Here hydrogen-fuel-cell-powered cars could have an advantage if the costsassociated with the fuel stack can be resolved.

Source: IEA 2008a.

10363_Pg51_94:10363_Pg51_94 11/29/08 7:07 AM Page 68


69

demand for coal would decline most sharplyunder such a scenario, with natural gas andother low-carbon energies increasing theirshare in total demand.

An even more aggressive set of policies,including a significant policy initiative to in-crease energy efficiency and reduce carbonemissions to below their 2005 levels, could seeenergy demand fall even further (table 2.7).

Demand prospects for metalsand mineralsDemand for metals and minerals is alsoclosely related to GDP and the mix of GDP—with manufacturing and investment activities

associated with relatively high commodityintensities. Like oil, the evolution of metalsintensities reflects technological change, thegrowing importance of services in the eco-nomies of high-income countries, and otherstructural changes in demand.

After falling for years, metals intensities indeveloping countries are rising, especiallyin ChinaThe reversal of the trend decline in metals in-tensities that began in the mid-1990s (see fig-ure 2.10) reflects very different trends in high-income countries, most developing countries,and China (figure 2.13). The trend decline ob-served for all three groups between 1970 and1990 has continued among high-income coun-tries, apace with the continued transfer ofcommodity-intensive manufacturing activitiesto developing countries. In developing coun-tries excluding China, the same process hasdriven a slight rise in metal intensities begin-ning in 1992, after their fall attributable to theefficiency improvements associated with theend of the FSU.

China stands out as the country where in-tensities have increased the most. After declin-ing for years, they began to rise gradually to-ward the beginning of the 1990s and thensharply accelerated around 1998, reflecting arapid increase in manufacturing activity and a

Table 2.6 Energy demand is projected toslow in the baseline scenario

Contributions to annual average global growth in energydemand (percentage points)

1990–2005 2005–15 2015–30

World 1.4 1.1 0.6High-income countries 0.7 0.4 0.3Developing countries 2.2 3.4 2.0

Middle-income countries �0.1 2.4 1.5Low-income countries 4.1 3.9 2.2

Shares in total energy demand (percent of total)

1970 1990 2005 2015 2030

Coal 26.0 25.3 25.3 27.8 28.2Oil 44.0 36.7 35.0 32.9 31.5Gas 16.0 19.1 20.6 21.2 22.3Nuclear 1.0 6.0 6.3 5.6 4.8Hydro 2.0 2.1 2.2 2.3 2.3Biomass, waste 11.0 10.3 10.1 9.3 9.1Other renewables — 0.4 0.5 1.0 1.7

Source: World Bank ENVISAGE model (forecast);IEA (historical data).— � Not available.

Table 2.7 Energy demand could declinefurther under more aggressive climatechange policies

Energy source Baseline Stable emissions Aggressive(percent change in energy consumption)

Coal 198 �15 �22Oil 57 10 �29Gas 96 68 25Biomass, waste 48 144 214

Source: IEA 2008a.

Quantity of metals used per unit of GDP, index (1971 5 1.00)

Source: World Bank.

1.6

1.2

1.4

1.0

0.6

0.8

0.2

0.4

1970 1975 1980 1985 1990 1995 2000 2005

Rest of the world

OECD

China

Figure 2.13 Metal intensities have declinedsteadily in high-income countries but havereversed in China since 1993

10363_Pg51_94:10363_Pg51_94 11/29/08 7:07 AM Page 69

sharp uptick in investment. The increase in theChinese investment ratio came partly from theneed to create capacity to meet the manufac-turing boom, but the increase also reflects sig-nificant investment in support of infrastruc-ture in response to increased urbanization(box 2.5).

Except for a few export- and manufactur-ing-intensive Asian economies, other develop-ing countries, including those at much higherlevels of income than China, have not seenmetal intensities rise in this way. Metal inten-sities in Brazil, India, and South Africa, for ex-ample, remained flat or continued to declineduring the same period.15 As a consequence,the strong acceleration in metal demand ob-served in China is not expected to be repeatedin other developing countries.

Not only have Chinese metal intensitiesbeen rising, they are also as much as 7.5 timesas high as in high-income countries and 4times as high as in other developing countries(figure 2.14). While some of the same factors(high investment rate, large manufacturingsector) that explain the increase in Chinese in-tensities likely explain these differences, the

fact that the former Soviet Union also hadsimilarly high intensities before its economictransition suggests that perhaps nonmarketfactors continue to influence allocation ofthese resources in a way not seen elsewhere.


70

China’s accession to the World TradeOrganization and the boom in manufacturing

activity that accession generated certainly played arole in increasing metals demand. However, the long-term investments in new capacity and infrastructurethat began at the same time as WTO accession werelikely just as important. Overall investment in Chinaincreased from 36 percent of GDP in the early 1990sto around 45 percent currently, a result both of in-creased manufacturing and rapid urbanization (overthe same period, the share of the population living incities increased from 30 to 40 percent). Similarly,part of the increase in China’s energy demand wasassociated with an acceleration in steel and cementproduction (Lin and others 2006). These structural

Box 2.5 Understanding the rise in Chinese metalintensities

changes entailed substantial investments in infrastruc-ture and were associated with a rapid increase in au-tomobile production—all heavy consumers of metals.

As of 2007, more than 50 percent of Chinesesteel and 44 percent of copper demand was used inconstruction and infrastructure. While China’sspecialization in manufacturing is likely to persist,investment rates are projected to decline over time(the average life span of infrastructure investmentsexceeds 50 years) so China’s metal intensity is ex-pected to stabilize and then decline, as did the metalintensities of other Asian countries, such as Japanand the Republic of Korea, that followed a manufac-turing- and export-intensive development path(Mitchell, Tan, and Timmer 2007).

0

2

4

6

8

10

Nickel

Metal intensity indexes, 2000–06 averages (high-incomecountries 5 1)

Figure 2.14 Metal intensities in China aremuch higher than elsewhere

Source: World Bank.

High-income countriesDeveloping countries (excluding China)

China

Copper Aluminum Zinc Tin Lead

10363_Pg51_94:10363_Pg51_94 11/29/08 7:07 AM Page 70

Slowing global growth and a declinein Chinese metals intensity should seedemand growth for metals slow over thenext 25 yearsOver the next quarter of a century, metal in-tensities in developing countries are likely tostabilize and begin declining once again. Sev-eral factors should contribute to the reasser-tion of the earlier downward trend.

A slowing in the pace at which global man-ufacturing capacity is transferred to the devel-oping world is projected to result in a levelingoff and eventual decline in manufactures’ sharein Chinese GDP, from about 40 percent in2005 to around 33 percent in 2030. This slow-ing in turn should be reflected in a decline inmetals intensities. Less-rapid growth in manu-facturing and the gradual completion of invest-ment projects are expected to cause the shareof investment in GDP to decline considerably,which should also serve to lower Chinese metalintensities. Finally, the rising influence of mar-ket forces in determining allocation decisionsin China should also cause a drop in the quan-tity of metal used per unit of output.

In the rest of the developing world, similarforces should be at work, which, coupled withrising incomes and increased service-sectordemand, is expected to reduce the metalsintensity of demand.16

Nevertheless, growth in China and devel-oping countries more generally is expected tocontinue to outpace growth in the rest of theworld throughout the projection period.Given China’s high metal intensities, develop-ing-country growth should keep global metalintensities from falling, at least initially. How-ever, the beginning of the decline in Chinesemetal intensities should be reflected in a sig-nificant weakening in the rate of growth ofmetals demand during the period 2015–30.Overall, global demand for metals is expectedto continue to grow somewhat more quicklythan global GDP, at about 4.0 percentthrough 2015, before slowing to around 2.5percent in the period 2015–30, a pace signifi-cantly slower than that of projected GDPgrowth itself.

Demand for food and otheragricultural productsThe weaker growth in population and GDPexpected over the next few decades (see table2.4) should cause global demand for food togrow less quickly over the next 25 years.

Overall, the global population growth rateis projected to decline from an annual averageof 1.6 percent between 1970 and 2005 toabout 1.0 percent over the following 25 years.While most of the slowdown is expected totake place in high-income countries, popula-tion growth rates in every developing regionare expected to decline between 0.4 and 0.8percentage points (figure 2.15).

Rising incomes in developing countriesimply that per capita food consumption willincrease in most of these countries, but theimpact on overall demand is expected to besmall. As the earlier analysis suggested, a10 percent increase in per capita income willincrease grain demand by 6 percent in poorcountries (those with per capita incomesbelow $2,000), but only by 2 percent in mid-dle-income countries.17 Most of the heavilypopulated developing regions have already


71

20.5

0.0

0.5

1.0

1.5

2.0

2.5

3.0

Average annual % change in population

High-in

com

eco

untri

es

Midd

le-inc

ome

coun

tries

Low-in

com

eco

untri

es

Latin

Amer

ica

North

Amer

ica

Ocean

ia

WORLD

Africa

Asia

Europ

e

Figure 2.15 Weaker population growth should slow demand for food

Source: UN 2006.

Population growth, 1970–2005

Population growth projections, 2005–30

10363_Pg51_94:10363_Pg51_94 11/29/08 7:07 AM Page 71

achieved incomes associated with income elas-ticities close to 0.2 (figure 2.16).18

Demand for meat and dairy products (andfeed grains) will likely expand more rapidlybecause these products tend to be more in-come elastic than basic food stuffs.19 For ex-ample, in Asia, demand growth for meat andedible oils outstripped population growth by awide margin over the past 15 years, even ris-ing somewhat faster than GDP in the case ofedible oils (figure 2.17).

Slower population growth will dampendemand for agricultural productsOverall demand for food should slow over thenext few decades, despite income gains. TheFood and Agriculture Organization (FAO)estimates global food demand will increase byabout 1.5 percent a year between now and2030, with cereals, edible oils, and meats grow-ing at 1.2, 2.3, and 1.7 percent, respectively—somewhat slower than they did between 1990and 2006 (table 2.8). Developing countrieshave higher income elasticities, faster incomeand population growth, and relatively largepopulations, compared with high-incomecountries. Thus, three-quarters of the addi-tional global demand for food between nowand 2030 will emanate from developingcountries.

The implications of biofuels demand foragricultural pricesThe production of biofuels in Brazil, theUnited States, and the European Union (whichtogether account for more than 90 percentof global output) has increased by 18 percenta year since 2000. Biofuels now use 16 percentof global sugarcane production, 9 percent ofglobal vegetable oils production, and 13 per-cent of global maize production, and havebeen the key contributor to the rise in foodcrop prices in recent years (Mitchell 2008).


72

0

2

4

6

8

10

Meats

Average annual % growth, 1990–2005

Figure 2.17 Demand for edible oils grew much faster than population in Asia

Source: World Bank.

Wor

ld

Asi

a

Population GDP Cereals Edible oils

BrazilChina

India

RussianFederation

0.4

Grain consumption per capita (kg)(Log scale)

1.1 2.9 8.1 22.0 59.8

GDP per capita (1,000 PPP $)

Note: Curve fitted on a log scale.

18

45

135

367

1,000

Figure 2.16 Per capita grain demand tendsto stop rising when income reaches around $5,000

Source: World Bank.

Table 2.8 Developing countries willaccount for most of the projected demandfor various foods, 2000–30

All Edibleagriculture Cereal oils Meats

WORLD 1.5 1.2 2.3 1.7Developed 0.7 0.9 2.0 —Transition 0.5 0.8 1.7 —Developing 2.0 1.4 2.5 2.4

Sub-Saharan Africa 2.8 2.5 2.9 3.3Middle East and 2.2 2.1 2.3 3.3

North AfricaLatin America and 1.8 1.2 2.6 2.0

the CaribbeanSouth Asia 2.3 1.6 2.7 4.0East Asia and Pacific 1.7 1.2 2.4 2.1

Source: FAO (2006, pp. 33, 39–42, 47).— � Not available.

10363_Pg51_94:10363_Pg51_94 11/29/08 7:07 AM Page 72

The rapid expansion of production capac-ity in the United States and Europe wasprompted by generous subsidies and use man-dates, but high energy prices have made con-tinued production without subsidies profitablein many cases. As a result, demand for biofu-els may mean that in the future prices forcrops used to produce biofuels will be higher,and more volatile, than if these crops wereused only for food.

Indeed, when oil prices exceed the thresh-old of roughly $50 a barrel, a strong correla-tion can be observed between the price ofcrude oil and crop prices that does not existwhen prices are below $50 a barrel (fig-ure 2.18). At oil prices below $50 a barrel,ethanol production is not very profitable.However, at $50 a barrel, a 1 percent increasein oil prices results more or less in a 0.9 per-cent increase in maize prices, because everydollar increase in the price of oil increases theprofitability of ethanol and hence biofuel de-mand for maize.20 Since the oil market ismuch larger than the market for maize (if allthe maize currently produced in the worldwere converted into ethanol, it would equalonly 8 percent of global gasoline supplies), theprice of maize is now effectively determinedby the price of oil.

The impact of biofuels is not limited to thecrops used for biofuel production. As morecropland shifts to produce the now-more-profitable biofuel crops, then the supply ofother crops declines (or less productive land isbrought under cultivation), thus raising foodprices in general. As a consequence, the priceof wheat and soybeans have also become moresensitive to oil prices in excess of $50.

The future impact of the oil market on thedemand for food crops and their prices is un-certain. Technological improvements maylower the cost of producing ethanol, in turnlowering the threshold oil price above whichcrops used for biofuels become sensitive to oilprices. But technological change may also giverise to other nonfood sources (such as cellulose)for biofuel production or to other energy alter-natives such as solar, wind, and hydrogen-based


73

Figure 2.18 Food crop prices have becomesensitive to oil prices

Oil price per barrel versus food price per ton

Crude oil ($/bbl)

b. Wheat vs. Crude Oil Prices

0

100

c. Soybeans vs. Crude Oil Prices

0

100

200

300

400

500

600

700

Crude oil ($/bbl)

Soybeans ($/ton)

0 20 40 60 80 100 120 140

0 20 40 60 80 100 120 140

200

300

400

500

Wheat ($/ton)

Crude oil ($/bbl)

0 20 40 60 80 100 120 140

a. Maize vs. Crude Oil Prices

0

50

100

150

200

250

300

350

Maize ($/ton)

Source: World Bank.

10363_Pg51_94:10363_Pg51_94 11/29/08 7:07 AM Page 73

Indeed, over the past 50 years these forceshave enabled global production of most com-modities to rise despite falling, or at best sta-ble, real prices. Production of aluminum, forexample, increased fivefold between 1965 and2007, while that of crude oil, copper, andwheat increased 2.6, 3.2, and 2.8 times, re-spectively (figure 2.19).

Technological change has kept extractioncosts in check even as the quality of minesand wells declinedAlthough the quality of newly discoveredmines and oil wells (and the ease with whichthey can be exploited) tends to be lower onaverage than older ones, technological im-provements have reduced the cost of produc-ing most commodities over the past 50 years,allowing effective supply to keep pace withdemand (box 2.6).

In the case of oil, declining yields from on-shore wells pushed exploration into offshorefields that are much more difficult and ex-pensive to exploit. Improved technologies al-lowed these sources to be exploited prof-itably even at low prices and even thoughthey are much more challenging to drill thanexisting wells. As a result, nearly all of theadditional increase in global oil productionsince 1978 has come from offshore wells(figure 2.20).21


74

systems. Should this occur, demand for biofuelfood crops would drop off and food priceswith it.

Long-term supply prospects

The slowing of growth should bring com-modity prices down by roughly 25 percent

in 2009 (see chapter 1). But over the mediumto long term, they are not expected to declineto the levels observed in the 1990s. How farthey come down, and their future trajectory,will depend not only on the demand factorsalready discussed but also on the pace at whichfinite resources are exhausted; improvementsin the efficiency with which commodities arefound, extracted, and grown; and the policiesthat are put into place to promote long-termsupply.

Energy and metals supplySupply prospects for both oil and metals de-pend on the competing forces of resource ex-haustion and the declining quality of newsources, on the one hand, and the pace of newdiscoveries and improvements in the technol-ogy with which commodities are discoveredand extracted, on the other.

The world is unlikely to run out of oil,metals, and minerals in the foreseeablefutureDespite ultimately finite quantities of oil, met-als, and minerals in the earth’s crust, there islittle likelihood that the world will run out ofnatural resources (or food) in coming decades.The existence of ample (and growing) re-serves, and a history of significant improve-ments in the technology with which resourcesare found and extracted, suggests that supplywill continue to rise in pace with demand.True resource exhaustion is unlikely not leastbecause, as resources become scarcer, theirprices rise, consumption declines, and alterna-tives that once may have been uneconomic aresubstituted for the scarce (and expensive)commodity.

Output volumes, 1965–2007, index (1965 5100)

Source: World Bank.

100

150

200

250

300

350

400

450

500

550

2005

Crude oil

Aluminum

Copper

Wheat

Figure 2.19 Output of virtually all commodi-ties has increased since 1965

1965 1970 1975 1980 1985 1990 1995 2000

10363_Pg51_94:10363_Pg51_94 11/29/08 7:07 AM Page 74

Technology has also helped maintainsurprisingly stable ratios of reservesto outputAdvances in the technology with which newreserves are discovered and in the efficiencywith which the final product is extracted from

ore beds or wells has meant that known re-serves of most extractive commodities have in-creased over time—despite rising production.

Such technological improvements help ex-plain the substantial rise in estimates of re-serves over past decades. Two authoritativesources of such data for oil are the Oil andGas Journal, which reports annual estimatesof proven reserves (figure 2.21), and the


75

Rising costs for producing a unit of output repre-sent a good a priori indicator of increasing

scarcity. The fact that the prices of most commodi-ties have remained stable or declined for most of thepast 100 years is therefore a good indicator that atleast until 2003 the world was not running out ofthem (Radetzki 2008a).

Production costs—especially for the marginalproducer—are an even better indicator. For the me-dian producer, the real cost of producing a ton ofmetal between 1985 and 2002 declined by 28 per-cent for aluminum and copper and by 21 percent fornickel (IMF 2006). For high-cost producers, thedecline was the same for aluminum but was only18 percent for copper and nickel. Those numberssuggest that while new projects to extract copperand nickel were more expensive than preexisting

Box 2.6 Declining costs of resource extractionones, technological change had nevertheless reducedthe costs of production by more than the lowerquality of the underlying vein or its remoteness hadraised them.

Similarly, the average cost of bringing a new oilfield into production declined from $29 a barrel in1981 to $9 in 1999 (IEA 2001). These cost reduc-tions would be all the more marked if the numberswere expressed in real terms. And although not allof this cost decline can be attributed to technologicalchange, much can (Bohi 1999). Indeed, improve-ments in extractive technology allowed copper pricesto decline more or less continuously between 1890and 1970 even as the average grade of copper ore inthe United States fell from 6 percent to less than2 percent between 1890 and 1920 and to less than1 percent by 1960 (Lowell 1970).

0

15

30

45

60

75

90

Figure 2.20 Almost all of the additional oilsupply since the 1970s has come from nontraditional sources

Source: Sandrea and Sandrea 2007.

World oil production, millions of barrels per day

Offshore Onshore

1970 1974 1978 1982 1986 1990 1994 1998 2002

0

200

400

600

800

1,000

1,200

1,400

World crude oil reserves, billions of barrels

Figure 2.21 Rather than declining, known oil reserves keep rising

Source: Oil and Gas Journal.

1950 1958 1966 1974 1982 1990 1998 2006

10363_Pg51_94:10363_Pg51_94 11/29/08 7:07 AM Page 75

United States Geological Survey (USGS),which attempts to quantify the resource baseof the world’s major basins by including as-sessments of known reserves, undiscovered re-sources, and reserve growth (table 2.9).22

Estimates from the Oil and Gas Journal,which include unconventional sources of hy-drocarbon fuels such as Canadian oil sandsand oil shale, show known reserves risingfrom just over 600 billion barrels in 1980 to1.3 trillion barrels by 2008. Furthermore, thereserve estimates for a number of major pro-ducers in the Middle East have not changedfor years, reflecting both their current sizecompared with production levels (national-reserve-to-production levels imply adequatereserves for 82 years of production for theMiddle East–producing countries) and the factthat for decades these countries have not feltan incentive to explore in more depth thepotential for additional reserves nor to verifyexisting reserve estimates.

On the other hand, according to the USGS es-timates (which include undiscovered resources),

total resources increased from 1.7 trillion bar-rels in 1981 to 3.0 trillion barrels in 1996, so theamount of known oil still in the ground re-mained stable, at around 70 percent of the totalof oil ever found (see table 2.9).

Reserves of natural gas estimates are equallyhigh. According to the USGS, they were at 2.3trillion barrels of oil equivalent in 2003, almostas large as crude oil reserves (figure 2.22).

Yet, among the key hydrocarbon sources ofenergy, coal is perhaps the most abundant. Asof 2007, the reserves-to-production ratio wasestimated at 133 years, according to BP. How-ever, as mentioned earlier, the use of coal willdepend on the degree to which new techno-logical advances will be able to ameliorate theenvironmental concerns.

Finally, expansion of nuclear energy (andother renewable fuel) supplies could lessen therelative importance of hydrocarbon-basedfuels. For example, in addition to the exis-tence of abundant feed stocks (at current con-sumption rates, known uranium reserves areexpected to last almost a century), currentmodern nuclear technologies not only producemuch less nuclear waste but also have lowerlikelihood of accidents compared with nuclearpower plants in the past.


76

Table 2.9 Historically, estimates of oilreserves have kept pace with production

Date of assessment

Category 1981 1985 1990 1993 1996

(billions of barrels)

Cumulative 445 524 629 699 710production

Known reserves 724 795 1,053 1,103 891

Undiscovered 550 425 489 471 732conventionalresources

Expected reserve — — — — 688growth

Estimated total 1,719 1,744 2,171 2,273 3,021resources

Total resources 74 70 71 69 76still in ground(percent)

Source: U.S. Geological Survey, World Bank calculations.Note: Estimated total resources is the sum of the first threerows. Total resources still in ground is one minus the ratio ofcumulative production over total resources.— � Data are not available (the concept of reserve growthwas first introduced in 1996).

0

500

1,000

1,500

2,000

2,500

Billions of barrels (equivalent)

Figure 2.22 Gas reserves are almost as largeas oil reserves

Source: U.S. Geological Survey.

Reservegrowth

Undiscoveredresources

Knownreserves

Oil Gas

10363_Pg51_94:10363_Pg51_94 11/29/08 7:07 AM Page 76

Reserves of metals and minerals have alsotended to rise with outputThe story for metals and minerals is somewhatmore nuanced. Reserves expressed as a shareof production for a number of metals diddecline during the 1980s and 1990s. In partthis reflected their relative abundance (re-serves exceeded more than 40 years for baux-ite, copper, iron ore, and nickel), continuedrising production levels, declining prices, andunderinvestment. It also reflected the fact thatreserves are really a measure of the inventorythat producers have readily available for fu-ture delivery, rather than a measure of thephysical quantity remaining of a commodity.With demand and prices weak, and invento-ries (reserves) ample, firms had little incentiveto invest in additional inventory.

Since 2003, when metal prices began risingand production accelerated, explorationexpenditures have picked up (see earlier dis-cussion). For some metals, the reserve-to-production ratios have increased as a result(table 2.10).

Increasing scarcity is unlikely to result inresource exhaustionAlthough the history of reserves data suggeststhat much more oil is likely to be discovered,ultimately the quantity of available oil is finite.

Long before the world begins to run out of oil,however, prices would begin to rise and con-sumption growth would slow. As a result, al-ternatives such as natural gas, nuclear power,and renewable energy sources would increasetheir output share (see earlier discussion onlong-term demand). Reserves of crude oilwould not decline as rapidly as they wouldhave had prices not increased, and its usewould be reserved for those products (plastics,chemicals, and polymers) where few alterna-tives exist.

Overall, high prices will encourageincreased supply and substitution ofalternative sourcesAs indicated earlier, the supply of crude oil isexpected to continue to expand over the nextfew decades, reaching about 112 mb/d by2030. The supply of other energy sources isexpected to increase more rapidly than thatfor oil, with coal and natural gas projected toincrease their shares in total energy supplyfrom 46 percent in 2005 to 51 percent in2030. Renewable energy sources are projectedto see their share in total energy supply risefrom about 0.45 percent to about 1.7 percentover the same period (table 2.11).

Biofuels are a source of renewable energywhose share of global liquids production has


77

Table 2.10 Increased investment has stabilized reserve-to-production ratios for somecommodities

Year Oil Coal Bauxite Iron ore Copper Lead Nickel Tin Zinc

Proven reservesbillions of

barrels (Millions of metric tons)

1980 667 — 25,000 250,000 493 127 55 10 1621990 1,003 — 22,000 150,000 350 70 49 8 1472000 1,104 984,211 24,000 140,000 340 64 58 7 1902007 1,238 847,488 25,000 150,000 490 79 67 6 180

Reserves/production ratio(Years of production equivalent)

1980 29 — 280 280 64 36 77 42 261990 42 — 193 178 41 20 53 37 212000 40 230 178 132 26 21 46 29 222007 42 133 132 79 31 22 40 20 17

Source: Radetzki (2008a, 2008b), British Petroleum, U.S. Geological Survey.

10363_Pg51_94:10363_Pg51_94 11/29/08 7:07 AM Page 77

reached 1.6 percent, largely because of govern-ment encouragement (box 2.7). Recent projec-tions suggest that biofuel production will reachthe equivalent of 1.95 mb/d of oil by 2013 (a45 percent increase over 2008), correspondingto 2.1 percent of the projected global oil de-mand (IEA 2008b; FAPRI 2008).23 While pop-ular, biofuels are controversial, in part becauseenergy is required to produce energy, so the netaddition to the global energy supply fromcorn-based ethanol is relatively small (Kojima,Mitchell, and Ward 2006), and in part becausebiofuels yield only limited environmental ben-efits (Searchinger and others 2008; Fargioneand others 2008).

Long-term projections for metals and min-erals supplies are optimistic, with expectationsthat production will increase by a further 3.0percent a year between now and 2030. At thesame time, the trend toward substitution of al-ternative metal products is likely to continue.For example, copper initially displaced lead inplumbing applications, only to be displaced by

plastics most recently and by sand (fiber op-tics) in telecommunications applications. Andthe rapid expansion in demand for aluminum,shown in figure 2.19, partly reflects its increas-ing use as a lightweight alternative for steel.

Another element of growing importance inthe metals markets is the role of recycling,which currently ranges from 55 percent offinal demand in the case of lead to about5 percent in the case of zinc. In developedeconomies, the proportion of metal availablefrom scrap is higher because of greater inven-tories embodied in old cars and infrastructurethat can be recycled. Future increases inscrap’s share of the metal supply in emergingeconomies will slow the rate of growth of de-mand for mined metal.24

Actual results will depend on policychoices and technological progressSupply of both energy and metals over thelong term depends critically on policies andthe pace of technological change. Rising con-cerns about the environmental consequencesof economic activity, notably but not exclu-sively those associated with climate change,may alter the regulatory environment in im-portant ways.

Emissions abatement policies may restrictthe use of hydrocarbons, either through man-dates or tax policy that alters the economicsof both demand and supply—potentially ex-tending reserve-to-production ratios signifi-cantly. Environmental concerns may also re-strict the use of extraction and productiontechniques in other primary sectors in waysthat reduce supply or significantly raise pro-duction costs. In the IEA’s aggressive emissionsabatement scenarios, global oil demand fallsby 29 percent.

How successful alternative fuels andimproved extraction technologies will be inenabling the kind of substitution and in-creased supply that has been observed in thepast will depend on how successful policy is insupporting the creation and diffusion of newtechnologies. Particularly important for poorcountries will be efforts to create affordable


78

Table 2.11 Oil’s share in global energysupply is projected to decline

Average annual growth rate(%)

Energy source 1990–2005 2005–15 2015–30

Coal 1.8 3.3 1.5Oil 1.5 1.7 1.1Gas 2.3 2.6 1.7Nuclear 2.1 1.1 0.4Hydro 2.1 2.7 1.6Biomass & Waste 1.6 1.5 1.3Other renewables 3.8 9.0 5.2

Total 1.8 2.3 1.4

Share in total energy supply (percent)

1990 2005 2015 2030

Coal 25.3 25.3 27.8 28.2Oil 36.7 35.0 32.9 31.5Gas 19.1 20.6 21.2 22.3Nuclear 6.0 6.3 5.6 4.8Hydro 2.1 2.2 2.3 2.3Biomass, waste 10.3 10.1 9.3 9.1Other renewables 0.4 0.5 1.0 1.7

Total 100 100 100 100

Source: IEA 2008a.

10363_Pg51_94:10363_Pg51_94 11/29/08 7:07 AM Page 78

and durable solar cells, whereas at the globallevel efforts to reduce dependence on liquidsfor transportation—such as a breakthrough inbattery technologies or hydrogen generation—will be key.25

The structure of energy markets, includ-ing the market power and supply decisionsmade by OPEC may also play a role. Theconcentration of oil reserves in the hands ofa few countries could limit the increase inexploration and production anticipated in re-sponse to high prices (box 2.8). OPEC con-trols three-quarters of the world’s oil reservesand dominates export markets.26 Moreover, anumber of producers have made their reservesand fields off limits to private investors; twoof these countries (Mexico and Saudi Arabia)officially prohibit the participation of foreign

companies, even in a consultancy capacity. Inthe baseline scenario, more than 75 percent ofthe increase in global production is expectedto come from OPEC member countries.Should they decide to restrict supply, oil pricescould be sharply higher in the medium term,and demand much lower. Although such anepisode would likely be very painful, ulti-mately it would speed the switch into alterna-tive energy sources (much as it did in the1980s) and result in a significant decline in thelong-term demand for oil.

Agricultural supplyIncreases in cultivated land and yields arelikely to result in strong growth in agriculturalproduction and declines in prices from theircurrent high levels, as has occurred during the


79

While biofuels have been used since the earlydays of the automobile (Henry Ford’s 1908

Model T car was designed to run on maize-basedethanol), limited supplies and the availability ofcheaper and more efficient petroleum productsdiminished the use of biofuels (except for a briefrevival during the petroleum shortages of WorldWar II).

In the United States, various amendments to the1970 Clean Air Act and the 1992 Energy Policy Actwere instituted that favored the use of biofuels, espe-cially maize-based ethanol. More recently, the 2007Energy Independence and Security Act called for afourfold increase in biofuel production by 2022. Asa result, an estimated 25 percent of U.S. maize out-put in 2007–08 was diverted to ethanol production.In 2007, the United States produced 6.6 billion gal-lons of ethanol, roughly equivalent to 4.5 percent ofits gasoline consumption.

The European Union began instituting mandatoryuse of biodiesel (mostly from rapeseed oil) as early as1992. During 2008, its biofuel output was expectedto reach 225,000 barrels a day of oil equivalent, rep-resenting about 1.5 percent of its crude oil consump-tion. The European Union has a target 5.75 percent

Box 2.7 The rise of biofuel productionof biofuel use by 2010, whereas a 2008 EuropeanCommission directive proposed a 10 percent usemandate by 2020.

In the 1970s, Brazil offered incentives to bothsugarcane producers and its car industry to encour-age biofuels, and by the mid-1980s, Brazil was pro-ducing 3 billion gallons of sugarcane-based ethanola year, while 90 percent of Brazilian-made cars weredesigned to run on ethanol. The biofuel programalmost collapsed during the 1990s when the price ofoil was low, offshore oil discoveries weakened politi-cal support for biofuels, and high sugar pricesstrained the subsidy program and diverted sugarcaneto the world market. However, the recent crude oilprice spike along with the introduction of “flex-fuel”cars that can use any combination of gas and ethanolhas encouraged reliance on ethanol.

In Brazil, the government no longer providessubsidies to either the car or the sugar industry, andthe cost of producing ethanol is $1.40 a gallon (verylow compared with maize-based ethanol or edibleoil–based biodiesel), making the industry competitiveeven if crude oil prices decline to $40 a barrel(Kojima and Johnson 2005).

10363_Pg51_94:10363_Pg51_94 11/29/08 7:07 AM Page 79

past 50 years. However, supply growth willremain sensitive to public policy as well as toinvestments in infrastructure and research.Furthermore, prospects are subject to signifi-cant risks, both upside (rapid technologicalchange) and downside (impacts of environ-ment and climate change and links to oilprices through inputs and biofuels demand).

Rising productivity and land undercultivation have boosted agriculturalproductionThe past half century has witnessed a steadyincrease in agricultural output, both in ab-solute and per capita terms. Total factorproductivity in the agricultural sector has in-creased by between 2.1 and 2.5 percent eachyear (Coelli and Rao 2005; Martin and Mitra2001) over the past 20 years, with the largestproductivity gains recorded in Asia and NorthAmerica (figure 2.23).

Reflecting this strong productivity growth,most of the increase in agricultural output overthe past 40 years is attributable to increasedyields rather than to increases in the quantityof cropped land (figure 2.24). Similar gainswere observed in the livestock sector, with thequantity of meat produced per animal rising by

1.7 percent for chicken and 3.5 percent forpork between 1980 and 2005 (FAO 2006).Growth in productivity was responsible forhalf of the increase in output since 1960 inChina and India and between 30 and 40 per-cent of the increase in other East Asian coun-tries (World Bank 2008). These productivityimprovements enabled a decline in the share oflabor force employed in agriculture (even as


80

The rising share of oil reserves and global produc-tion controlled by state-owned firms has

prompted concerns about future supply. Theconcerns are about:

• Cartel-like behavior• The efficiency and responsiveness of state-owned

firms to economic incentives• The denial of access to multinational firms, which

have historically been among the most efficient

State-owned firms need not be less responsive orless efficient than privately owned ones. To maximizeproductivity, however, policy makers need to ensurethat government-owned or -controlled firms are notoverburdened with very high effective tax rates

Box 2.8 State-owned firms and output efficiency(including profit remittances to the state andobligations to sell oil at below-market prices) or socialmandates that limit the extent to which they are ableto invest in new technologies, infrastructure, and fields.

In some countries, such responsibilities have beenassociated with disappointing results. For example,oil production in República Bolivariana de Venezuelahas declined 19 percent since 2000, while it has beenstagnant and is now declining in Mexico; both arecountries with restrictive legislation or practice.

This contrasts with the 45 percent increase inproduction and 49 percent increase in reserves (notincluding the new Tupi field) recorded by Brazil’sstate-owned Petrobras, which has been encouraged toreinvest profits and hire foreign experts when needed.

0.0

1.0

0.5

1.5

2.0

3.0

2.5

3.5

Figure 2.23 Agricultural productivity has been rising rapidly over the past 20 years

SouthAmerica

AsiaAfrica Austral-asia

NorthAmerica

Source: Coelli and Rao 2005.

Europe

Average, annual percentage change in agricultural TFP, 1980–2000

10363_Pg51_94:10363_Pg51_94 11/29/08 7:07 AM Page 80

production and population increased) and a25 percent increase in the average caloric percapita consumption in developing countriesduring the past 30 years.27

Among developing countries, crop produc-tivity increases (which control for increases ininputs such as capital and labor) have beendriven mainly by the expansion of irrigation,improved seed varieties, and increased use offertilizer. Worldwide, the area devoted to im-proved varieties has been expanding continu-ously. In 2000, high-yielding grain varietieswere used on 90 percent of planted area inSouth and East Asia, up from 10 percent in1970 (World Bank 2007b). The use of im-proved varieties is expanding in all regions, in-cluding Sub-Saharan Africa, where it now rep-resents almost one-quarter of cropped land.

Fertilizer use is also up. In developingcountries it has risen from only 10 percent ofglobal use in the 1960s to 77 percent now(FAO 2008). However, fertilizer use in sub-Saharan Africa is minimal, accounting for lessthan 3 percent of global use versus a 40 per-cent share in East Asia.

Most recently, yield growth has declinedfor some commodities, notably wheat, rice,and soybeans (figure 2.25). While such weak-ening in yield gains has been attributed to

exhaustion of the gains that came from the in-troduction of green revolution technologies,persistently low commodity prices have alsoplayed a role. Yields gains in other commodi-ties have accelerated because of greater use ofgenetically modified varieties, which boostedyields in cotton in China and India by 19 and26 percent, respectively (World Bank 2007b).In addition, maize yields have benefited fromthe more extensive use of techniques madeeconomically profitable by high prices.

The recent slowing of productivity gainsand the spike in food prices have raised con-cerns about long-term output trends. Fears of afood shortage over the long term are unwar-ranted, however, given the enormous potentialfor increasing agricultural output through cul-tivating unused land and increases in yields.

Although much of the best agriculturalland is already in use, significant opportunitiesfor increasing output remain simply by in-creasing the amount of land under cultivation.About 12 percent of arable land worldwidethat is not currently forested could be broughtinto agricultural production relatively easily(Thompson 2008). Considerable amounts ofarable and unforested land in Africa could bebrought into production assuming appropri-ate infrastructure were put into place, while in


81

0.0

1.0

2.0

3.0

4.0

5.0

Contribution to average annual increase in output, 1965–2008

Figure 2.24 For key crops, most of theincrease in output was due to increasedyield, not increased area planted

Rice MaizeWheat CottonSoybeans

Source: World Bank calculations based on U.S. Department of Agriculture data.

Area growth

Yield growth

0.0

Wheat

Source: World Bank calculations based on U.S. Department ofAgriculture data.

Rice Maize Soybeans Cotton

0.5

1.0

1.5

2.0

2.5

3.0

3.5

4.0

Annual % change in yields, 1965–99, 2000–08

Figure 2.25 Yield growth has deceleratedrecently

1965–1999 2000–2008

10363_Pg51_94:10363_Pg51_94 11/29/08 7:07 AM Page 81

Brazil about 180 million arable hectares thatare currently used as pasture could eventuallybe brought into food crops. Sizable amountsof unused or underutilized land also exist inUkraine and Russia.28

Another source of additional farmland isthe 18 million hectares in the United States andEurope that have been set aside to reduce sup-ply and keep producer prices high (Normila,Effland, and Young 2004). Recent changes tothe Common Agricultural Policy have autho-rized European farmers to use about half ofthat land, which could see the amount of landapplied to agriculture in Europe rise by 3.5percent this year. Similarly, the United Statesrecently released 1.5 million acres of the landfallowed by its conservation program.

However, the new land is less productivethan existing land and will be more costly toexploit, especially in an environment of highprices for energy and equipment. Further-more, the expansion of new land (especially inAfrica) will require large investments in infra-structure and likely will take decades to ex-pand significantly.

These calculations do not include land thatis currently forested but that is suitable forrainfed crop production. Such lands exceed byone and one-half times the total currently usedfor agriculture (figure 2.26). Bringing all ofthis land into crop production is probably nei-ther desirable nor likely, but its existencemeans that the agricultural supply potential ofthe planet is far from exhausted.

Technological gains are likely to drivecontinued increases in yieldsMuch of the increase in agricultural productiv-ity over the past 50 years came about throughoften scientifically simple improvements inagricultural technique, including increased useof irrigation, fertilizers, and commerciallyoptimized seeds. The adoption of these tech-niques in the developing world is most ad-vanced in Asia, and its impact on yields is evi-dent in the very strong productivity growthenjoyed by the region over the past half cen-tury (table 2.12). Considerable potential exists

for extending the same kind of gains to otherregions, particularly Sub-Saharan Africa andmany countries in Europe and Central Asia,that have adopted these techniques less exten-sively (table 2.13).29 However, such expansionwill require policies to encourage research and


82

0

200

400

600

800

1,000

1,200

High-in

com

e

coun

tries

East A

siaan

d

Pacific

Europ

ean

d

Centra

l Asia

Latin

Amer

ica

and

Caribb

ean

Midd

leEas

t and

North

Africa

Sub-S

ahar

anAfri

ca

Millions of hectares of arable rain-fed cropland

Figure 2.26 The stock of unused but poten-tially arable land is enormous

Already in use

Source: Food and Agriculture Organization.

Currently unused

South

Asia

Table 2.12 Potential gains from extendingthe green revolution remain large

Potentialgain

(PercentActual Potential Poential

ofRegion production production gain

current(Millions of metric tons) production)

High income 423 440 17 3.9East Asia and

the Pacific 501 508 7 1.4Europe and

Central Asia 130 191 60 46.5Latin America and

the Caribbean 140 161 21 15.0Middle East and

North Africa 50 57 7 14.3South Asia 250 259 9 3.7Sub-Saharan

Africa 56 81 25 43.9Total 1,551 1,697 146 9.4

Source: World Bank.

10363_Pg51_94:10363_Pg51_94 11/29/08 7:07 AM Page 82

development (R&D) and extension directedparticularly at small-holders. If these countrieswere to adopt more intensive techniques likethose used in Asia and elsewhere, global pro-duction of cereals could be increased by asmuch as 9.4 percent, enough to meet severalyears’ worth of increasing global demand.

Based on similar observations, the FAO inits most recent long-term forecasting exerciseexpects global agricultural production to riseby 1.5 percent a year for the next threedecades, somewhat slower than over the past50 years but still significantly faster than pro-jected population growth.

Prospects will depend on a number ofuncertain factorsOf course, the long-term supply prospects foragricultural commodities are far from certain.Past productivity gains are an imperfect indi-cator of what might be expected in the future.Moreover, a number of looming issues in theglobal economy could affect supply conditionsin important ways.

Public investment in infrastructure andR&D will be critical to realizing potential

productivity gains and to ensuring that suchgains benefit the poor. About 95 percent ofdeveloping-country R&D expenditures in agri-culture is publicly funded. As a result, thisR&D is mainly dependent on administrativedecisions, which may or may not respond tomarket conditions. Therefore, it is imperativethat efforts to increase food production inlow-income countries should be part of acomprehensive effort that includes investmentin R&D as well as dissemination efforts.30

Notwithstanding the swings in the prices ofagricultural products, agricultural R&D hasremained remarkably stable as a share of agri-cultural value added, at about 0.85 percent be-tween 1981 and 2000. Moreover, developingcountries are spending much less on R&Dthan are high-income countries, both in ab-solute terms and as a share of agricultural GDP(figure 2.27).

Recent advances in biotechnology may offerdeveloping countries additional improvementsin yields through the introduction of new plantvarieties with heightened resistance to drought,rain, diseases, and pestilence—characteristics


83

Table 2.13 With some exceptions, yieldgrowth for key agricultural commoditieshas been highest in South and East Asia

Category Wheat Rice Maize Soybeans Cotton

(Annual percent change in yields, 1965–2006)

World 2.0 1.7 1.8 1.5 1.7Income level

High income 1.6 0.9 1.6 1.3 1.6Middle income 2.0 1.9 2.6 2.8 2.3Low income 2.6 2.0 1.1 1.4 3.1

RegionEast Asia and

the Pacific 3.8 1.8 2.9 1.9 2.7Europe and

Central Asia 0.1 0.0 0.8 �0.1 0.7Latin America and

the Caribbean 2.0 2.5 2.6 1.3 2.1Middle East and

North Africa 2.5 1.2 2.7 3.0 1.2South Asia 2.6 2.1 1.6 1.5 3.1Sub-Saharan Africa 2.2 0.7 0.7 3.2 1.6

Source: World Bank calculations based on U.S. Departmentof Agriculture data.

Sub-S

ahar

an

Africa

Asiaan

dPac

ific

Wes

t Afri

caan

d

North

Africa

Latin

Amer

ica

and

Caribb

ean

Develo

ping

coun

tries

High-in

com

e

coun

tries

0.0

0.5

1.0

1.5

2.0

2.5

R&D spending as a share of agricultural GDP

Figure 2.27 Developing countries spend less on agricultural R&D than high-income countries

Source: Pardey, Alston, and Jones 2008.

1991

Wor

ld

1981 2000

10363_Pg51_94:10363_Pg51_94 11/29/08 7:07 AM Page 83

that might be especially desirable during a pe-riod of climate change. However, like the chem-ical-based pesticides and fertilizers that helpedgenerate substantial improvements to yieldsduring the green revolution, they may alsocarry with them hidden risks such as cross-plant genetic contamination and potentialhealth impacts because of unexpected interac-tions with human biology. Transparent andcost-effective regulatory systems that inspirepublic confidence will be needed to evaluaterisks and benefits on a case-by-case basis.

Moreover, the diffusion of these innova-tions into developing countries has been un-even, partly because of the high cost of theseseeds and their incompatibility with traditionalagricultural methods and partly because of theunwillingness of seed companies to marketthem into countries with weak regulatoryframeworks and intellectual property regimes(box 2.9).

In the long term, climate change andwater scarcity could have significantimpacts on yieldsGlobal temperatures are expected to rise by0.4 degrees Celsius between now and 2030.This could lead to an overall decline in agri-cultural productivity of between 1 and 10 per-cent by 2030 (compared with a counterfactualwhere average global temperatures remainedstable), with India, Sub-Saharan Africa, andparts of Latin America being most affected(see next section).

Over the longer term, the impacts of cli-mate change could be much more serious,with agricultural productivity in many devel-oping regions, notably Africa, potentiallydeclining by as much as 25 percent as com-pared with a baseline of temperatures remain-ing stable at their 2030 levels (Cline 2007).

Sustainable water supply forms anotherlonger-term risk facing future agricultural


84

The most important recent technological break-through in agriculture has been the development

of genetically modified (GM) crops. These crops tendto be more disease resistant than traditional varietiesand lower cost because of increased yields and theneed for fewer pesticides.

Originally developed in the United States, theyhave spread to many countries, including many inthe developing world. In 2006, farmers in 22 coun-tries planted GM seeds on 100 million hectares,corresponding to about 8 percent of global crop area(World Bank 2007b). Although GM crops were ini-tially taken up by commercial farming, increasinglysmall farmers are now using the technology.

Yet GM crops have not been adopted widely indeveloping countries despite considerable potential incrops, such as bananas, that suffer large losses fromdisease. This lack of uptake results partly from con-cerns over environmental and food safety risks andpartly from private producers of these seeds thatare unwilling to allow them to be distributed in

Box 2.9 Genetically modified crops—the next greenrevolution?

countries where they are unable to enforce theirproperty rights.

Some countries (such as China) have gotten aroundthis problem by developing their own varieties in pub-lic research agencies that they make available to small-holders. Other countries (such as Burkina Faso) haveentered into agreements with private companies thatallow them to develop GM seeds to be used by small-holders under a general licensing agreement.

The current generation of GM technology hasconcentrated on internalizing resistance to diseasesand pests. Research in the pipeline, however,focuses on developing varieties with other charac-teristics such as increased tolerance to drought,wetness, and temperature, as well as slowingproduct deterioration. As a result, whereas thefirst generation of GM crops was tailored to theagriculture of the developed world, the secondgeneration may be better suited to resolving thekinds of problems found in the production systemsof developing countries.

10363_Pg51_94:10363_Pg51_94 11/29/08 7:07 AM Page 84

supply. About 85 percent of water use indeveloping countries goes to agriculture, withless than one-fifth of the cultivated area in de-veloping countries producing two-fifths of thevalue of agricultural output (World Bank2007b). Already 15–35 percent of water with-drawals worldwide are not sustainable, in thesense that the amount being withdrawn fromaquifers or rivers exceeds the rate at which thesource is naturally resupplied. Perhaps themost notable example of unsustainable usewas the rapid expansion of cotton productionin the Aral Sea basin, which has resulted in thedisappearance of 90 percent of the sea’s sur-face area and a broadly based environmentaldisaster. Improving water management will re-quire countries to take more responsibility forshared water resources, ensuring that they arepriced appropriately and that adequate watermanagement institutions are put in place toprevent a recurrence.

Projections

As anyone following commodity marketsover the recent past can attest, forecasting

future demand, supply, and prices in commod-ity markets is—at best—a hazardous under-taking. While some commodities, especiallyextracted commodities such as oil and metals,may become more scarce in coming decades,there is little likelihood of a serious shortfall insupply. Nevertheless, the overall balance be-tween demand and supply is very uncertain.It will depend on a wide range of factors,including climate change, productivity develop-ments in commodity supply and commoditydemand markets, GDP and population growth,and the policy environment. The remainder ofthis chapter attempts to quantify the range ofpossible outcomes in commodity markets.

Agricultural prices are likely to declineover the long termAs discussed in chapter 1, agricultural pricesare forecast to decline over the next two yearsbut remain well above the levels of the first halfof this decade. While the long-term outlook for

agricultural prices is particularly uncertain, thisdecline is expected to continue through theforecast period.

As outlined earlier, the growth in demandfor agricultural products is expected to besomewhat weaker in the next several decadesbecause of slower population growth and thelimited impact of higher incomes on fooddemand. On the supply side, the availabilityof additional land and further productivityimprovements should enable production tokeep pace with demand even as the agricul-tural sector continues to release labor to workin other parts of the global economy.

Based on long-term forecasts of populationand incomes and a continuation of the histori-cal experience of rising productivity, annual de-mand and supply are projected to grow byabout 1.7 percent on average between 2008 and2030. This would imply a continued decline inagricultural prices of about 0.7 percent a yearrelative to manufacturing prices and the share ofthe unskilled labor force working in agriculturedeclines by 6 percentage points (table 2.14).

One particularly difficult issue in the long-term forecasts for agricultural production andprices concerns the impact of climate change.Human-induced global warming has begun tochange growing conditions around the world,particularly in developing countries. In manycountries, and for many crops, ideal growingtemperatures have been surpassed, stressing thegrowth of plants. Perhaps more significantly,more-extreme water-related events are occur-ring, including more periods of persistentdroughts, drier soils from higher temperatures,changing patterns of rainfall (for example themonsoon arriving earlier or later), and more se-vere rainfall falling in shorter periods. These cli-mate events can reduce immediate productionand impair agricultural development, as poorfarmers faced with drought may be forced tosell or eat animals, while severe storms damageother types of capital such as irrigation canals.

Forecasts of the rise in temperature and theimpact on agriculture over the next twodecades are extremely uncertain. Lobell andothers (2008) anticipate that southern Africa,


85

10363_Pg51_94:10363_Pg51_94 11/29/08 7:07 AM Page 85

South Asia, and parts of Latin America willrank among the hardest-hit areas, with maizeproduction in southern Africa, for example,potentially falling as much as 30 percentbelow what it would have been without cli-mate change by 2030.

Our base case in table 2.14 assumes signif-icant damage from climate change over thelong run. However, over the projection period2030, the impacts are relatively modest. Todate, global temperatures have risen 0.8° Csince 1900 and are projected to rise a further0.4° C by 2030 (Cline 2007). Scaling Cline’s2080 estimates of damage to agriculture bythe estimated temperature change in 2030leads to an overall decline in agricultural pro-ductivity of between 1 and 10 percent by 2030(compared with a future where average globaltemperatures remain stable), with Canada andEurope least affected and India, Sub-SaharanAfrica, and parts of Latin America most af-fected.31 Were there to be no climate changebetween now and 2030, global agriculturalproductivity would be nearly 4 percent higher,and the world price of food 5.3 percent lower.

These projections are subject to other im-portant uncertainties. In particular, the pro-jected productivity gains are contingent onpolicies being put in place that permit produc-tivity gains to continue rising as they have inthe recent past. The policies include the re-moval of trade distortions, progress to limitthe increase in carbon emissions, construction

of infrastructure, and R&D investments in de-veloping countries with lagging productivity.

As shown in Scenario I, should global agri-cultural productivity rise by only 1.2 percent ayear on average instead of the 2.1 percent pro-jected in the baseline, then prices, rather thandeclining, can be expected to rise by as much as0.3 percent a year relative to manufactures—reversing the trend decline of the past 100 years.Reduced productivity includes increasing thequantity of cereal required to produce meat andas a result total agricultural output rises, eventhough final consumption declines by 0.3 per-cent per annum. Final demand does not declineby more, because lower productivity is partiallycompensated for by increased inputs, includinga 1.2 percentage point increase in the share ofagricultural workers in the labor force com-pared with the base case. Overall, by the end ofthe projection period, real incomes in develop-ing countries would be lower by about 3.4 per-cent compared with the baseline.

Consistent with Scenario II, should theweaker productivity be limited to developingcountries, in part because climate change is ex-pected to affect them more adversely and per-haps because policy fails to step up infrastruc-ture, R&D, and dissemination of investments,the overall impact in markets would be atten-uated somewhat. Prices would fall by only0.17 percent a year (compared with a declineof 0.7 percent a year in the base case), andagricultural sector employment would rise


86

Table 2.14 Agricultural sector simulation results, 2005–30

Results in 2030 by scenario

I. II. III.Global Developing- Strong demand

Baseline productivity slowdown country slowdown for biofuels

Total factor productivitya

Developing countries 2.1 1.2 1.2 2.1High-income countries 2.1 1.2 2.1 2.1

Outputa 1.7 1.9 1.9 2.4Pricesa �0.7 0.3 �0.1 �0.5Employment (unskilled in develoing countries)a �6.0 �4.8 �5.3 �5.4Change in real incomeb �3.4 �2.3 �1.8

Source: World Bank ENVISAGE model.a. Change in share of total employment between 2005 and 2030.b. Percent of base income.

10363_Pg51_94:10363_Pg51_94 11/29/08 7:07 AM Page 86

slightly compared with the baseline. But devel-oping countries, especially those whose popu-lations continue to grow relatively rapidlywould become much more dependent on high-income countries for their food supply.

Scenario III examines the potential impactof biofuels production on food prices. Whilebiofuels have made a major contribution tothe rise in food prices over the past two years,their impact in the future is difficult to esti-mate. The decline in oil prices has already con-tributed to the decline in food prices via its in-fluence on biofuel demand for food crops.Should oil prices remain moderate as pro-jected (see below), the influence of biofuels onfood prices should also stabilize. If technolog-ical progress improves the attractiveness ofnonfood biofuels inputs, the link between oiland food prices may be broken. Alternatively,biofuels could have a significant impact onfood prices if oil prices remain high or the costof biofuels production declines.

The simulation reported in table 2.14 ex-plores the implications of a permanent increasein the rate of growth of demand for food prod-ucts as source material for biofuels. Under this

scenario, global demand is expected to growtwice as fast as it does in the baseline. In thisinstance, agricultural employment increases by0.6 percent of the labor force, output of othergrains (including maize) rises by 350 percent,but food prices increase by much less, due tosubstitution away from these products.

Over the long run oil prices are expectedto stabilize (in real terms) at around $75As described in chapter 1, despite rather recentvolatility, oil prices are not expected to fallmuch below $60 in the medium term. Oil de-mand should pick up as the global economy re-covers, but supply conditions should also haverecovered, enabling the real price of oil to risegradually to around the $75 range. This fore-cast assumes that, in the absence of policychanges, demand for energy will continue torise faster than GDP. The actual rate ofincrease of demand and how it is met willdepend critically on policies, technologicalchange, and the level of reserves.

The simulations presented in table 2.15 il-lustrate the potential impacts of four alterna-tive scenarios.


87

Table 2.15 Energy sector simulation results, 2005–30

Results in 2030 by scenario

I. II. III. IV.High Carbon Alternative Weak oil

2004 Baseline demand tax energy supply

Energy demand (average annual percent growth)Coal 4.9 5.7 2.2 2.5 5.4Oil 1.6 1.8 1.6 1.4 0.2Natural gas (excluding distribution) 1.3 1.8 1.0 �0.2 1.3Total 3.0 3.5 1.7 1.6 2.9

Prices ($ per ton of oil equivalent)Coal 59 60 62 55 54 60Crude oil 256 428 475 420 219 760Natural gas 157 288 306 281 231 296

Production levelCoal (metric tons) 5,680 18,312 21,907 10,184 10,185 19,993Crude oil (mbd) 75 113 117 112 78 78Natural gas (1e12 BTU) 59,435 82,951 93,105 76,020 56,156 84,356

Share in total energy supply (percent)Coal 33.4 53.9 57.2 37.7 42.3 63.1Oil 47.3 33.5 30.5 46.2 45.6 23.8Natural gas 19.3 12.5 12.3 16.1 12.1 13.0

Source: World Bank ENVISAGE model.

10363_Pg51_94:10363_Pg51_94 11/29/08 7:07 AM Page 87

In Scenario I, energy demand rises 0.5 per-cent faster (3.5 percent versus 3.0 percent) thanin the baseline each year because energy-savingtechnologies and conservation measures fail tocome onstream as rapidly as anticipated.32 Thisresults in higher prices for all forms of energy.The higher price of energy means that globalGDP grows somewhat more slowly, with thecumulative impact on the level of output, com-pared with the base case, equal to 2.7 percentin 2030. Most of the increase in demand is con-centrated in the use of coal (in absolute andpercentage terms). Relatively higher supplyelasticities for coal and gas lead to higher vol-ume shifts for these two fuels, whereas thetighter supply of the oil markets leads to a con-comitantly higher price rise for oil and a rela-tive shift away from oil consumption.

Scenario II examines the impact of a moreconcerted effort to limit carbon emissions. Inthis scenario, it is assumed that policies are putinto place beginning in 2011 that are consistentwith achieving a target concentration of 500parts per million of carbon dioxide in the at-mosphere by 2050. This implies a shadowprice of carbon of $21 per ton of CO2 in 2030and a stock of emissions of around 11 gigatonsof carbon in 2030, a reduction of 32 percentfrom the base-case level.33

Such a carbon price would lead to a signif-icant drop in energy demand, with coal takingthe largest hit (from 4.9 percent to 2.2 per-cent). Coal would be most affected because itreleases the most carbon emissions per unit ofequivalent energy. But a more significant fac-tor is the large wedge in the price of coal (perunit of energy) compared with oil and gas. Inother words, the uniform price of carbon hasa much larger percentage increase on the priceof coal than on oil and natural gas. As a corol-lary, the countries with the greatest coal con-sumption experience the largest decline inenergy demand.

Scenario III illustrates a situation where acombination of policies to promote conserva-tion, increase fuel efficiency, and invest in al-ternative sources of energy such as solar andwind power succeeds in reducing the demand

for traditional fossil fuels. In this scenario, theglobal energy demand is lower by about thesame amount as in the carbon tax scenario, butprices of crude oil and natural gas are muchlower. By the end of the period, coal consump-tion is down by 45 percent from the base case(from 4.9 percent to 2.5 percent), while nat-ural gas and crude oil are 30 percent lowerthan they are in the carbon tax scenario.

The price of various forms of energy in thelong run is little different from the baselinescenario because the additional carbon tax in-duces sufficient reductions in energy demandto lower the final price by almost as much asthe tax itself.

Finally, under Scenario IV, oil reserves de-plete more quickly than in the baseline scenario,either because current estimates of reservesprove too optimistic or because additional tech-nology improvements do not materialize. In thisscenario, oil supply, instead of growing at about1 percent a year, is broadly stable, with pro-duction of about 78 mb/d in 2030. Oil pricesare about 80 percent higher and demand 32percent lower than in the baseline, with thedifference being made up by about a 9 percentstronger growth in consumption of coal.

The price of oil in this scenario rises toabout $122 a barrel but not higher because ofincreased supply from alternative energysources induced by the higher prices.

Overall, the impact on global growth inScenario IV would be limited. By 2030 globalGDP would be only 1.4 percentage pointsbelow the level in the base case. The bulk ofthis decline would be felt by the middle-income developing countries, where energyintensities are highest.

Taken together, these scenarios illustratethe considerable uncertainty surrounding theassumptions of the base case. Nevertheless,even the pessimistic scenarios have a limitedimpact on global welfare. Over the long run,economies have considerable potential to ad-just to higher oil prices through switching toother energy sources and conservation, thusmoderating the impact of higher oil prices ongrowth and poverty reduction.


88

10363_Pg51_94:10363_Pg51_94 11/29/08 7:07 AM Page 88

Conclusions

The almost unprecedented duration andsize of the recent commodity price boom

gave the impression, at least as can be judgedby the popular press, that the world is runningout of natural resources. This is not true. Acombination of circumstances have shaped thisboom: an unusually long period of above-potential growth among developing countries;a long period of low oil and metals prices thateroded supply capacity, in part driven by theexpansion of net oil exports from the transi-tion economies of Eastern Europe once domes-tic prices increased to world levels; the depreci-ation of the dollar; the increase in subsidies forbiofuels that diverted resources from growingcrops for food; declines in grain stocks; increas-ing demand from developing-country consumersof oil and raw materials; and continued globaleconomic expansion in the face of rising com-modity prices. As the rapid decline of commod-ity prices since mid-2008 attests, the currentboom is best understood as yet another cycle ina long history of commodity price cycles.

This does not mean that commodity pricesare necessarily going to fall all the way back tothe levels of the 1990s, nor are they likely toreturn to recent heights when demand recov-ers. In the oil and metals markets, it will taketime to build the machines and train the engi-neers required to find and exploit new re-sources, and this kind of exploration will re-quire that oil prices be maintained at around$75 a barrel in real terms.

However, in the long run, it will be difficultto sustain very high oil prices (in excess of$100 a barrel) for a lengthy period, because al-ternative sources of oil (such as Canadian oilsands and more-expensive offshore sources)and substitutes for oil (such as solar, wind, andbiofuels) would become profitable, while thepotential for reductions in demand from con-servation remain large. On average, the weak-ening of global demand and increased supplyhave caused metal prices to fall by more than40 percent from their recent peaks. Neverthe-less, they remain 2.5 times higher than theywere in the 1990s and even though they are

projected to decline a further 20-odd percent(see chapter 1) in 2009, these prices are highenough to ensure that sufficient further supplywill be forthcoming over the medium term. Inthe longer run, metals demand should slow asChinese metal intensities first stabilize andthen fall, both because of lower investmentrates and because of a higher share of servicesin Chinese GDP.

In agriculture, slower population growthshould slow demand for food, while productiv-ity growth should be sufficient to ensure futuresupply at the global level. However, prospectsfor individual countries are less clear. Yieldshave been declining among many of the coun-tries that had the strongest gains from theGreen Revolution, unless they step up invest-ments in infrastructure and R&D and remainopen to new technologies, agricultural pro-ductivity growth in developing countries maydecline. Moreover, for those countries withrelatively high population growth, many ofwhich are in Africa, failure to make invest-ments to boost agricultural productivity maysee them cease to be self-sufficient and forcedto import increasingly expensive food fromhigh-income countries where agricultural pro-ductivity continues to rise much faster thanthe population.

Central to these forecasts, and particularlyuncertain, are the prospects for technologicalprogress. Technology will determine the avail-ability of oil reserves and the costs of extrac-tion, the price levels at which different oil sub-stitutes become profitable, the potential foreconomizing on scarce oil and metals, and thelikelihood of rapid increases in crop yields.Making assumptions for technologicalprogress 25 years in the future is a perilous un-dertaking. Most likely to be missed are tech-nological surprises that enable rapid increasesin productivity. So in a sense these forecasts areconservative. But even without counting ontechnological miracles, under reasonable as-sumptions the supply of commodities is likelyto increase rapidly enough over the long run tomeet anticipated increases in demand at pricesthat are lower than the current levels.


89

10363_Pg51_94:10363_Pg51_94 11/29/08 7:07 AM Page 89

Notes1. The 1916–17 boom was associated with the

First World War. Similarly, all three booms since 1945have been associated with a major, though geographi-cally confined, military conflict (Korea, Vietnam, andIraq) and heightened geopolitical uncertainty, whichtranslated into market fears about the availability ofsupplies.

2. However, real prices of domestic food commodi-ties in developing countries increased by an additional28 percent during the first three quarters of 2008.

3. This capacity was partly and temporarily utilizedduring the time of the first Gulf war, when 5 mb/d ofcapacity was shut in Iraq and Kuwait.

4. OPEC surplus capacity typically refers to capac-ity that can be brought onstream within 90 days. Here,OPEC’s surplus is conservatively estimated as thatlying dormant from previously higher (though notpeak) levels.

5. Although less important than in the past, thesefirms still account for almost 50 percent of global up-stream spending.

6. Upstream expenditures and the price of crude oilare highly correlated; the correlation coefficient betweenspending per barrel of oil and the price of oil is 0.95.

7. The balance of supply was made up from OPECnatural gas liquids (1.8 mb/d); non-OPEC, non-FSUproduction growth (3.1 mb/d); and rising OPEC capacity.

8. The pickup in oil demand was led by China,where demand for electricity had outstripped supplyfrom public sector utilities, resulting in a spike in theprivate use of diesel oil for electrical generators.

9. Private communication with David Humphreys,chief economist at Norilsk Nickel.

10. The contrast between inputs in metal and in-puts in the agricultural sectors is noteworthy. In agri-culture, the same type of machinery can be used for vir-tually all crops in all countries of the world. However,machinery in metals is custom-made for each mine.

11. Jatropha curcus L. is a bush or small tree usedas a hedge by farmers in developing countries becauseit is not browsed by animals. It produces a fruit withhigh oil content, suitable for biodiesel production.

12. The poor harvests in Australia come against abackdrop of an unprecedented, decade-long period ofunusually low rainfall and record-high temperatures,which are at least partly a result of climate change.These events have severely stressed water supplies inthe east and southwest of the country (http://www.bom.gov.au/climate/drought/drought.shtml).

13. When hoarding and real-side speculation occurin response to expectations of a future shortfall, stocks(and prices) tend to increase in the short run (relative toa baseline where the behavior did not occur). In turn,

this ensures that future stocks are higher and futureprices lower than they would have been otherwise. Atthe same time it encourages producers to increase out-put, thereby accelerating a return to more normal prices.

14. In member countries of the Organisation forEconomic Co-operation and Development, high pricesinduced a substantial switch away from oil and towardcoal, natural gas, and nuclear power for electricalgeneration.

15. Exceptions include nickel, which has been ris-ing in Brazil; copper, which has been rising in India;and aluminum, which has been rising in South Africa.

16. Although developing countries now account foralmost half of the world’s metal consumption, it shouldbe noted that their average per capita use of metals isonly a fraction of that in the developed economies.

17. These income levels correspond roughly to themidpoint in the World Bank’s official range forlower-middle-income countries and close to theupper range for upper-middle-income countries.

18. At incomes of less than $1 a day (or annual percapita income of less than $350), consumption of basicstaples such as maize, wheat, and rice tends to increasealong with income. At higher incomes, per capitaconsumption of staples tends to remain stable, so thegrowth of staples consumption falls below incomegrowth.

19. The income elasticity for meat products exceeds3.0 for per capita incomes below $4,500 and declinesto 2.6 for countries with incomes in excess of $25,000.

20. When oil costs $120 a barrel (as it did in early2008), wholesale gas prices would be around $3.25 agallon in the United States, and the fuel-equivalentethanol price would be $2.44 a gallon. At that price,ethanol production from maize is profitable as long asmaize prices do not exceed $245 a ton, which wasmore or less the price of maize at that time.

21. Global production from onshore sources in2004 was 54 mb/day, almost identical to the 1973 level.

22. Reserve growth refers to the increase in the es-timated sizes of fields that occurs as oil and gas fieldsare developed. In the United States, the world’s mostintensely explored country, reserve growth is a majorcomponent of remaining oil and gas resources. It is hy-pothesized that reserve growth can occur worldwide insimilar proportions as exploration of new fields ma-tures. Undiscovered resources, on the other hand, areresources postulated from geologic information andtheory to exist outside of known oil and gas fields(Kleitt and others 2000).

23. If all announced projects materialize, potentialcapacity could reach 3.3 mb/d.

24. Concerns over the environment are likely to beanother constraint in future mining activities.


90

10363_Pg51_94:10363_Pg51_94 11/29/08 7:07 AM Page 90

25. Like electric cars, hydrogen-powered vehiclesallow the consumption of the power and the consump-tion of the propellant to be geographically separated.For electric cars, the energy source (be it coal, nuclear,or solar) that powers the car is consumed at the power-producing plant, whereas for hydrogen-powered cars(as distinct from hydrogen fuel-cell cars), it is expendedin the plant that separates the hydrogen from water.Thus a hydrogen-powered car can be considered justanother form of battery-powered car.

26. If OPEC is considered as a single producer, thenboth oil-export and oil-reserve markets are highlyconcentrated (Herfindahl index of 0.53 and to 0.57,respectively). However, if the member countrieswere to act independently, the market would not beparticularly concentrated (Herfindahl index of 0.07and 0.09, respectively).

27. Specifically, average caloric consumption in de-veloping countries rose from 2,110 kilocalories perperson per day to 2,650 (FAO 2006, p. 3).

28. Much of the underutilized land in the formerSoviet bloc was cultivated in the Soviet era but was leftto fallow when price signals rather than command andcontrol began to determine land use decisions.

29. The potential gains in the table reflect estimatesof the increase in production that could be expected iffertilizer production in countries in each of these re-gions were brought up to the 75th percentile level—roughly the level in Pakistan. In contrast with the cal-culations of Coelli and Rao (2005), which yieldbroadly similar results, these control for climatic con-ditions, income per capita, and soil conditions.

30. Investment expenditures in extractive industriesare highly correlated with the respective prices. Forexample, when crude oil prices declined by 50 percentfrom 1980 to 2000, investment expenditures followedsuit.

31. These estimates do not include the carbon fer-tilization effect (on which scientific evidence is mixed),whereby increases in atmospheric carbon concentra-tion enhance plant growth The simulations may beoverestimating the negative impacts. The Cline esti-mates have been scaled assuming linearity, but someevidence suggests that the actual damage functions arenonlinear. These simulations may therefore overesti-mate damages in the short run and underestimate themin the long run.

32. The baseline scenario incorporates an increasein efficiency of energy use of 1 percent a year. This rep-resents the culmination of new scientific advances (forexample, the use of carbon fiber materials instead ofmetals); the replacement of old, more-energy-intensivecapital with new capital; and changes in behavior(for example, switching from large vehicles to smallerones).

33. The questions regarding who should bear theburden on reducing carbon emissions are criticallyimportant but are set aside in this simulation to investi-gate the impact on overall demand and prices. The rev-enues generated by the price of carbon are assumed tobe recycled domestically with no international transfers.

ReferencesBohi, D. R. 1999. “Technological Improvements in

Petroleum Exploration and Development.” In Pro-ductivity in Natural Resource Industries: Im-provements through Innovation, ed. R. D. Simp-son. Washington, DC: Resources for the Future.

Büyüksahin, Bahattin, Michael S. Haigh, and MichelA. Robe. 2008. “Commodities and Equities: ‘AMarket of One’?” U.S. Commodity Futures Trad-ing Commission, Washington, DC.

Cline, William R. 2007. Global Warming and Agricul-ture: Impact Estimates by Country. Washington,DC: Center for Global Development and PetersonInstitute for International Economics.

Collins, Keith. 2008. “The Role of Biofuels and OtherFactors in Increasing Farm and Food Prices: AReview of Recent Development with a Focus onFeed Grain Markets and Market Prospects.”Kraft Foods Global, Inc., June 19.

Coelli, Tim J., and D. S. Prasada Rao. 2005. “TotalFactor Productivity Growth in Agriculture: AMalmquist Index Analysis of 93 Countries,1980–2000.” International Association of Agri-cultural Economists 32 (s1): 115–34.

Connor, Mike. 2008. “Could the EPA Cripple theChevy Volt?” http://blogs.motortrend.com/6293345/government/could-the-epa-cripple-the-chevy-volt/index.html (accessed October 14, 2008).

FAO (Food and Agriculture Organization). 2006.World’s Agriculture towards 2030/2050. InterimReport. Rome: FAO.

———. 2008. Current World Fertilizer Trends andOutlook to 2012. Rome: FAO.

FAPRI (Food and Agriculture Policy Research Insti-tute). 2008. U.S and World Agricultural Outlook.Ames, IA: Iowa State University.

Fargione, Joseph, Jason Hill, David Tilman, StephenPolasky, and Peter Hawthorne. 2008. “LandClearing and the Biofuel Carbon Debt.” Science319: 1235–38.

Gilbert, Christopher L. 2008. “Commodity Specula-tion and Commodity Investments.” Revised ver-sion of paper presented at the conference on TheGlobalization of Primary Commodity Markets,Stockholm, October 22–23, 2007.

Grilli, Enzo R., and Maw Cheng Yang. 1988. “PrimaryCommodity Prices, Manufactured Good Prices,and the Terms of Trade of Developing Countries:


91

10363_Pg51_94:10363_Pg51_94 11/29/08 7:07 AM Page 91

What the Long Run Shows.” World Bank Eco-nomic Review 2: 1–47.

GTZ (Gesellschaft für Technische Zusammenarbeit).2006. “Biofuels for Transportation.” WorldwatchInstitute for the German Federal Ministry ofFood, Agriculture and Consumer Protection(BMELV) in cooperation with the Agency forTechnical Cooperation (GTZ) and the Agency ofRenewable Resources (FNR), Washington, DC.

IEA (International Energy Agency). 2001. Oil SupplySecurity: The Emergency Response Potential ofIEA Countries in 2020. Paris: IEA.

———. 2007. Energy Use in the New Millennium:Trends in IEA Countries. Paris: OECD/IEA.

———. 2008a. Energy Technology Perspectives:Scenarios & Strategies to 2050. Paris: OECD/IEA.

———. 2008b. Medium Term Oil Market Report.Paris: IEA.

IMF (International Monetary Fund). 2006. World Eco-nomic Outlook. Washington, DC: IMF.

———. 2008. World Economic Outlook. Washington,DC: IMF.

ITF (Interagency Task Force on Commodity Markets).2008. Interim Report on Crude Oil. Washington,DC: ITF.

Kaldor, Nicholas. 1934. “A Classificatory Note on theDetermination of Equilibrium.” Review of Eco-nomic Studies 1: 122–36.

Kleitt, T. R., James W. Schmoker, Ronald R. Charpen-tier, Thomas S. Ahlbrandt, and Gregory F.Ulmishek. 2000. “Glossary” in U.S. GeologicalSurvey: World Petroleum Assessment 2000.Washington, DC: USGS.

Kojima, M., and T. Johnson. 2005. Potential for Bio-fuels for Transport in Developing Countries.Washington, DC: World Bank.

Kojima, Masami, Donald Mitchell, and William Ward.2006. Considering Trade Policies for Liquid Bio-fuels. Washington, DC: World Bank.

Lee, J. J., S. P. Lukachko, I. A. Waitz, and A. Schafer.2001. “Historical and Future Trends in AircraftPerformance, Cost and Emissions.” AnnualReview Energy Environment 26: 167–200.

Lin, Hian, Nan Zhou, Mark D. Levine, and David Fri-dley. 2006. “Achieving China’s Target for EnergyIntensity Reduction in 2010: An Exploration ofRecent Trends and Possible Future Scenarios.”Ernesto Orlando Lawrence Berkeley NationalLaboratory, Berkeley, California.

Lipsky, John. 2008. “Commodity Prices and GlobalInflation.” Remarks at the Council on ForeignRelations, New York City, May 8. http://www.imf.org/external/np/speeches/2008/050808.htm (ac-cessed October 14, 2008).

Lobell, David B., Marshall B. Burke, Claudia Tebaldi,Michael D. Mastrandrea, Walter P. Falcon, andRosamond L. Naylor. 2008. “Prioritizing ClimateChange Adaptation Needs for Food Security in2030.” Science February 319: 607–10.

Lowell, J. D. 1970. “Copper Resources in the 1970s.”Mining Engineering (April).

Martin, Will, and Devashish Mitra. 2001. “Productiv-ity Growth and Convergence in Agriculture ver-sus Manufacturing.” Economic Development andCultural Change 49: 403–22.

Masters, Michael W. 2008. “Testimony before theCommittee of Homeland Security and Govern-ment Affairs.” United States Senate, Washington,DC, May 20.

Mitchell, Donald. 2008. “A Note on Rising FoodPrices.” Policy Research Working Paper 4682.World Bank, Washington, DC.

Mitchell, Donald, Shuo Tan, and Hans Timmer. 2007.“China’s Demand for Primary Commodities andthe Impacts on Global Markets.” World Bank,Washington, DC.

Normila Mary Anne, Anne B. W. Effland, and C.Edwin Young. 2004. “U.S. and EU Farm Policy—How Similar.” In U.S.-EU Agricultural Compar-isons. Agriculture and Trade Reports. Washing-ton, DC: U. S. Department of Agriculture.

OECD (Organisation for Economic Co-operation andDevelopment). 2007. “Biofuels for Transport:Policies and Possibilities.” Policy Brief. Novem-ber. OECD, Paris.

Padgett, Marty. 2008. “2011 Chevy Volt Creates NewClass of Green Vehicle: E-REV.” http://blogs.thecarconnection.com/blogs/marty_blog/2011/chevy-volt-creates-new-class-of-green-vehicle-e-rev/ (accessed September 21, 2008).

Pardey, Phillip G., Julian M. Alston, and Jennifer S.Jones. 2008. “Agricultural R&D Policy: ATragedy of the International Commons.” WorldBank, Washington, DC.

Plastina, Alejandro. 2008. “Speculation and CottonPrices.” Cotton: Review of the World Situation61 (4): 8–12.

Radetzki, Marian. 2008a. A Handbook of PrimaryCommodities in the Global Economy. Cam-bridge, U.K.: Cambridge University Press.

———. 2008b. “Prospects for Long Run Metal andMineral Supply.” Background paper. World Bank,Washington, DC.

Rosegrant, Mark W., Tingju Zhu, Siwa Msangi, andTimothy Sulser. 2008. “The Impact of BiofuelProduction on World Cereal Prices.” InternationalFood Policy Research Institute, Washington,DC.


92

10363_Pg51_94:10363_Pg51_94 11/29/08 7:07 AM Page 92

Sandrea, Ivan, and Rafael Sandrea. 2007. “GrowthExpected in Global Offshore Crude Oil Supply.”Oil & Gas Journal 105: 34–36.

Searchinger, Timothy, Ralph Heimlich, R. A. Houghton,Fengxia Dong, Amani Elobeid, Jacinto Fabiosa,Simla Tokgoz, Dermot Hayes, Tun-Hsiang Yu.2008. “Use of U.S. Croplands for Biofuels In-creases Greenhouse Gases through Emissions fromLand Use Change.” Science 319: 1238–40.

Soros, George. 2008. “Testimony before the U.S. Sen-ate Commerce Committee Oversight Hearing onFTC Advanced Rulemaking on Oil MarketManipulation.” Washington, DC, June 4.

Thompson, Robert. 2008. “Farming—the Essential ofLife: A Global View of the Future.” NationalFarmer’s Union Conference, February 18. http://www.nfuonline.com/x25294.xml.

UN (United Nations). 2006. World PopulationProspects: The 2006 Revision. New York: UN.

UNCTAD (United Nations Conference on Trade andDevelopment). 2007. World Investment Report.Geneva: UNCTAD.

USDA (U.S. Department of Agriculture). 2008. Pro-duction, Supply and Distribution. Online data-base, accessed August 23, 2008.

World Bank. 2007a. Global Economic Prospects:Managing the Next Wave of Globalization.Washington, DC: World Bank.

———. 2007b. World Development Report 2008:Agriculture for Development. Washington, DC:World Bank.

———. 2008. Global Economic Prospects: Technol-ogy Diffusion in the Developing World. Washing-ton, DC: World Bank.


93

10363_Pg51_94:10363_Pg51_94 11/29/08 7:07 AM Page 93

10363_Pg51_94:10363_Pg51_94 11/29/08 7:07 AM Page 94

95

Dealing with ChangingCommodity Prices

3

As discussed in chapter 2, the rise in pri-mary commodity prices between 2003

and mid-2008 was much larger and more sus-tained than those of earlier decades. Althoughcommodity prices have fallen sharply fromtheir recent highs, they remain well above theirlevels in the early 2000s and are projected toremain high relative to their levels in the 1990sfor a significant period of time.

The boom in commodity prices has gener-ated dramatic transfers of income within andamong countries. While high commodityprices have imposed a severe burden on manyconsumers, they have also created significantopportunities for producers. The short-termmacroeconomic, balance of payment, infla-tionary, and growth implications of thesehigher prices were discussed in chapter 1,while long-term prospects for commoditymarkets were discussed in chapter 2.

This chapter focuses on the challengesthat prolonged periods of high and then lowcommodity prices pose for developing coun-tries. In particular, it evaluates the policiesadopted by both commodity-producingand -consuming countries during this boom,as well as the potential role of the interna-tional community in managing the commodityprice boom to maximize the development im-pact and protect the most vulnerable.

The main messages arising from this analy-sis are:

Commodity dependence need not hurt long-term growth. Indeed, high commodity pricesprovide a development opportunity but onlyif the proceeds are not squandered and if theright policies are adopted.

• Although commodity-dependent econo-mies have, on average, grown more slowlythan more diversified economies, for mosteconomies dependence on commodities isthe result of slow growth, not the cause.Several countries have achieved rapid de-velopment based on the exploitation ofnatural resources.

• To achieve the growth potentially inherentin commodity riches, countries need to im-plement policies that minimize the poten-tial disruptive effects of volatile export rev-enues, exchange rate appreciation that canerode the competitiveness of manufactur-ing, and incentives for rent seeking andcorruption.

Higher food prices, while damaging to urbanconsumers, may help lower poverty in thelong run.

• Higher agricultural prices provide addi-tional income in the rural economy, wheremore than 75 percent of the world’s poorlive. Some of this income will go directly to

10363_Pg95_140:10363_Pg95_140 11/29/08 7:08 AM Page 95

farmers, potentially helping them move be-yond precarious forms of subsistence agri-culture. Another part will go to raise in-comes of farm workers and increasedemand for related services such as trans-portation, inputs, and processing.

• For these potential gains to be realized gov-ernment will need to pursue policies thatinvest in infrastructure, including roadsand marketing institutions to move farmproducts to markets and inputs to farmers.

Resource-dependent developing countrieshave done a better job than in the past ofmanaging the macroeconomic consequencesof rapidly rising foreign currency earnings.

• Government spending in most countrieshas responded more prudently to increasedcommodity revenues than in the past. In-stead of spending temporary windfall re-serves, many governments have accumu-lated foreign reserves, and created andaugmented sovereign wealth funds. As a re-sult, real effective exchange rates in mostresource-rich countries have appreciated byless than in the past. Finally, resource-dependent countries are less corrupt andmore transparent when compared withmore diversified economies than in thepast.

• As a result, the nonresource sectors of thesecountries are more likely to have avoided alarge deterioration in international compet-itiveness, and a strong procyclical cut inspending is less likely to accompany the re-cent decline in commodity prices. Improve-ments in governance may also have con-tributed to these developments and haveincreased the chances that revenues arebeing allocated toward projects that en-hance the long-term development potentialof countries.

• Although in aggregate the story is encour-aging, some countries are experiencingstrong inflationary pressures that may re-duce their competitiveness and the sustain-

ability of growth. Others that lack a longhistory of oil or mineral development havepursued less prudent policies that may havesewn the seeds of future difficulties.

High food and oil prices may have increasedthe number of people living in extremepoverty by between 130 and 150 million.

• High food and fuel prices have impliedenormous transfers in incomes betweenproducers and consumers. High fuel priceshave reduced real incomes in oil-importingdeveloping countries by some $162 billiondollars but increased them by some $400billion in oil exporters. With the exceptionof a few import-dependent countries, foodis mainly consumed in the same countrywhere it is produced. As a result, the redis-tributive impact of high food prices ismainly between domestic producers andamounted to some $277 billion betweenJanuary 2007 and August 2008.

• Within countries, the largest poverty im-pacts have been among urban populations,which have not benefited from increasedearnings to the same degree as the ruralpopulation. Impacts were also larger incountries with fewer domestic alternativesto internationally traded grains, whoseprices rose the most (maize, wheat, andrice).

To mitigate the poverty impacts of higher foodprices in a fiscally responsible way, countriesneed to respond with targeted measures. Therecord so far is mixed at best.

• Strict targeting of assistance programs is es-sential to reach those most affected whilelimiting the strain on fiscal accounts. Thecosts of fully compensating people in devel-oping countries for higher food and fuelprices would be prohibitive both to coun-tries and to the aid community. Costs rangebetween 6 and 27 percent of the GDP of in-dividual countries.


96

10363_Pg95_140:10363_Pg95_140 11/29/08 7:08 AM Page 96

• Many policies imposed by countries so far(lower taxes, export restrictions, and pricesubsidies) have been costly and have im-peded adjustment. Increased fiscal outlayshave exceeded 2 percent of GDP in manycountries. Moreover, policies designed tokeep domestic prices low have exacerbatedand prolonged high market prices by re-ducing incentives to increase productionand reduce consumption.

• Countries should seek to expand or createmore-targeted safety net programs. Foodsubsidy programs, fuel subsides, and taxexemptions tend to be regressive, withmost of the benefits accruing to the non-poor. In contrast, well-targeted schemes,involving some form of means testing orselection mechanisms such as geographictargeting or a work requirement, are mostsuccessful in reducing costs and concentrat-ing benefits among the poor.

Some modest steps have been taken, but theinternational community can do much moreto mitigate the impact of high prices and re-duce the likelihood of further spikes and newcommodity booms.

• Given the magnitude of the problem, inter-national efforts to assist the poor need tofocus on the most vulnerable. One ap-proach would be to direct aid to assistingthe extreme poor in IDA-eligible countries(countries whose poverty and lack of accessto market-based finance make them eligiblefor concessional lending and grants fromthe World Bank Group). The cost of com-pensating the poor in these countries forthe rise in food prices between January2005 and December 2007 would be about$2.4 billion.

• International agreement is needed to placemore effective restrictions on the use of ex-port bans, which have become too com-mon. These bans have increased globalfood price volatility and reduced confi-dence in the reliability of world food mar-

kets, with potentially long-term impacts onfood policies.

• Efforts to improve information about andcoordination of global grain stocks couldreduce the probability of another food cri-sis. Similarly, the effectiveness of humani-tarian aid would be enhanced if the WorldFood Programme (WFP) were providedwith a stable source of financing and a lineof credit that would allow it to respondrapidly to emergencies.

• Biofuels policies that subsidize production,impose high tariffs, and mandate consump-tion need to be reconsidered in light of theirimpact on food prices and their trade-distorting effects. Such policies have led torapid expansion of biofuels productionfrom food crops, such as maize and veg-etable oils, and have contributed to higherfood prices as well as to environmentaldegradation. These policies have also re-duced opportunities for lower-cost devel-oping-country producers to expand pro-duction and exports.

• A successful conclusion to the World TradeOrganization’s Doha Round will not re-duce food prices in the near term, but itdoes offer the prospect of greater disciplinein agriculture and more-rapid incomegrowth in developing countries.

The remainder of this chapter is organized asfollows. The next section considers the perspec-tive of commodity-producing countries, evalu-ating the extent to which their policies have suc-ceeded in coping with volatility fromcommodity prices, thus avoiding some of thepitfalls that have typically caused such countriesto grow less quickly than resource-dependentcountries. The following sections examine theboom from the perspective of consumers, focus-ing on the impact of high prices on the poor andthe effectiveness of the antipoverty measuresimposed and their impact on long-term adjust-ment. The chapter then considers the interna-tional response to the rise in food prices and setsout some concluding remarks.

D E A L I N G W I T H C H A N G I N G C O M M O D I T Y P R I C E S

97

10363_Pg95_140:10363_Pg95_140 11/29/08 7:08 AM Page 97

Commodity dependence andgrowth

Economic dependence on primary com-modities has been long associated with

slow growth in development.1 While commod-ity booms are often associated with a pickup ingrowth, countries heavily dependent on the ex-ports of commodities have slower growth overthe long term than those with more diversifiedexports (the so-called resource curse). This sec-tion argues that this relationship should not beinterpreted as causal and is, in fact, far frominevitable. Provided the right policies areadopted, the resource-rich developing coun-tries have much to benefit from a period ofhigh commodity prices.

The idea that there exists a resource cursederives from the observation that countriesdependent on primary commodities for theirexport revenues have tended, on average, togrow more slowly than more-diversified ex-porters (figure 3.1). Developing countries,which in 1980 derived more than 70 percentof their export revenues from nonfuel primarycommodities, increased their per capita GDPby only 0.4 percent a year between 1980 and

2006, and countries that mainly exportedfuels raised their per capita GDP by 1.1 per-cent a year (figure 3.2). By contrast, more-diversified exporters achieved per capitagrowth of 1.6 percent a year. The same rela-tionship holds if countries severely affected byconflict are excluded, although the nonfuelprimary commodity exporters fare somewhatbetter in this case.

Moreover, low-income countries tend tobe more dependent on nonfuel commodityexports than high-income countries (see fig-ure 3.2). More than 60 percent of the exportsof low-income countries derives from nonfuelcommodities compared with about 33 percentfor high-income countries.

Resource dependency reflects low GDP,not resource wealthHowever, resource dependence is not the sameas resource richness. Most countries that areresource dependent (measured as the share ofnon-oil primary commodities in exports) actu-ally have relatively poor resource endowments(measured as per capita income derived fromnon-oil primary commodities). Conversely,many countries that are rich in resources havelow resource dependencies because, in addi-tion to having ample resources and large


98

Figure 3.2 Poorer countries are moredependent on nonfuel primary commodities

Low-incomecountries

Source: World Bank.

Upper-middle-income countries

Lower-middle-income countries

Non

fuel

com

mod

ities

Fue

ls

Oth

er

% share of merchandise export revenues, 2006

0

10

20

30

40

50

60

70

Figure 3.1 More-diversified developingcountries grew more rapidly from1980 to 2006

All countries

Source: World Bank.

Note: Diversified exporters include countries that depend onfuel and nonfuel primary commodities as well exports ofmanufactures.

Excluding conflict countries

Nonfuelprimarycommodityexporters

Fuelexporters


Average growth rate, %

0.0

0.2

0.6

0.4

0.8

1.0

1.2

1.4

1.6

1.8

10363_Pg95_140:10363_Pg95_140 11/29/08 7:08 AM Page 98

resource sectors, they also have thriving in-dustrial and service sectors. Oil-exportingcountries are excluded from this comparisonbecause most of them are both resource richand resource dependent.

Resource dependency primarily reflects lowlevels of GDP, not resource richness. Whilethe top 20 non-oil resource-dependent coun-tries have an average annual per capita incomeof just $1,099, the annual income of the top20 resource-rich countries is 11 times higher(table 3.1). These trends are reflected morebroadly. Even when oil exporters are includedin the mix, low-income countries have thehighest dependence on primary commodities,but the lowest level of primary commodity ex-ports per capita, and the inverse is true forrich countries (figure 3.3).

Considerable efforts have been made to de-termine if, after controlling for other determi-

nants of growth, dependence on primary com-modities is associated with slower growth.Several authors have found a negative rela-tionship in cross-section regressions betweennatural resource abundance and growth.2

Others find that natural resource abundance isnot responsible for the slow growth of re-source-rich developing countries (Manzanoand Rigobon 2007), and that there is a posi-tive relationship between resource abundanceand both short-term (Collier and Goderis2007) and long-term growth (Lederman andMaloney 2007) after accounting for othergrowth determinants.

Commodity dependence may, but neednot, result in slower growthWhile the causality behind these correlationsremains unresolved in the literature, there isconsensus about the channels through whichcommodity dependence could contribute toweaker growth. These include:

• A tendency for significant fluctuations inexport revenues, often exacerbated by


99

0

10

20

30

40

50

60

70

0

10

20

30

40

50

60

70

Primary exports per capita(left axis)

Primary exports/exports(right axis)

Value of per capita primarycommodities in exports(US$ thousands)

Source: World Bank.

Share of primarycommodities in total

merchandise exports (%)

Figure 3.3 On average, poor countries aredependent on commodities but relativelyresource poor

Low-incomecountries

Lower-middle-incomecountries

Upper-middle-incomecountries

High-incomecountries

Table 3.1 Non-oil or resource-richcountries have higher per capita incomesthan resource-dependent countries, 2006

Share of Net nonfuelnonfuel primaryprimary commodity

Real GDP commodities exportsper capita in exports per capita

(US$) (percent) (US$)

Top countries dependent on non-oil primary commodities

1 Gambia, The 320 97 �812 Uganda 275 91 173 Cuba — 85 494 Ethiopia 146 84 65 Niger 168 83 36 Malawi 145 82 247 Jamaica 3,357 81 2768 Rwanda 262 80 �49 Chile 5,896 79 2,596

10 Burundi 102 79 �4

Top countries rich in non-oil primary commodities

1 New Zealand 15,199 62 2,5972 Chile 5,896 79 2,5963 Australia 23,262 48 2,3894 Netherlands 25,678 16 1,4475 Norway 41,446 14 1,4366 Ireland 30,736 10 1,2657 Denmark 32,484 23 1,1428 Canada 25,894 17 1,0829 Estonia 6,938 26 675

10 Kazakhstan 2,166 28 533

Source: World Bank.Note: — � Not available.

10363_Pg95_140:10363_Pg95_140 11/29/08 7:08 AM Page 99


100

As discussed in chapter 2, at the national level therevenues from commodities tend to be much

more volatile year to year than at the global level,and they are more volatile than manufactures. As aresult, countries for whom primary commodities rep-resent a large share of exports experience higher lev-els of GDP volatility than countries with more diver-sified exports.a Indeed, export revenues, the realexchange rate, and per capita output were all morevolatile over the past 25 years among those develop-ing countries where primary commodity exportsrepresented more than 70 percent of total exports(box figure).b

High volatility in these annual data reflectspronounced economic cycles that can have adverseimplications for growth and development.c Sharpbooms and busts can lead to unemployment and un-derutilized capital during downswings and to bottle-necks during upswings. High levels of uncertaintyconcerning future prices and demand can depress

Box 3.1 The impact of severe shocks on economicprogress

the average level of investment over the cycle. Higherrisks may bias lenders toward shorter maturities,further raising the risks of investment. And volatilityof consumption reduces welfare directly if mostconsumers are risk averse.

For countries with the same level of primary com-modity dependence, less-developed economies tendto be more sensitive to such swings because they lackthe means of coping with volatility. In countries withmore-developed financial systems, individuals canborrow to smooth consumption over the cycle, firmscan borrow to sustain operations in bad times, andgovernments can run countercyclical fiscal policy toreduce the macroeconomic implications of adverseshocks. By contrast, in less-developed countries withunderdeveloped domestic financial systems and weakaccess to international finance, these adjustmentmechanisms tend to function poorly. As a result, theimpact of volatility on long-term growth and welfareis more severe.

Moreover, poor households suffer most from ad-verse shocks, because they tend to have lower levelsof savings, have limited access to credit (and interestrates from informal lenders tend to be high), andmust therefore respond to negative shocks by cuttinginto already low levels of consumption. In addition,if workers lose labor experience and connections andchildren leave school, these permanent losses inhuman capital may increase long-term poverty(Ocampo 2003).

Whether month-to-month or day-to-day volatilityhas similarly deleterious economic impacts is lessclear. High-frequency volatility tends to increasetransaction costs and reduce activity levels, but it isless likely to cause the kind of cycles in investmentbehavior and economic activity described above.Moreover, high-frequency volatility is easier to over-come through traditional financing mechanisms,such as short-term credit and inventory adjustments.

An illustration of the difference between eco-nomic cycles and measured volatility based on morefrequent data is provided by the recent boom incommodity prices. While this was the longest andlargest commodity price boom in the past 100 years(see chapter 2), price volatility, as measured bychanges in monthly data, increased only modestly

Box figure 3.1 The impact of severe shockson economic progress

Export revenues GDP per capitaReal exchangerate

Nonfuel primary exporters


Standard deviation of percentage change

Economies dependent on primary commodities experiencemore volatility

0

5

10

15

20

25

30

35

40

Source: World Bank.

Note: Volatility is defined as the standard deviation of percentagechanges over time (annual data). Commodity concentrationmeasured in 1980. Excludes countries with population ofless than 1 million.

Fuel exporters

10363_Pg95_140:10363_Pg95_140 11/29/08 7:08 AM Page 100


101

a. See Turnovsky and Chattopadhyay (1998) and Van der Ploegand Poelhekke (2007) among many others. Cashin, Cespedes,and Sahay (2002) show that volatile commodity prices in-creased the volatility of real exchange rates for 58 countriesover 1980–2002.

b. The more diversified exporters include countries that dependon both fuel and nonfuel primary commodities, as well as ex-porters of manufactures.

c. In cross-country regressions, Aghion and others (2005) findthat real exchange rate volatility lowered growth performancein developing countries over 1960–2000. Fatas and Mihov(2005) find that variability in inflation and government spend-ing were related to lower growth in a cross-section of 91 coun-tries. Aizenman and Marion (1996) find a negative relationshipbetween volatility and private (but not total) investment, andBleaney (1996) and Ramey and Ramey (1995) find a negativerelationship between volatility and growth but not betweenvolatility and investment. Empirically, there is a relatively ro-bust negative relationship between high volatility of growthrates and the level of development (Koren and Tenreyro 2003).However, the direction of causation is unclear. Rather than sug-gesting that volatility causes underdevelopment, the greater de-pendence of poorer countries on relatively volatile primarycommodities may explain the correlation.

procyclical government spending, to accen-tuate economic cycles, tending to depressgrowth over the medium term (box 3.1);

• A tendency for exchange rate appreciationsassociated with commodity booms toweaken the competitiveness of the non-commodity sectors of the economy (theso-called Dutch disease); and

• A tendency for high commodity revenuesto incite individuals to attempt to appro-priate the wealth generated by the resourcewithout investing in productivity or value-enhancing activities (rent-seeking behavior)or, in the worst cases, to engage in outrightcorruption.

Of course, abundant commodity wealth, ora large rise in the value of commodities stem-ming from higher prices, can also contributeto a country’s development, if the implied in-come generated is fruitfully invested—for ex-ample, in infrastructure, education, and healthor in additional productive capacity when the

rents accrue to the private sector. Althoughmore easily said than done, when governmentcontrols the resource rents, care must be exer-cised to avoid forcing the economy down anartificial capital-intensive path instead ofusing the commodity rents to exploit the econ-omy’s comparative advantage, which could bebased on a combination of commodities, com-modity-intensive sectors, and labor-intensiveservices.

What determines whether resource wealthgenerates wider development is the extent towhich the proceeds are consumed (appropri-ate for a permanent increase in income) orsaved (appropriate for a temporary increase);whether they are invested in high- or low-return enterprises; the extent to which rentsaccrue to the population at large or arechanneled through the government; andwhether they are deployed responsibly andtransparently by governments, or used tofund a bloated civil service or are even stolenoutright.

Price volatility has not increasedsystematically

Average absolute monthly percent price change

Crude oil Copper Aluminum Coal

2000–03 8.4 3.4 3.1 4.02004–07 6.9 6.2 4.6 5.72008 7.6 6.3 6.5 15.0

Wheat Corn Rice

2000–06 4.5 5.0 2.92007 7.9 6.1 1.82008 9.5 9.4 18.3

Source: World Bank.Note: Volatility is defined as the average of the absolute valueof the month-to-month percentage change in detrended prices.

until 2008. Indeed, only some of the commoditiesthat have experienced a sharp rise in price experi-enced greater volatility during the price rise thanthey did previously (box table). Volatility didincrease for almost all of the principal commoditiesin 2008, reflecting the rise in prices earlier in theyear and their subsequent decline.

10363_Pg95_140:10363_Pg95_140 11/29/08 7:08 AM Page 101

Overall, an abundance of natural resourcesdoes not necessarily impair development andcan in fact promote it, but it does presentparticular challenges that require appropriatepolicies to overcome.

Managing primary commoditybooms

While dependence on primary commodi-ties does not condemn a country to slow

growth, it does require careful management ofmacroeconomic policy to reduce the impact ofvolatile export revenues (see box 3.1).

In past decades, the governments of sev-eral developing countries failed to reactappropriately to commodity price booms,increasing public expenditures on inefficient,import-intensive investment projects (Cashin,Cespedes, and Sahay 2002) and borrowingexcessively—expecting export revenues to re-main high for longer than was the case.3 As aresult, many of them faced severe economicdifficulties when prices declined. For example,the seeds for the Latin American debt crisis ofthe 1980s were sown by the accumulation ofdebts by countries during a period of highcommodity prices. The payments for theseloans proved to be unsustainable when inter-est rates rose and commodity prices declined,resulting in years of slow growth or economicstagnation (Manzano and Rigobon 2007).

Commodity revenuesand fiscal spendingThe tendency for a temporary rise in revenuesto be reflected in an unsustainable rise in gov-ernment spending has historically been an im-portant explanation for the poor long-termgrowth performance of commodity-dependentdeveloping countries. Countries that are depen-dent on point resources—oil and metals—areparticularly vulnerable because the governmentis the direct recipient of a large share of boomrevenues, either through ownership of the re-source or through taxing the rents accruing toa limited number of private firms. By contrast,

government revenues are less sensitive tobooms in agriculture prices because agricul-tural export crops are produced in manylocations by many producers, so productionexpands to the point where, in normal times,there are no rents for governments to appropri-ate and no special tax regimes (Collier 2007).

Although the evidence is not conclusive,the tendency for government spending to risewith windfall revenues, while still present dur-ing the current commodity boom, is less pro-nounced than in the past. This in turn suggeststhat perhaps the strong growth that has beenassociated with higher commodity prices thistime may prove more sustained than in pastbooms.4

Resource-rich developing countries haveshown greater fiscal restraint during thecurrent boomDuring this boom, resource-rich developingcountries appear to have shown greater fiscalrestraint than they did during earlier booms,thereby reducing the risk of a procyclical cutin government spending now that commodityprices are declining.5 The average general gov-ernment budget surplus of oil-exporting coun-tries improved from 0.6 percent of GDP in2001 to 7.7 percent in 2007. Among develop-ing-country exporters of oil, minerals, andagricultural products, public consumption hasincreased more slowly than private consump-tion, external debt has risen more slowly thanduring past booms, and government borrow-ing has increased more slowly than privateborrowing (IMF 2008b).

While fiscal policy responses have been ex-tremely diverse,6 government expenditures ofprimary commodity exporters have increasedless strongly than during the 1980s, a periodlike the current boom when the export rev-enues of resource-dependent developing coun-tries increased by about 7 percent of GDP(figure 3.4).7

In the 1980s, government spendingtended to increase procyclically—rising in linewith the boom in GDP caused by windfallcommodity revenues. As a result, the ratio of


102

10363_Pg95_140:10363_Pg95_140 11/29/08 7:08 AM Page 102

government expenditure to GDP was broadlystable. On a cyclically adjusted basis, however,government spending rose. Because much ofthe additional money went to governmentspending and transfer programs of a quasi-permanent nature, the increased spendingproved hard to reverse when GDP slowed andcommodity prices reversed. Governmentswere either obliged to cut spending procycli-cally as commodity prices fell, which exacer-bated the cycle, or allow the deficit and debtto build up, increasing their macroeconomicvulnerability

Most recently, governments have reactedmuch more prudently. As a consequence,while government expenditure has increasedin real terms, it has declined as a share of GDPby almost 5 percentage points. Governmentexpenditure among nonfuel exporters has de-clined the most, perhaps reflecting concernthat nonfuel commodity prices would remainhigh only temporarily and the tendency forgovernments to absorb a smaller share ofwindfall revenues from high prices for nonfuelcommodities than from those for hydrocarbonresources.8 Fuels (and minerals) exporters

have also taken steps to increase the share ofthe windfall revenues that accrue to the state,although care must be taken to avoid harmingincentives for production (box 3.2).

Much of the difference between the twoperiods reflects more prudent behavior bygovernments in Sub-Saharan Africa. Duringthe 1980s boom, government expenditures incountries dependent on primary commoditiesin Sub-Saharan Africa rose even more quicklythan GDP. In this decade, the ratio of govern-ment expenditures to GDP has declined by al-most 8 percentage points (figure 3.5). Thistrend contrasts with the spending pattern inLatin America and the Caribbean and theMiddle East and North Africa (other regionshave too few observations to report useful


103

Figure 3.4 Government spending by primarycommodity exporters responded less toexport booms in this decade than in the1980s

Percentage change of the share of GDP

26

24

22

0

2

4

6

8

1980s 2000s

Source: World Bank.

Change inexports/GDP

Change in governmentexpenditures/GDP

Note: The country sample includes developing countries whereprimary commodities account for more than 70 percent ofmerchandise exports. The figures represent the percentage pointchange in merchandise exports divided by GDP, and governmentexpenditures divided by GDP, during the boom.

Figure 3.5 Public expenditures in Sub-Saharan Africa grew much less quickly in the2000s than in the 1980s


a. 1980s

25

0

5

10

15

20

25

Latin Americaand Caribbean

Middle East andNorth Africa

Sub-SaharanAfrica

Latin Americaand Caribbean

Middle East andNorth Africa

Sub-SaharanAfrica


b. 2000s

210

25

0

5

10

15

Source: World Bank.

Source: World Bank.

Change in exports/GDP

Change in government expenditures/GDP

10363_Pg95_140:10363_Pg95_140 11/29/08 7:08 AM Page 103

averages), where government spending hasbeen more procyclical—rising at about thesame rate as GDP as during the 1980s.

Surprisingly the extent to whichgovernments are saving from increased oilrevenues is only loosely correlated withthe size of their reservesFor countries dependent on nonrenewable re-sources, the optimal fiscal response to primarycommodity price booms in part depends onthe importance and expected life span of theresource.9 Some countries, such as República


104

As commodity prices increased, a number ofcountries sought to increase the share of the

windfall that accrues to the state. Several energyproducers (including Argentina, Bolivia, Colombia,Ecuador, and República Bolivariana de Venezuela)have increased, or are considering increases in, therates for royalties or taxes. A few countries haveforced the renegotiation of contracts or nationalizedexploitation rights, which has had a chilling effect oninvestors’ willingness to participate in some markets.Developed-country governments (for example,Alaska in the United States and Alberta in Canada)also are increasing their revenue share.

The governments of several metal-producingcountries also have attempted to increase their shareof the rising profits in recent years (UNCTAD 2006).For example, Mongolia instituted increased rightsfor the government to acquire equities in newventures. The Democratic Republic of Congo isreviewing contracts for mineral extraction signedsince 1995 with the purpose of increasing thegovernment’s stake. Governments, including Chile,Mongolia, Peru, South Africa, and Zambia, havetaken steps or are considering proposals to raisemineral taxes or royalty fees.

Countries that contract with private (often inter-national) firms to exploit nonrenewable resourceshave revised contracts to reflect higher prices. Thedanger here is that arbitrary changes in their share of

Box 3.2 Efforts to capture a larger share of windfallcommodity revenues

revenues will reduce the companies’ incentive to in-vest and lower confidence in the broader investmentclimate. An alternative approach, which is now beingconsidered by several countries, is to base the gov-ernment’s revenue share on the price. For example,Colombia has proposed imposing an additional5 percent tax on every $30 increase in the price of abarrel of oil, thereby raising the tax rate to 75 per-cent when oil exceeds $140 a barrel. This kind ofarrangement holds some promise of creating a stableframework so that firms can evaluate investmentsaccurately and governments can capture a fair shareof windfall revenues when price increases.

It is understandable that countries wish to capturea rising share of revenues from nonrenewable re-sources as prices increase. However, such effortsneed to be carefully calibrated to maintain appropri-ate incentives for making new investments and maxi-mizing current output. Countries with state-ownedcompanies that control resource extraction have toensure that incentives facing these companies encour-age efficiency. For example, whereas some state-owned energy firms (for example Brazil’s Petrobras)continue to enjoy very positive relations with service-providing firms and are efficiently managed, others(such as in Mexico and República Bolivariana deVenezuela) face very high effective tax rates thathave resulted in chronic underinvestment, decliningoutput, and poor efficiency.

Bolivariana de Venezuela, could continue toproduce oil at current rates until almost theend of this century before exhausting all ofthe oil deposits detected under their soil(table 3.2). However, other countries that areheavily dependent on deposits of oil or mineralresources could exhaust their reserves (as cur-rently estimated) within one or two decades.10

If resources are viewed as a national assetof both current and future generations, thencountries with low reserves should be savinga much larger proportion of permanent (andwindfall) revenues—investing them in either

10363_Pg95_140:10363_Pg95_140 11/29/08 7:08 AM Page 104

productive potential or financial assets thatwill continue to generate an income even asthe original resource is depleted.11 To a de-gree, this is what countries are doing. Theshare of government spending in total GDPamong countries with low reserves has de-clined, whereas those with high reserves havebeen more procyclical (figure 3.6).12

Countries like Algeria, Angola, the Repub-lic of Congo, Turkmenistan, and the Republicof Yemen, all of which have less than 20 yearsworth of reserves and rely upon hydrocarbonexports for 80 or more of their merchandiseexports, face serious challenges. Unless theirsavings from oil revenues are high, associatedexpenditures are likely to lead to exchangerate appreciation, with serious negative im-pacts on the non-oil sectors of their economies(see below).

Private sector saving fromcommodity revenuesWhile governments appear to be saving moreof the windfall than they did in the 1980s,private sector spending is rising rapidly—especially among non-oil primary commodity

exporters. However, much of the demand isgoing to investment goods. Investment de-mand in commodity-dependent economies in-creased 7.5 percentage points faster duringthis boom than during the 1980s. As a result,the current private sector boom should beincreasing domestic productive capacity thatwill help countries sustain the high growth ofthe past several years.

Reflecting the large share of commodityrevenues that accrue to the government in oil-exporting countries and the relative prudencethat these governments have displayed, im-ports in these countries have increased lessrapidly than GDP, and current account sur-pluses have improved significantly as a shareof GDP during the recent oil price rise. Thispattern is similar to, but more pronouncedthan, that prevailing during the 1980s boom(figure 3.7).

In part because the benefits of high agricul-tural prices accrue to a much wider segmentof the population, the private sector in non-oil-commodity exporters appears to have in-creased spending sharply during the recentboom, with much of the increased demand


105

Table 3.2 Ratios of reserves to productionvary greatly among oil exporters(Percent)

Share of oil in Ratio of oil reservesCountries merchandise exports to current production

Algeria 95.7 16.8Angola 92.0 17.6Azerbaijan 85.1 29.3Equatorial Guinea 83.8 13.8Gabon 71.1 25.3Iran, Islamic Rep. of 89.8 86.7Iraq 88.1 157.6Kazakhstan 52.8 76.5Libya 98.7 61.9Nigeria 95.6 40.3Oman 85.2 20.5Congo, Rep. of 92.1 19.9Sudan 74.8 44.2Syria, Arab Rep. of 58.3 19.7Turkmenistan 81.0 9.2Venezuela, R. B. de 80.5 77.6Yemen, Rep. of 80.9 20.0

Source: World Bank, British Petroleum.

Percentage point change of the ratio

25

0

5

10

15

20

High reserves Low reserves

Figure 3.6 Oil-exporting countries with largereserves spent a smaller portion of theirrevenue from the recent boom in oil prices,2000–06

Source: OPEC Secretariat, World Oil, Oil and Gas Journal,World Bank staff calculations.

Note: Includes countries where oil accounts for more than 70percent of merchandise export revenues and data on oil reserves,oil production, and government expenditures are available(Angola, Republic of Congo, Equatorial Guinea, Islamic Republicof Iran, Kazakhstan, Libya, Nigeria, Oman, Syrian Arab Republic,Turkmenistan, República Bolivariana de Venezuela, and Republicof Yemen).

Governmentexpenditures/GDP

Exports/GDP

10363_Pg95_140:10363_Pg95_140 11/29/08 7:08 AM Page 105

having been met through imports. The ratioof imports to GDP increased by 6 percentagepoints, and the current account balance hasremained roughly stable despite a 23 percentrise in export revenues.

Real currency appreciationThe rapid increase in imports and the stabilityof the current account in the face of rising ex-port revenues and domestic demand is poten-tially disturbing, because it suggests that thedomestic supply response in these countrieshas been relatively weak. This situation is es-pecially problematic if the increased importsare consumption goods, and if they are associ-ated with a real effective appreciation of thecurrency that has impaired the competitive-

ness of the noncommodity sectors of the econ-omy. To the extent that the imports reflect in-vestment, they are less worrisome if they arecreating the future productive potential thatwill allow these countries to continue growingstrongly when commodity prices and incomesweaken.

Most resource-rich countries are showingfewer signs of real effective exchangerate appreciationThe relationship between export revenues andthe exchange rate is complex. While a real ex-change rate appreciation is the appropriate re-sponse to a long-term improvement in theterms of trade, it may have a deleterious im-pact on the economy if the appreciationproves short-lived. Potential negative effectsinclude adjustment costs, such as increasedunemployment or the bankrupting of mar-ginal firms, and reductions in potential posi-tive externalities in tradable goods sectors,such as

• More-rapid technological progress throughlearning by doing in industries character-ized by firm-specific knowledge

• Demonstration effects, where the gains inefficiency of one firm are easily copied byothers

• Increased incentives for accumulation ofhuman capital

• More-stable and faster-growing markets inmanufactures than primary commodities13

During the most recent boom, there is someevidence that developing countries have suc-ceeded in limiting the appreciation of their cur-rencies, thus reducing potential adjustmentcosts as prices decline (figure 3.8). On average,the currencies of non-oil primary commodityexporters have actually depreciated by a modest4 percent in real effective terms, while the cur-rencies of developing-country oil exporters haveappreciated only 8 percent in real effectiveterms—although most recently domestic infla-tion has risen to more than 10 percent inAngola, the Islamic Republic of Iran, República


106

a. Oil exporters

25

210

215

220

0

1980s 2000s

5

10

15

20

25

30

0

21

22

1

1980s 2000s

2

3

4

5

6

7


Source: World Bank.

Figure 3.7 Imports and current accountpositions suggest more savings from commodity revenues by oil exporters than by nonfuel commodity exporters

b. Non-fuel exporters


Imports/GDP Current account/GDP

10363_Pg95_140:10363_Pg95_140 11/29/08 7:08 AM Page 106

Bolivariana de Venezuela, and the Republic ofYemen (see chapter 1 for a discussion of infla-tion and commodity prices) (figure 3.9).14

The limited currency appreciation in re-sponse to the commodity price boom is in partattributable to the fiscal restraint discussedearlier. Government expenditures fall mostheavily on nontraded goods. As a result, in-creasing government expenditures tend toraise the price of nontraded goods relative totraded goods, which causes the real exchangerate to appreciate.

Commodity-dependent countries alsoavoided real appreciations by sterilizing theinflows of foreign currency by converting

them into foreign-denominated assets. Oil-exporting developing countries doubled theirofficial foreign reserves from $36 billion in2000 to $70 billion by mid-2008, or fromabout four months of import cover to aroundeight months in 2008. At the same time, someof these countries created new sovereignwealth funds (Algeria, Kazakhstan, and Libya)or greatly expanded preexisting sovereignwealth funds (Azerbaijan, Russian Federation,and República Bolivariana de Venezuela)(Griffith-Jones and Ocampo 2008). The assetsof developing-country exporters of oil andminerals in such funds reached $367 billion bymid-2008 (table 3.3).

New entrants into oil production may beexceptions to these welcome trendsSeveral resource-rich developing countries areenjoying the fruits of newly found naturalwealth or are experiencing their first com-modity boom as an independent state, notablythe oil-producing countries of central Asiathat were formerly part of the Soviet Union.These countries have less experience in


107

Figure 3.8 Primary commodity exporterslimited the real appreciation of theircurrencies during the recent boom

Percentage change in trade-weighted real effectiveexchange rate

220

215

210

25

0

5

10

Non-oil exporters Oil exporters

Source: IMF data. World Bank staff calculations.

1980s boom

Recent boom

Chad

Congo

, Rep

. of

Algeria

Gabon

Nigeria

Oman

Libya

Sudan

Yemen

, Rep

ublic

of

Kazak

hsta

n

Angola

Azerb

aijan Ira

n

Venez

uela,

R. B. d

e

25

210

0

5

20

15

10

Figure 3.9 Many oil exporters are suffering significantly higher inflation


Annual % increase in CPI, 2007

Table 3.3 Assets in sovereign wealthfunds grow in commodity-exportingcountries($US billions)

Country As of mid-2008

Algeria 47.0Azerbaijan 5.0Botswana 6.9Chile 15.5Equatorial Guinea 2.9Iran, Islamic Republic of 12.9Kazakhstan 21.5Libya 50.0Mexico 5.0Nigeria 11.0Russian Federation 162.5Timor-Leste 3.0Trinidad and Tobago 2.0Venezuela, R. B. de 22.0Total 367.2

Source: Sovereign Wealth Fund Institute(www.swfinstitute.org).Note: Latest available information as of June 2008, but allestimates may not refer to 2008. Excludes funds with assetsunder $1 billion. Data for Equatorial Guinea as of 2005.

10363_Pg95_140:10363_Pg95_140 11/29/08 7:08 AM Page 107

managing a resource boom than countries thathave been producing substantial amounts ofoil for many years.

Perhaps because of this lack of experience,many of these countries show signs of experi-encing the same kind of macroeconomicvolatility that characterized developing, re-source-rich countries in the 1980s. Their cur-rencies have appreciated in real terms (againstthe U.S. dollar) by 43 percent from 2001 to2007, their inflation rates are higher, and gov-ernment expenditures have been rising in linewith GDP (figure 3.10).

While these developments may be consistentwith prudent management of newfound wealthand a careful investment strategy designed toenhance future production capacity, they mir-ror, disconcertingly, those of the 1980s amongmore established producers. New producers

must therefore pay particular attention tomacroeconomic management going forward toensure that the current downturn in primarycommodity prices does not lead to a sharp re-versal of economic progress.

Another troublesome aspect of the currentboom, especially given the financial crisis, is therapid increase in bank lending to commodity-dependent economies in Sub-Saharan Africa,in part to finance investments in oil and min-eral projects. Despite enjoying substantial in-creases in their export revenues, many of theseeconomies remain poor and need to be partic-ularly careful in incurring foreign currencyliabilities on market terms. Commercial bankcommitments to these economies rose from anaverage of just under $2 billion a year in1995–2000 to more than $5 billion a year in2004–06, and to $11 billion in 2007 (fig-ure 3.11). These countries’ total stock of private-source external debt has not increased signifi-cantly above the $35 billion level reached in2000 and has fallen as a share of GDP. Thedownturn in commodity prices could result indisappointing returns to these projects anddifficulties in servicing this debt on the part offirms, especially as existing loans come due inthe current environment of much tighter creditconditions and higher risk premiums fordeveloping countries. Should companies have


108

Figure 3.10 New oil exporters are experienc-ing more macroeconomic volatility thanestablished producers

% change

210

0

10

20

30

40

50

60

Real exchangerate with US$

Percentagechange inCPI, 2008

Change ingovernmentexpenditure/

GDP, 2001–07

Source: World Bank and IMF data.

Note: New producers are defined as countries dependent on oilthat began production after 1985 or were established as a countryafter 1985, including Azerbaijan, Chad, Equatorial Guinea,Kazakhstan, Sudan, and the Republic of Yemen (Turkmenistanlacks data for inflation and the real exchange rate). Theestablished producers include Algeria, Angola, Republic of Congo,Gabon, Islamic Republic of Iran, Libya, Nigeria, Oman, andRepública Bolivariana de Venezuela. We use the real exchangerate with the United States (rather than the trade-weighted realexchange rate as in figure 3.5), to include sufficient countries fora useful comparison between the two groups.

a. Real exchange rate with the U.S. dollar, where increaseindicates appreciation. Data for Equatorial Guinea are for2001–04.

b. Percentage change in consumer price index in 2008.

c. Change in ratio of government expenditure to GDP from2001 to 2007.

New producers

Established producers

4,000

2,000

0

6,000

12,000

10,000

8,000

Figure 3.11 Commercial bank lending to commodity-dependent economies in Sub-Saharan Africa is rising

Source: Loanware 2008.

Gross commitments, US$ millions

1995

–200

0,

aver

age 20

0120

0220

0320

0420

0520

0620

07

10363_Pg95_140:10363_Pg95_140 11/29/08 7:08 AM Page 108

difficulty refinancing, this could transfer into asovereign risk—–especially in those caseswhere the debtor firms are state-owned.

Governance and transparencyResource riches can yield disappointinggrowth outcomes by creating incentives andopportunities for corruption, mismanage-ment, and political instability. Resourcewealth has been a source of political conflictsin Africa (Gelb 1998) that have been enor-mously destructive of wealth, while in coun-tries with weak governance and institutions,the concentrated wealth deriving from pointresources too often lends itself to corruptpractices by politicians and civil servantscharged with overseeing the firms exploitingthem.15 Indeed, some econometric analyseshave found that dependence on oil, metals,and minerals, where the government plays acentral role in determining the allocation ofrents, lowers the quality of institutions.16

Partly reflecting the influence of these in-centives, countries dependent on nonrenew-able resources (equal to more than 70 percentof merchandise exports) tended in 1996 to bemore corrupt than those dependent on agricul-tural commodities and more diversified exports(figure 3.12).17 More recently, corruption levelsin the oil, metals, and mineral exporters havedrawn much closer to the developing-country

average. These are relative rankings and thuscannot indicate absolute improvements in in-dividual countries. Nevertheless, this progressmay reflect the reforms instituted over the past10 years to counter the corrupting influence ofhigh resource rents and may also indicate thatresource wealth is being more effectively de-ployed in promoting the overall developmentof these economies (box 3.3).


109

Figure 3.12 Corruption is highest amongfuel exporters, although the difference hasnarrowed

Index

21.0

20.8

20.6

20.4

20.2

0.0

1996

Bet

ter

Wor

se

2006

Note: The lower the value of the index, the worse the level ofcorruption relative to other countries. Countries are classified asoil or mineral, or agricultural exporters if they earn more than 70percent of merchandise export revenues from these sources.Diversified exporters are all other developing countries.Classification is based on shares in 2000.

Source: Kaufmann and others 2007; World Bank data.

Oil and mineral exporters

Agricultural exporters


Arecent example of efforts to reduce the scope forcorruption in commodity-rich countries is the

Extractive Industries Transparency Initiative.Launched in 2002, it aims to increase the account-ability of governments in resource-rich countriesthrough the publication of company payments andgovernment revenues from oil, gas, and mining. Asof July 2008, 23 countries were in the process ofmeeting the conditions for transparency supportedby the initiative, and 17 of 42 major oil companies

Box 3.3 Combating the corrupting influence of highcommodity revenues

were supporting the initiative.a These developmentscould be strengthened if the home countries of multi-national companies were to require these firms to ac-count more explicitly for the funds they disburse tolocal governments.b

a. See the Transparency International Web site, transparency.org.

b. Statement by Michel Roy, from the French NGO SecoursCatholique, published in a press release from Publish What YouPay (www.publishwhatyoupay.org).

10363_Pg95_140:10363_Pg95_140 11/29/08 7:08 AM Page 109

Sovereign wealth fundsThe increased prevalence of sovereign wealthfunds among resource-rich countries is anotherrecent innovation aimed at increasing the de-velopment impact derived frommineral wealth,both by increasing the returns that countries re-ceive on their savings from resource revenuesand by insulating those savings from procycli-cal spending and corrupt practices.

The success of these funds in managing nat-ural resource revenue and reducing procyclicalspending has been mixed (Asfaha 2007). Ingeneral, countries with sovereign wealth fundshave tended to experience less-procyclical fis-cal policies and less-volatile macroeconomicoutcomes.18 However, the commodity here isunclear. Such funds tend to be most successfulin countries that are already fiscally prudentand are most likely to be established in coun-tries with strong institutions. As such, sover-eign wealth funds are no substitute for strongfiscal institutions (box 3.4).19

Dealing with revenue volatilityThe volatility of commodity prices and outputmeans that revenues also tend to be volatile

(see chapter 2). At the macroeconomic level,this manifests itself as greater GDP, exchangerate, and export volatility (see box 3.1). Forindividual producers, this volatility increasesthe riskiness and quantity of investment, espe-cially in developing countries where financialsystems that could provide temporary financ-ing to bridge shortfalls are underdeveloped.As a result, the overall production potential ofthe sector rises less quickly, which may be re-flected in poor growth outcomes. Perhapsmore importantly, for the poor who aredependent on farm-related incomes (close to75 percent of all poor; see below) and livingclose to the subsistence level, the impacts canbe particularly devastating.

Traditionally, developing (and developed)countries have sought to offset this kind ofvolatility with price stabilization schemes,marketing boards, and the like (box 3.5).However, the track record of these schemeshas not been good and they have fallen intodisfavor. More recently, countries are enteringinto more market-based mechanisms such aslong-term contracting arrangements andmarket-based conditional contracts.


110

For a sovereign wealth fund to be successful, trans-parent procedures must be established for manag-

ing the allocation of resources to the fund and the in-vestment of these resources. For example, clear rulesfor forecasting prices (necessary for the calculationof permanent income that underlies allocation deci-sions) and, where available, reliance on independentforecasts can help insulate allocation decisions frompolitical pressures. National revenue funds in Nor-way and Botswana benefited from stable and democ-ratic political systems that encouraged decision mak-ing based on long-term considerations (Eifert, Gelb,and Tallroth 2002).

Rules for the allocation of a share of resource rev-enues to a wealth fund must not be too rigid. Severalcountries have changed, bypassed, or eliminated such

Box 3.4 Successful sovereign wealth fundsrules when conflicts arose (IMF 2007). Such changes,although often needed, can limit the impact of thefund if they occur too frequently as has happened inOman (UNCTAD 2006).

Transparency in the procedures governing thefund must be matched by overall strong governanceto ensure that fiscal policy is consistent with the allo-cation of resources to the fund. For example, insome instances governments have effectively circum-vented the goals of a sovereign wealth fund by bor-rowing (using the fund as collateral). In RepúblicaBolivariana de Venezuela, for example, resourceswere deposited in the national revenue fund accord-ing to the rules, but at the same time the governmentborrowed heavily to finance procyclical expenditures(Fasano 2000).

10363_Pg95_140:10363_Pg95_140 11/29/08 7:08 AM Page 110

Long-term contracting provides large-scaleproducers with some protection fromoutput volatilityOver the past decade or so, a number of re-source-dependent developing countries haveentered into long-term contracts with clientcountries that guarantee sales volumes and insome cases prices. These contracts cover anextended period, sometimes with specific esca-lator clauses that ensure that prices, whilemore stable than market prices, do not varytoo far from market norms, causing one part-ner or the other to renege on the deal.

Russia and oil-producing countries in Eu-rope and Central Asia have engaged in suchcontracts with Hungary, the Czech Republic,Poland, and Ukraine as well as with severalhigh-income countries. Because these con-tracts specify prices over the duration of thecontract, these consuming countries have notobserved as large a swing in energy costs asother countries (and supplier countries havenot experienced as large a boom).

Such contracts are sometimes entered intoin the context of a foreign direct investmentdeal by either the resource-exporting countryor, increasingly, a resource-importing countryhoping to gain security of future supply.20 Sev-eral African countries have entered into suchrelationships with Brazil, China, India, andMalaysia, among others, in exchange for astable demand-supply relationship and accessto foreign capital (most often in the form offoreign direct investment) to develop domesticresources.

China, or Chinese state-owned firms, havetaken equity positions in oil ventures in Africaequal to some $13.5 billion as of early 2007.Investments have been made in Angola, Chad,Côte d’Ivoire, Equatorial Guinea, Kenya,Mauritania, Niger, Nigeria, São Tomé andPrincipe, Somalia, and Sudan, but the bulkof production is currently concentrated inAngola, Nigeria, and Sudan (Downs 2007).Chinese companies also have invested in thedevelopment of minerals, such as copper and


111

Marketing boards in developing countries typi-cally got their start during colonial times as a

way to facilitate the export of agricultural commodi-ties to Europe and to stabilize prices for food crops.Newly independent governments generally retainedmarketing boards because they provided a conve-nient way for the governments to maintain controlover the distribution of strategic commodities suchas food staples and export crops.

Marketing boards are state-controlled or state-sanctioned entities legally granted control over thepurchase or sale of agricultural commodities (Barrettand Mutambatsere 2008). They flourished in the20th century in both developed and developingcountries but have declined in number under pres-sure for domestic liberalization and internationaltrade rules. Where reforms have been widespreadand successful, marketing boards have vanished orretreated to providing public goods, such as strategic

Box 3.5 National and international marketing strategiesgrain reserves or insurance against extraordinaryprice fluctuations. Where reforms have been lesssuccessful, the weakness of private agriculturalmarketing channels has been revealed by the rollbackof marketing boards, often leading to calls for rein-statement of the powerful boards.

Similar efforts to minimize volatility have beentried at the global level as well. These included theInternational Sugar Agreement of 1954 and interna-tional agreements for tin (1956), coffee (1962),natural rubber (1980), and cocoa (1981). Theseagreements used some combination of supplycontrol, buffer stocks, and export controls to limitprice changes. All of these commodity agreementsbroke down or lapsed in the 1980s and 1990s eitherbecause of their ineffectiveness or because of difficul-ties in coordinating production among members(Gilbert 2005).

10363_Pg95_140:10363_Pg95_140 11/29/08 7:08 AM Page 111

other resources in Zambia, and cobalt and cop-per in the Republic of Congo (Lyman 2005).Chinese companies also have invested in LatinAmerica, with the bulk of this investment re-lated to the production of primary commodi-ties, such as oil in Ecuador (Caspary 2008).

Market-based conditional contracts offerprotection from both price and volumevolatility for large-scale marketparticipantsSome countries are attempting to reduce theimpact of volatile commodity prices throughmarket-based derivative instruments. Unfortu-nately, developing-country producers, andparticularly agricultural small-holders, havelittle access to the market-based risk manage-ment instruments now available, because of alack of knowledge; lack of collateral for mar-gins; the small scale of their operations; andthe complexities of executing, monitoring,and administering hedging transactions.

These hurdles can be overcome through alarge domestic entity that pools price riskfrom many small producers and hedges themin international markets. In Mexico, the gov-ernment organization, ASERCA, does this tohedge price risks for cotton farmers. ThroughASERCA, the government offers farmers thechance to participate in a program to guaran-tee a minimum cotton price for a fixed fee.ASERCA then hedges its price risk by usingthe fee to purchase a “put” option in interna-tional financial markets, which pays if theinternational price of cotton falls below thespecified price. This payoff is in turn paid outto farmers, effectively providing them withmarket-based insurance against the cottonprice falling below the specified minimum thatis demand driven and inexpensive to adminis-ter (Larson, Varangis, and Yabuki 1998).

The over-the-counter market is very activefor oil (over-the-counter risk management in-struments are highly liquid and can extend asfar as seven years in the future) and preciousmetals (contracts are considered competitiveover the three-to-five-year time horizon).Exchange-traded instruments also exist for

highly traded tropical products such as coffeeand cocoa, and for maize, soybeans, andwheat, which are produced and exported bythe United States. However, the over-the-counter market is more limited for the basemetals exported by many developing coun-tries.21 Moreover, many agricultural productsproduced and consumed by developing coun-tries are difficult to hedge efficiently.

In any event, small-holders in developingcountries have little access to these instru-ments. The provision of agricultural risk insur-ance to small-holders also has proven difficult.State-managed insurance schemes have beenlargely ineffective and unsustainable withoutsubsidies to cover premiums. One hopefuldevelopment is the advent of index-basedweather insurance. These schemes, which pro-vide for a different way of underwriting, andtransferring, weather risk to the market, arenow being scaled up by private initiatives inIndia and elsewhere. In addition to the directbenefits these contracts provide to producers,by reducing overall revenue volatility, theyreduce the risk by potential lenders and canimprove farmers’ access to credit.

So far these efforts have been limited tolarge-scale farms. To bring similar benefitsto small-scale producers, more direct govern-ment involvement may be required to ensurethat supply-chain actors, who are the only ac-tors large enough to enter into such contracts,have the incentives to share their benefits withsmall-holders.

Food markets are more complicatedpoliticallyFood markets present a particularly difficultrisk management challenge, because the re-quirements (objectives) of consumers and pro-ducers are often in conflict. Historically, gov-ernment interventions in food markets havehad significant adverse effects on the supplyside, creating strong disincentives for privatesector storage, finance, and trade. All toooften, the ensuing shortfall in private sectorinvestment in these markets—and the corre-spondingly weak development of local and


112

10363_Pg95_140:10363_Pg95_140 11/29/08 7:08 AM Page 112

regional trade—exacerbate the price and sup-ply volatility that the interventions were at-tempting to mitigate in the first place.

More recently, governments have used cus-tomized price and supply risk managementcontracts to help reduce volatility and ensuresecurity of supply in a way that strengthensrather than weakens private sector trade.Trading companies and banks in southernAfrica are now offering contingent purchaseagreements that use “call” options as a basisfor physical supply contracts (box 3.6). Riskmanagement can also be enhanced by more-open borders and private trade, as in the suc-cessful management of flood-induced riceshortages in Bangladesh in 1998.

Poverty impacts of highercommodity prices

While resource-rich countries have facedchallenges in capitalizing on the rise in

commodity prices, poor consumers confrontsevere difficulties in coping with the substan-tial decline in real incomes. The rise in realcommodity prices in developing countries wasmuch less marked than in the United States(see chapter 1); nevertheless, the increases weresubstantial and imply severe consequences forthe poor in developing countries.

The rise in food prices presents the greaterchallenge for the poor, most of whom spendmore than half of their incomes on food. The


113

In 2005–06, southern Africa experienced a severedrought-related food shortage. Affected countries

included Malawi, Mozambique, Zambia, andZimbabwe. Initial estimates suggested that as muchas 2 million metric tons of maize imports might berequired.

The government of Malawi, with assistance fromthe World Bank and the British government, usedcall options from the South Africa Exchange Market(SAFEX) to help cap the cost of managing the foodshortage. The government was concerned about bothhigh price increases and its ability to secure addi-tional grain on world markets. As a result, a cus-tomized call option for 60,000 metric tons of whitemaize with a total value of approximately $17 mil-lion and a premium payment of $1.53 million waswritten. To ensure that the maize was delivered (ifneeded), the contract was written on a deliveredbasis, thus combining the price for white maize onthe SAFEX exchange with the transport costs toMalawi.

In the event, with spot prices rising and the foodshortage growing more severe in November andDecember 2005, the government exercised the calloption, elected physical settlement, and allocated the

Box 3.6 Malawi government hedging of maize priceand supply risks, 2005–08

majority of the maize to humanitarian operations.During the delivery period, spot prices for a metricton of white maize rose $50–$90 above the ceilingprice of the contract following increases in theSAFEX white maize price and transport costs overthe October-January period. The maize purchasedthrough the option contract had a better deliveryperformance than most other procurementprocedures.

Since then the government, facing a projectedmaize surplus, worked with the World Bank to struc-ture contingent export contracts. These were put op-tions structured to ensure foreign markets wouldtake up any surplus grain and provide a price floorin the case that maize prices fell. Although the con-tracts were not taken up, they did demonstrate howcontingent contracting could be used to help managerisk associated with surpluses. In May of 2008, theMalawi government issued a request for proposalsfor a repurchase option, which will be based on atrade finance structure for grain held in the countrycombined with a call option. The objective of thisapproach is to set up a second layer of grain reservesthat operates financially through the private sector(Dana, Gilbert, and Shim 2006; Dana 2008).

10363_Pg95_140:10363_Pg95_140 11/29/08 7:08 AM Page 113

urban poor are most directly affected, bothbecause they consume more commerciallyproduced foods and because they are muchless likely than the rural population to benefitfrom increased revenues from food sales orimproved employment opportunities arisingfrom higher food prices. The poor are less af-fected by rising fuel prices because they spendless of their incomes on fuel; however, highfuel prices are still a burden to the poor, espe-cially those in colder climates.

The remainder of this section explores inmore detail the impacts that higher prices havehad on the poor in developing countries.

Higher oil prices and povertyAs discussed previously, oil price increasessince 2003 pushed up consumer spending inoil-producing developing countries by some$400 billion in 2008, while the annual in-crease in the food bill due to the price in-creases between January 2007 and May2008 was some $240 billion—assuming inboth cases that international prices werefully passed through to consumers. Ofcourse not all of these price increases havebeen passed through. In these cases, the costsare either being borne by governments as in-creased expenditures or by firms in the formof forgone revenues when price increases arecontrolled.

Most estimates suggest that the poverty im-pact of higher oil prices was smaller than theimpact of higher food prices, mainly becausein most developing countries, the poor spendonly about 10 percent of total householdspending on energy, compared with 50 percentfor food (Grosh, del Ninno, and Tesliuc2008). For example, the poorest 20 percent ofBolivians, Malians, and Sri Lankans spendmore than 40 percent of their income on food,but only 3 percent on energy (World Bank2008a). Moreover, when energy costs rise, theextremely poor tend to turn to alternativesources of energy (principally biomass). Evenwhere the poor receive subsidized fuel forcooking, consumption tends to be low, in partbecause they resell it on the black market.22

At the same time, the direct cost of higherenergy prices may well underestimate theirtotal cost. While direct energy consumptionmay be low, higher energy prices increase theprices of energy-intensive goods and servicesconsumed by the poor. For example, surveysof poor communities in China, India, Indone-sia, and the Lao People’s Democratic Republicindicate that households have reacted toenergy-induced increases in the prices of elec-tricity and transportation by reducing lightingand increasing their isolation (UNDP 2007).Moreover, the switch to lower-cost biomassenergy sources carries with it hidden costsin the form of increased indoor pollution, in-creased incidence of respiratory disease, blind-ness, heart disease, and obstetrical problemssuch as stillbirth and low birth weight (IEA2002).

Many efforts to measure the poverty im-pact of higher oil prices have taken an indirectroute because few household expenditure sur-veys have enough detail on the consumptionof petroleum or petroleum products to esti-mate poverty impacts directly.23 Some countrystudies have relied on input-output tablescombined with household surveys, or on com-putable general equilibrium models, to esti-mate the impact of an oil price rise on poverty.The results are mixed, with most studies con-cluding that a 20 percent rise in oil pricescould impose a 1–3 percent reduction in theincomes of poor households (table 3.4).

Global studies of the impact of oil prices onpoverty have first estimated the impact ofhigher fuel prices on GDP and then the impactof lower GDP on poverty. For example,Herrera and others (2005) estimate that a $10increase in the price of a barrel of oil wouldreduce GDP in the short run by about 0.8 per-cent in developing-country oil importers. Theycalculate that poverty rates would increase inthe more severely affected countries by a rangeof 1.4–1.5 percentage points.

A simplistic extrapolation of these results(which are based on an everything-else-equalassumption) to the $110 increase in crudeprices between 2003 and mid-2008, would


114

10363_Pg95_140:10363_Pg95_140 11/29/08 7:08 AM Page 114

lead to the conclusion that the GDP in devel-oping-country oil importers would havedeclined by more than 8 percent and that theincidence of extreme poverty in developingcountries would have increased by some15 percentage points. However, everythingelse was not equal and for most of the periodduring which oil prices were rising, GDP inoil-importing developing countries was ex-panding by more than 6 percent a year (muchfaster than in the past). At least for the initialincreases in oil prices between 2003 and mid-2006, such a simplistic calculation substan-tially overstates the impact of higher oil priceson GDP and poverty.

That said, the most recent oil-price hikesoccurred under very different conditions thanthe initial ones. Global capacity was con-strained, inflation was rising, and the initialcushions that allowed the first oil price hikesto be absorbed were exhausted (see chapter 1).Partly as a consequence, global growth in oil-importing developing countries slowed by1.7 percentage points between 2007 and2008. Although not all of that slowdown canbe attributed to oil prices, if it were, applyingthe poverty elasticities used by Herrera andPang (2006) would lead to a conclusion thatthe most recent hike in oil prices may have in-creased headcount poverty rates by as much as1.7 or 2.0 percentage points.

The rise in the food bill is attributable tohigher pricesThe balance of payments implications of therise in food prices are important for a fewcountries, including some oil or metals ex-porters, a few countries beset by civil conflicts,and several small island states that sell servicesand import most of their needs, includingfood. However, with the exception of a fewfoods such as palm oil and a few countries, in-cluding several island states and some MiddleEastern countries, the bulk of food productsare consumed in the same country where theyare produced.

Nevertheless, the increased food bill facingconsumers has been extremely large, equalingon average about 2.4 percent of gross nationalincome in developing countries, or 8.0 percentof government expenditures. For some coun-tries, the costs rise as high as 20 percent ofgross national income, equal to the total ofgovernment expenditures (figure 3.13).

The magnitude of these costs would makeit impossible for most governments (or the in-ternational community) to completely financethe rise in expenditures on grains required tomaintain consumption at 2006 levels. As a re-sult, the greater part of the adjustment mustbe borne by consumers, while government in-terventions need to focus on programs thatstrictly target the poor.


115

Table 3.4 Country studies suggest that high oil prices have large poverty impacts

Study Country Impact on poor of 20 percent increase in oil price (unless otherwise specified)

Coady and Newhouse (2005) Ghana Poor incomes decline by 3.6 percent

World Bank (2003) Iran Cost of living of rural poor rises by 3.1 percent

ESMAP Report done in 2001a Pakistan Cost of living of poor rises by 1.15 percent

McDonald and van Schoor (2005) South Africa Rural households suffer drop in income of 0.76 percent, versus0.83 percent for urban households, with poor households lessaffected than rich households

Clements, Hong-Sang, and Gupta (2003) Indonesia 25 percent rise in oil prices reduces average real consumption by2.5 percent, with high-income groups slightly more affected thanlow-income groups

ESMAP (2005) Yemen Increasing fuels to import parity (62 percent) increases householdexpenditures by 14.4 percent for poorest decile

Kpodar (2006) Mali Household expenditures of poor rise by 1.8 percent

Source: Kpodar 2006.a. As cited in Kpodar (2006).

10363_Pg95_140:10363_Pg95_140 11/29/08 7:08 AM Page 115


116

Higher food prices and povertyAlthough estimating the direct poverty effectsof high oil prices is difficult, a more direct ap-proach is possible for analyzing the poverty ef-fect of higher food prices, because householdexpenditure surveys tend to provide more de-tail on the consumption of food.

Changes in food prices can affect povertythrough consumption and income channels.On the consumption side, as food prices rise,the cost of a given basket of food increasesand consumer welfare declines. However, forthe segment of the population whose incomedepends directly or indirectly on agriculture(that is, farmers, wage workers in agriculture,and rural landowners), higher food prices rep-resent an increase in income. Thus, for eachhousehold, the net welfare effect of an in-crease in food prices depends on the combina-tion of a loss of purchasing power (consump-tion effect) and, for some households, a gainin income (income effect). At the countrylevel, the poverty effect of higher food pricesdepends on

• The initial incidence and depth of poverty• The proportion of the poor that have little

or no direct income from agriculture, suchas the urban poor

• The importance of food in the budgets ofthe poor

• Households’ ability to substitute betweenfood items

A rise in the price of food relative to othergoods and services tends to raise poverty inthe short term. The recent increase in interna-tionally traded food prices (mostly grains andoilseeds) is estimated to have increasedpoverty in eight of nine developing countriesstudied by Ivanic and Martin (2008). Thisfinding reflects the fact that most of the poorin developing countries (including those inrural areas) are net food buyers, as demon-strated by a number of studies based on de-tailed household surveys (Christiaensen andDemery 2007; Seshan and Umali-Deininger2007; Byerlee, Meyers, and Jayne 2006).

Analyzing the poverty impact of higherfood prices is complicated, however, becausenet sellers are disproportionately poor (Aksoyand Isik-Dikmelik 2008). As a consequence,high food prices can transfer income fromricher to poorer households. Moreover, overthe longer run, higher food prices that boostfarm income may also increase other ruralincomes by boosting employment and wagesamong the landless rural poor. Thus the im-pact of rising food prices on poverty can differsubstantially between urban and rural areas.

Higher food prices increase urban povertyunambiguouslyThe overall impact of higher prices on povertymay be complicated to sort out, but there isbroad consensus that higher food prices in-crease urban poverty, mainly because most ofthe urban poor have no offsetting income ef-fects. The upper panel of table 3.5 reports theestimated effects on urban poverty levels inthe six World Bank regions of a hypothetical10 percent increase in food prices. The esti-mates are calculated using the Bank’s modelfor Global Income Distribution Dynamics(GIDD) (see box 3.7 for a discussion of theassumptions underlying this and other model-ing exercises reported here).24

10

15

5

0Countries

20

25

Figure 3.13 The increased grain bill could exceed 5 percent of GDP in more than20 countries

Source: World Bank.

Estimated change in grain expenditures, % of GNI

10363_Pg95_140:10363_Pg95_140 11/29/08 7:08 AM Page 116

The largest impacts, both in the increasein the proportion of individuals in the urbanpopulation living in absolute poverty (theheadcount poverty rate) and in the extent towhich the average income of the poor fallsbelow the poverty line (the income gap ratio),are observed in East Asia, South Asia, andSub-Saharan Africa and are attributable to theheavy weight that food plays in the householdconsumption basket in these regions and tothe high initial poverty headcounts in these re-gions (see table 3.5). The increase in head-count poverty in Sub-Saharan Africa is some-what lower than in South Asia because foodrepresents a smaller share of the urban poor’soverall budget.25 Low food shares in LatinAmerica and the Caribbean and very low ini-tial poverty levels in Europe and Central Asiamean that the urban poverty effects of higherfood prices in those regions are close to zero.

In a similar exercise, Dessus, Herrera, andDe Hoyos (2008) estimated that the increasein financial resources needed to alleviate


117

Table 3.5 Higher food prices raise poverty more in urban areas than in rural areasEstimated change in poverty from a 10 percent increase in food prices

Initial Change

Poverty headcount Income gap ratio Poverty headcount Income gap ratioRegion (percent) (percent) (percentage point) (percentage point)

Urban populationEast Asia and the Pacific 13.2 20.3 2.9 1.2Europe and Central Asia 2.5 8.7 0.6 2.5Latin America and the Caribbean 3.7 37.6 0.3 0.0Middle East and North Africa 2.7 17.8 0.6 1.1South Asia 32.3 25.0 4.4 1.5Sub-Saharan Africa 34.1 38.1 2.8 0.5


Rural populationEast Asia and the Pacific 31.9 23.2 1.8 0.3Europe and Central Asia 8.2 6.6 0.3 1.0Latin America and the Caribbean 18.6 43.9 �0.2 0.2Middle East and North Africa 15.4 22.9 0.3 0.2South Asia 43.3 24.0 1.7 0.5Sub-Saharan Africa 54.9 41.5 �0.2 �0.3


Source: Computations using data from the World Bank’s GIDD.Note: The poverty line is set at 1.25 international dollars (2005) a day per capita. The ratio of food in total consumption amongthe poor is computed as described in De Hoyos and Lessem 2008. East Asia excludes China. The Middle East comprises onlyJordan, Morocco, and the Republic of Yemen.

urban poverty arising from the recent increasein food prices is less than 1 percent of GDP forthe majority of countries, rising to 3 percentof GDP among those most affected.26 The au-thors find that around 90 percent of the in-crease in costs derives from a reduction in thereal incomes of households that were poor be-fore the price shock and that the rest is attrib-utable to an increase in the number of poorcaused by higher prices.

Higher food prices also tend to raisepoverty in rural areas, but by lessMost households under the extreme povertyline live in rural areas. In 2000, 7 out of every10 poor individuals lived in a household whereagricultural activities represented the mainoccupation of the head, with lower average in-comes among these households being a con-stant pattern across all regions and countries(Bussolo, De Hoyos, and Medvedev 2008).27

The lower panel of table 3.5 reports theeffect on rural poverty of the same uniform

10363_Pg95_140:10363_Pg95_140 11/29/08 7:08 AM Page 117

10 percent increase in food prices. It assumesthat farm-related incomes of rural householdsalso rise by 10 percent. This could be an un-derestimate, because total spending on foodincludes retailing and transportation margins.Assuming that all of the real increase in foodprices was attributable to increased food com-modity prices, then the percentage increase infarmgate prices would have been proportion-

ately larger than that of retail prices (see tech-nical appendix).

In every region, the deterioration in therural poverty indicators is milder than it is forurban poverty, primarily because of the effectof increased prices on the incomes of farmers.Rural poverty actually declines somewhat inLatin America and Sub-Saharan Africa,whereas it increases a fair amount in East Asia


118

The poverty analysis reported in this chapter isbased on microsimulations using the World

Bank’s model for Global Income Distribution Dy-namics (GIDD). The GIDD comprises household-level data for 73 countries covering around 60 per-cent of the developing world population.

In the reported simulations a number of simplify-ing assumptions had to be made.

1. All households within a country face the same in-crease in the real price of food, measured as therise in the price of food deflated by the rise in theaverage price of all nonfood items. Data are takenfrom national consumer price indexes.

2. The income generated by the rise in food prices isredistributed to rural households in proportion totheir agricultural-generated incomes. Informationon the share of rural household income fromagricultural activities is taken from the “RuralIncome Generating Activities” (RIGA), a projectof the Food and Agriculture Organization (FAO)and World Bank based on 17 Living StandardsMeasurement Surveys. This information isextended to the remaining 56 countries in theGIDD by estimating a simple polynomial relation-ship between the share of agricultural-relatedincome and the level of income (at the centilelevel) across the 17 RIGA countries and thenapplying the estimated coefficients to theremaining countries in the GIDD.

3. One issue is whether self-employed workers andwage earners are likely to share in the rise in in-

Box 3.7 Critical assumptions underlying theestimation of the poverty impact of food price increases

come from higher food prices. Because it is notpossible to identify which households are self-employed and which are wage earners, the addi-tional income attributable to high food prices isdistributed equally among them. This approach isequivalent to assuming that all of the income goesto the self-employed (i.e., assuming that agricul-tural wages and employment are constant) andthat all of the agricultural wage earners in a givencentile work for a self-employed farmer from thesame centile.

4. Household-level information on food consump-tion is available for only 21 countries in theGIDD. Engel curves, relating food shares tohousehold per capita income (or consumption)and other household characteristics (see DeHoyos and Lessem 2008) are estimated, and esti-mated parameters plus a randomly drawn residualare used to impute food shares in countries thatdo not report this information.

5. The simulations show the instantaneous impactof the rise in food prices, assuming no substitu-tion or conservation on the part of consumers(or producers).

The technical annex to this chapter reports thesensitivity of the poverty estimates to variation inassumptions made concerning the size of the priceshock and the distribution of resources within boththe rural and urban sectors.

10363_Pg95_140:10363_Pg95_140 11/29/08 7:08 AM Page 118

and South Asia, reflecting the greater impor-tance of nonfarm incomes within the overallincomes of the rural poor in those regions andthe large share of food in consumption (see thesecond column of table 3.6).

The actual extent of food price increasesvaries widely across countriesThe analysis so far has assumed that all foodprices increased by a uniform 10 percent. Infact, observed changes have been very differ-ent. As discussed in chapters 1 and 2, whileprices of internationally traded commoditiesdenominated in U.S. dollars increased by asmuch as 74 percent between January 2005and December 2007, the real increase ob-served in individual developing countries wasmuch smaller. Indeed, among the 73 countriesfor which distinct monthly consumer priceindex and household survey data are avail-able, the majority had real food price increasesof 12 percent or less (figure 3.14).28 Only fourcountries saw real food prices rise by as much

or more than the average increase of real in-ternationally traded food prices. The differ-ence between domestic and internationalprices arises because internationally tradedfoods represent only a small share of totalfood consumption in most developing coun-tries. Moreover, different foods have very dif-ferent weights across developing countries,and many developing countries have policiesthat have prevented local prices from fully re-flecting changes in international prices.

Table 3.7 reports the result of simulationsof the poverty impacts of the observed in-crease in real food prices. Like the earlier sim-ulations, it assumes that the farm incomes inrural households rise in line with the real in-crease in national food prices.29

As with the uniform shock, all regions ex-cept Europe and Central Asia and Latin Amer-ica and the Caribbean experience a significantincrease in the incidence and depth of poverty.At the global level, the headcount ratio in-creases by 1.3 percentage points, representingan additional 130 million individuals fallingbelow the poverty line.30

The largest increases in the absolute num-ber of poor are in Asia and Sub-Saharan


119

Table 3.6 Observed real price shocks andfood shares of consumption vary acrossdeveloping regions(Percent)

Food shareRegion Price Shock among the poor

Rural populationEast Asia and the Pacific 12.4 71.5Europe and Central Asia �0.2 63.4Latin America and

the Caribbean 6.9 51.2Middle East and

North Africa 25.9 64.5South Asia 5.0 65.3Sub-Saharan Africa 9.6 68.0


Urban populationEast Asia and the Pacific 13.8 67.5Europe and Central Asia �0.5 57.9Latin America and

the Caribbean 1.6 44.1Middle East and

North Africa 12.5 57.1South Asia 4.8 64.4Sub-Saharan Africa 4.9 53.0


Source: World Bank.

0

5

10

15

20

40

Percentage of developing countries

Source: World Bank.

Note: Real local currency price increase of internationally tradedfood commodities is shown by vertical line.

220 0 20

Percentage change in real food prices

Distribution of cumulative increases in relative food prices(Local currency unit, January 2005–December 2007)

Figure 3.14 Real food prices in developingcountries rose less than prices of interna-tionally traded foods

10363_Pg95_140:10363_Pg95_140 11/29/08 7:08 AM Page 119

Africa, reflecting the large number of peoplein each of these regions living just above thepoverty line. The share of the urban popula-tion in extreme poverty is estimated to dou-ble from 2.7 to 5.2 percent in the MiddleEast and North Africa and to increase by al-most 50 percent in the East Asia and Pacificregion.

Some caution should be exercised in inter-preting the figure for East Asia because theGIDD data set does not include China, by farthe largest country in the region. As a result,the GIDD model reports an initial poverty

headcount ratio of 24 percent for urban andrural populations combined, a figure substan-tially higher than the 18 percent reported inChen and Ravallion (2008), which includesChina. The impact that this discrepancy hason the global poverty estimates depends onthe difference between the poverty effects ofhigher food prices in China and those effectsin the average East Asian country. In the ab-sence of household-level information forChina, the underlying assumption is that thepoverty impacts there (that is, the change inthe headcount ratio and the income gap ratio)will be equal to the average poverty effects forthe region.

Overall, the rise in food prices increases theglobal poverty deficit (the amount that a per-fectly targeted poverty alleviation programwould need to spend to bring all of those liv-ing on less than $1 a day up to the povertyline) from 8.2 to 13.4 percent of developing-country GDP, or an increase of $37 billion.The income gap ratio (the average differencebetween the incomes of poor people and thepoverty line, expressed as a percent of thepoverty line) rises by much more in urbanthan in rural areas, reflecting increased earn-ings in rural areas when food prices rise. Thedifference is particularly dramatic in East Asiaand the Middle East, where the increase in theincome gap ratio in urban areas is more than4 times larger than it is in rural areas.

The results presented in table 3.7 hide im-portant heterogeneities across countries. In-deed, the increase in the poverty headcountand the deficit resulting from the rise in foodprices is less than one-fifth of a percentagepoint for almost half of the countries ana-lyzed. In around 40 percent of the countriesanalyzed, higher food prices raise the head-count ratio by at least 0.2 percentage point;and in 6 countries, the change in relativeprices reduces the incidence of poverty by atleast 0.2 percentage point. In some countries,the measured impact of higher food priceson poverty is small, or even negative, becausenonfood prices rose more quickly than foodprices during the period in question.31


120

Table 3.7 Poverty effects of the changesin relative food pricesJanuary 2005–December 2007

Initial levels: Change in:

Poverty Income Poverty IncomeRegion headcount gap ratio headcount gap ratio

(percent) (percentage point)

Urban populationEast Asia and

the Pacific 13.2 20.3 6.3 2.7Europe and

Central Asia 2.5 8.7 0.0 0.2Latin America and

the Caribbean 3.7 37.6 0.1 �0.7Middle East and

North Africa 2.7 17.8 2.4 5.7South Asia 32.3 25.0 2.0 0.5Sub-Saharan Africa 34.1 38.1 1.7 0.3


Rural populationEast Asia and

the Pacific 31.9 23.2 4.9 0.7Europe and

Central Asia 8.2 6.6 0.0 0.0Latin America and

the Caribbean 18.6 43.9 0.1 0.1Middle East and

North Africa 15.4 22.9 0.7 0.9South Asia 43.3 24.0 0.8 0.3Sub-Saharan Africa 54.9 41.5 0.3 0.0


Source: World Bank.Note: Computations using data from the GIDD. Poverty lineof 1.25 international 2005 dollars per day. The ratio of foodin total consumption among the poor is computed as de-scribed in De Hoyos and Lessem 2008. East Asia excludesChina. The Middle East comprises Jordan, Morocco, and theRepublic of Yemen.

10363_Pg95_140:10363_Pg95_140 11/29/08 7:08 AM Page 120

Over the long term, higher food prices willraise incomes in the agricultural sectorIn most developing countries, higher foodprices raise the number of poor and lower in-comes of the existing poor in the short term.Over time, however, the impact on povertybecomes less clear. The increased incomes offood sellers will raise incomes in rural areas(where the majority of poor live). The simula-tions summarized here do not reflect the mul-tiplier effects of higher food prices on incomesin the agricultural sector nor any long-termdynamic effects that may arise because agri-culture has strong links to the rest of the econ-omy. These include backward links, whenfarmers purchase inputs such as chemicals,fertilizers, and farm equipment for agricul-ture, and forward links, when agriculturalproduction provides raw materials to foodand fiber processing in the nonfarm sector.

Moreover, increases in agricultural incomesare usually spent on locally produced goodsand services, which generate local employ-ment. In many African countries, for example,on average for every $1 of additional farmincome, an additional $1.47 in net incomeis generated in the wider economy, some ofwhich accrues to the poor (Delgado, Hopkins,and Kelly 1998).

The long-term impacts of higher agricul-tural prices are difficult to measure becausethey are lengthy and complex (World Bank2007). They depend in part on public invest-ments in roads, markets, irrigation, infrastruc-ture, education, and health as well as on in-vestments in the main factors of agriculturalproduction—land, labor, and capital—all ofwhich take a long time to adjust. Over time,increases in agricultural prices relative toother sectors slow migration out of agricultureand increase capital investment, which resultsin increased agricultural output.32

To the extent that agricultural sectors dosustain more rapid growth because of higherfood prices, rural poverty will be reduced, espe-cially where the concentration of land owner-ship is low and labor-intensive technologies areused (Gaiha 1993; Datt and Ravillion 1998).

Dealing with high foodand fuel prices

The priority for governments is to addressthe immediate needs of the poor while

minimizing the impact on already-strainedbudgets. Care must be exercised to do so in away that does not exacerbate the crisis or im-pair the economic adjustment of the economyto higher prices. Given the necessity to respondquickly and the time and cost involved in gath-ering information on the poor, governmentshave tended to respond to the rise in foodprices by increasing resources to existing an-tipoverty programs. While a logical response,in many cases care has not been taken toclearly define the temporary boost in spendingto compensate for a temporary rise in foodprices by announcing, for example, a limitedtime for improved benefits or by tying themexplicitly to food prices to avoid creating anunnecessary, permanent, and unsustainablefiscal burden.

Over the medium term, governments needto put in place more efficient policies for pro-tecting the poor and supporting agriculture, sothat the next crisis can be met without seri-ously impairing incentives for production orramping up wasteful spending. Such policieswould entail better targeted and more efficientsafety nets, along with steps to achieve the po-tential for strong improvements in agriculturalproduction described in chapter 2, includinginvesting in agricultural research and infra-structure, promoting the diffusion of bestpractices, and reducing carbon emissions tominimize the extent of climate change in thelong term.

The immediate response has been policiesdesigned to mitigate the impact of risingfood and fuel pricesThe immediate response of most countries tothe rapid rise in food and fuel prices duringthe course of 2008 has included a mix of mar-ket interventions and the scaling up of existingantipoverty measures. Almost three-quartersof the 80 developing countries surveyed by the


121

10363_Pg95_140:10363_Pg95_140 11/29/08 7:08 AM Page 121

World Bank in March 2008 have taken somepolicy action in response to the rise in foodprices (figure 3.15).

The most common response was reducedtariffs on imports combined with price con-trols or consumer subsidies, followed by bansor restrictions on exports and decisions to addto official grain stocks. Most oil-importingcountries have passed through all or morethan all of the fuel price increases since 2003,but on average oil-exporting countries havepassed through only about one-half of the in-crease (Mati 2008).33 Indeed, as oil prices hitthe $140 range, the fiscal cost of fuel subsidiesbecame very large in some oil-exporting coun-tries and represented a significant challenge tofiscal sustainability. Some 36 countries re-sponded to higher fuel prices by increasingsubsidies and 43 by lowering fuel taxes (IMF2008a). Those countries that have expandedexisting safety net programs have favored cashtransfers and school-feeding systems. Food forwork and food stamps were also popular op-tions (figure 3.16).

Overall, the additional fiscal costs of mea-sures aimed at offsetting higher fuel and foodcosts varies from zero to a maximum of4.8 percent of GDP, with food and fuel pricesubsidies the most costly measures imple-mented (table 3.8). However, individual coun-

try experience varied widely. Indeed, althoughthe majority of countries increased spending—either because preexisting subsidy policiesbecame much more expensive or because ofdirect measures—some actually reduced thescope of programs and cut into spending be-cause of increased budgetary cost.

Policies need to be more targeted andmore supportive of medium-termadjustmentAlthough subsidies and export restrictionshave helped dampen the immediate impactof higher prices in the countries where they areimplemented, they are very expensive andoften poorly targeted. Moreover, they tend toexacerbate the extent and duration of the


122

0

20

10

30

40

50

60

Percent

Figure 3.15 Developing countries have responded to rising food prices with a variety of policies

Source: Revenga, Wodon, and Zaman 2008.

NoneIncreasefoodgrain

stocks

Exportrestrictions

Pricecontrols/

consumersubsidies

Reducefoodgrain

taxes

0

20

10

30

40

50

70

Percent

60

Figure 3.16 Countries have tended to expand cash transfers and school feeding programs when responding to higher food prices

Source: Revenga, Wodon, and Zaman 2008.

NoneFood forwork

Food ration/stamp

Schoolfeeding

Cashtransfer

Table 3.8 Fiscal costs of selectedantipoverty measures vary widely

Number of Maximum Mediancountries increase increase

where (percent of (percent ofMeasure implemented GDP) GDP)

Food tax decreases 31 1.1 0.1Food price subsidies 28 2.7 0.2Targeted transfers 21 2.0 0.2Public sector wage hikes 10 1.9 0.6

Fuel subsidies 38 4.0 0.7Fuel tax reductions 37 1.3 0.3

Aggregate costs 79 4.8 0.7

Source: IMF 2008a.

10363_Pg95_140:10363_Pg95_140 11/29/08 7:08 AM Page 122

crisis by reducing producers’ incentives to in-crease output and consumers’ incentives to re-duce demand. Over the medium-term, policymakers need to redress the balance, placingmore emphasis on well-targeted antipovertymeasures and on policies that promote in-creased supply and more prudent use of nat-ural resources.

Subsidies and price floors are expensiveand poorly targeted antipoverty measuresFood and fuel subsidies tend to be costly andpoorly targeted, even when steps are taken tomake the subsidized material available only tocertain segments of the population. For exam-ple, the Egyptian system of food subsidies istargeted at the poor by restricting access tosubsidized flour to the truly poor, by locatingdistribution points in poor neighborhoods,and by using lower-quality products. Never-theless, the system is very expensive (with anestimated financial cost of 2 percent of GDP)and ineffective (World Bank 2005a). Betweenone-quarter and one-third of the poor do notbenefit from it, and fully 83 percent of thevalue of the food subsidies goes to the non-poor. Moreover, those poor and vulnerablehouseholds that do benefit receive so little thatthe net effect is to lift only 5 percent of thepopulation out of poverty.

General fuel subsidies tend to be even moreregressive and more costly than food subsidiesbecause they involve substantial leakages ofbenefits to higher-income groups. A study offive countries from various regions found thaton average 78 percent of fuel subsidies went tothe richest 60 percent of households (Coadyand others 2006). Even when targetedthrough voucher programs, fuel subsidies tendto be ineffective. In India, for example, abouthalf of subsidized kerosene34 (which is madeavailable to poor families on a quota basis at9 rupees a liter) is diverted to the black mar-ket where it is either sold at a higher price oris used to adulterate diesel, which sells forabout 30 rupees per liter.35

More generalized price subsidies or pricefloors (including indirect ones such as man-

dating the national oil company to sell atbelow cost) are also common and can be veryexpensive.36 Estimates suggest that India’stotal fuel subsidies amount to about 2 percentof GDP. Even after reform, the fuel subsidy inIndonesia is expected to total 127 trillionIndonesian rupiah ($13.9 billion) in 2008 andmake up about 13 percent of the country’stotal budget (more than the total of spendingon education and health).

The imposition of export bans by food-exporting countries has the same basic goal ofkeeping consumer prices below the marketlevel.37 Some 20 developing counties haveintroduced such bans since 2007, includingArgentina, China, Egypt, India, Kazakhstan,Pakistan, Russia, Ukraine, and Vietnam. Sev-eral different policies have been used, includ-ing export taxes on a particular commodity(India), taxes on transport (Kazakhstan), re-stricting licenses to export (Argentina), and acomplete ban on exports (Vietnam).

Price containment policies distortincentives, reducing supply, limitingconservation, and exacerbating andprolonging high pricesWhile expensive and generally poorly targeted,all of these policies (price subsidies, pricefloors, and export bans) do succeed in limitingthe immediate domestic impact of rising inter-national prices. However, they do so at a cost.Not only are they fiscally unsustainable inmany cases, but they also tend to exacerbateand prolong the price increase. Lower pro-ducer prices mean that less new supply is forth-coming, while lower consumer prices meansthat demand is not curtailed—both domesti-cally and internationally. For example a seriesof steps taken by Serbia in 2007 to secure do-mestic supply, including a temporary ban onexports of wheat, maize, soybeans, and sun-flower backfired. Serbia’s wheat plantings fellto a 90-year low (partly because of badweather) and prices rose (USDA 2007).

The problem with export bans is even moresevere. Although they are domestically appeal-ing, these bans decrease confidence of net


123

10363_Pg95_140:10363_Pg95_140 11/29/08 7:08 AM Page 123

importers in the international trading systemas a reliable source of food. For example, fol-lowing India’s ban on exports of premium riceon October 9, 2007, domestic prices remainedwell below international prices, but the with-drawal of supply from international marketssparked an almost immediate rise in interna-tional rice prices (figure 3.17).38

Although countries’ food security concernsare legitimate, a widespread return to policiesof food self-sufficiency could be very costlydepending on how quickly it is achieved, theresource endowments of the country and thepolicies used to achieve it.39 If investments inresearch and infrastructure are made to im-prove productivity, the costs may not be toohigh. Although the rate of return on suchinvestments is high, it can take many yearsto raise production enough to achieve self-sufficiency.

If price policies are used to boost domesticproduction, the costs could be very high andthe effectiveness uncertain. First, the supplyresponse of the agricultural sector as a wholeis low (Cavallo 1988).40 Raising the total ofagricultural production as opposed to produc-tion of a single crop takes many years. Thus,unless a policy is very carefully constructed, itrisks increasing production in one food item atthe expense of reduced production (and in-creased dependence) in another.

Moreover, using a price subsidy or importrestrictions to boost domestic prices and in-duce additional production is often a costlyalternative to importing (table 3.9). For exam-ple, to increase domestic rice output by10 percent, a country would have to increasedomestic prices by as much as 50 percent.41

For the 10 largest rice importers over 2000–05(who imported about 10 percent of their totalconsumption), achieving self-sufficiency inthis way would imply a $24 billion dollar in-crease in food costs compared with the currentsituation where the rice is imported—mainlybecause the extra 50 percent would have tobe paid both on the rice currently produceddomestically as well as on the new rice to beproduced (currently imported).

A better approach would be to enter intolong-term supply arrangements, such as thosediscussed earlier in the context of the oil mar-ket. Under these agreements, importing coun-tries could agree to buy a minimum amount ofgrain or other food crop each year in exchangefor a commitment by the exporting countryto meet larger imports when needed. Alterna-tively, countries might make more intensive


124

Figure 3.17 After India banned rice exports,international prices rose

300

07/1

7/07

07/3

1/07

08/1

4/07

08/2

8/07

09/1

1/07

09/2

5/07

10/0

9/07

10/2

3/07

11/0

6/07

11/2

0/07

12/0

4/07

12/1

8/07

315

330

345

360

375India banned riceexports

US$/ton

Source: International Grains Council, USDA.

Note: Price for Thailand's export price ($/ton) for 100 percent Bwhite rice.

Table 3.9 Increasing rice self-sufficiencycan be more costly than relying on imports

Cost of riceconsumption

Import Self-Production Consumption Imports strategy sufficiency

(millions of metric tons) ($US billions)

China 123.2 133.8 10.6 28.8 43.2Indonesia 33.8 36.1 2.3 7.8 11.6Nigeria 2.2 3.7 1.6 0.8 1.2Iran,

IslamicRep. 1.6 2.9 1.3 0.6 0.9

Iraq 0.1 1.1 1.0 0.2 0.4European

Union 1.7 2.6 0.9 0.6 0.8Philippines 8.7 9.6 0.9 2.1 3.1Bangladesh 25.3 26.0 0.8 5.6 8.4Senegal 0.1 0.9 0.7 0.2 0.3Côte

d’Ivoire 0.5 1.2 0.7 0.3 0.4Total 197.2 217.9 20.7 46.8 70.3

Source: World Bank.

10363_Pg95_140:10363_Pg95_140 11/29/08 7:08 AM Page 124

use of the kinds of conditional contracting re-cently used by Malawi (see box 3.6).

Over the medium term, countries need tomove toward more flexible and targetedsocial safety net schemesHaving weathered the initial consequences ofhigh food and fuel prices, countries need totransfer more of the burden of dealing withhigh prices to better-targeted social safety netsand market mechanisms. Doing so will bringboth fiscal and economic benefits, in the formof increased poverty reduction, reduced cost,and lower commodity prices.

There is no magic prescription for effectivesocial safety nets, especially among developingcountries where both fiscal and administrativeresources are often in short supply. Successfulsystems usually consist of several individualprograms that complement each other as wellas other public or social policies. Ultimately,the particular policy mix put into place willdepend on the country context.

Nevertheless, there is general consensus onthe relative strengths and weaknesses of dif-ferent forms of support. A loose ranking ofprograms would favor targeted cash transfersof adequate coverage, generosity, and qualityas the best option and could include increasingpensions and unemployment benefits whenthey target the poor (box 3.8).

Emergency food aid distribution, used inplaces like Afghanistan and Angola, often inpartnership with agencies such as the WorldFood Programme (WFP), ensure food securityfor vulnerable groups and are appropriatewhere markets are functioning poorly orwhere foreign assistance is only available in-kind, but the physical transfer and potentialleakages can make these programs costly.School feeding programs can be used for aquick response, but these do not typically ad-dress child malnutrition at its most criticalpoint—when children are in their infancy. Con-ditional cash transfer programs can help fosterincreased use of health and education servicesand are generally most efficient, but they arenot always a feasible option in low-income

countries with weak administrative capacity.Finally, public works programs, in food orcash (such as in Cambodia and Mozambique),can be effective only for a few areas and forpeople who are currently unemployed.

Household targeting systems—such asproxy means tests or means tests, sometimescommunity-based decision making, or hybridsamong these—can be effective in directing re-sources to the poor. Where a household tar-geting system is not in place, a combination ofgeographic targeting, self-targeting, or demo-graphic targeting can produce at least moder-ately good results, reducing the cost of admin-istrative targeting.42 For example, schoolfeeding programs targeted geographically topoor rural areas may have relatively lowerrors of inclusion. Self-targeting can beachieved by setting low wages for labor-inten-sive public works. Open market operationsfor food sales can be geographically targetedto slum areas, with a limitation on quantityand provision of an inferior staple commodityinducing some degree of self-targeting. Feesfor networked electricity can be differentiatedby use level or neighborhood. Provision of for-tified weaning foods that are culturally ac-ceptable for only very young children is agood use of demographic targeting.

Although the economics of reform aresolid, eliminating existing but inefficientantipoverty measures is politically difficultRemoving subsidies is difficult and can be metwith strong opposition and violent protest.Nevertheless, given the fiscal burden thatsuch subsidies impose—especially on oilimporters—governments have little choice butto reform. While many different approacheshave been followed, those that have workedhave tended to use a strategy that replaces thesubsidy with a better-targeted benefit, pre-ceded by an effective publicity campaign thatemphasizes the poorly targeted nature of theexisting subsidy (Kojima and Bacon 2006).

Several countries have used some variationof this approach. Chile made a one-time pay-ment of $28 to low-income households to


125

10363_Pg95_140:10363_Pg95_140 11/29/08 7:08 AM Page 125

compensate for higher fuel prices and providedextra cash compensation to 1.4 million house-holds consuming less than 150 kilowatt-hoursof electricity a month. Indonesia used an effec-tive public relations campaign, coupled with acash compensation scheme and general trust inthe government, to more than double gasoline

and diesel prices and nearly triple keroseneprices in 2005 with no substantial opposition.Ghana combined prior analysis of who bene-fited from fuel subsidies with a campaign pub-lishing the measures that would be used tocompensate for removing subsidies in a suc-cessful effort to remove subsidies (box 3.9).


126

Targeted cash transfers are the cornerstone ofsafety net programs in most of the countries

with safety nets. They help protect poor householdsby providing them with the resources they need tomaintain a minimum level of consumption. These arethe most flexible programs and can be adapted toparticular circumstances. It is not surprising that tar-geted cash transfers are used in countries of varyingincome level, from Albania to Mexico to Zambia.

Even poor countries can afford to allocate re-sources for safety net programs. The fiscal costs of awell-targeted safety net for the poorest need not beunduly high. For a large share of developing coun-tries, spending on overall safety nets has been on theorder of 1–2 percent of GDP in recent years. How-ever, the costs of the responses differ according tothe scope, generosity, and degree of targeting: rang-ing from a mere 0.04 percent of GDP in Chile (for awell-targeted response) to more than 1 percent ofGDP in Ethiopia (for lifting the value added tax onfood grains, raising the wage on the cash-for-workprogram, and distributing wheat to the urban poorat a subsidized price). A careful fiscal-planning exer-cise will be needed in each country. Such a planshould seek to protect critical growth-enhancingspending and prune low-priority expenditures, andbe embedded in a medium-term fiscal sustainabilitystrategy so that the longer-term fiscal sustainabilityof the program is ensured. For the poorest countries,international assistance will be essential.

The quality and care with which programs aredesigned and implemented, including the selection,provision, and monitoring of benefits, have a largeimpact on program efficiency and effectiveness.No program is a guaranteed success, and few areguaranteed failures. The role of good systems and

Box 3.8 Conditional cash transfers are most effectivein getting money to the poor

adroit managers in getting the most from a programcannot be overemphasized.

Conditional cash transfer programs have a goodreputation and are an effective mechanism for direct-ing assistance toward the poor. Large-scale condi-tional cash transfer programs were developed inMexico (Progresa, Oportunidades) and in Brazil(Bolsa familia) and later spread to other countries inLatin America and the Caribbean and to the rest ofthe world. Those programs are well targeted to poorfamilies through a combination of geographic priori-tization and household assessment mechanisms andare particularly efficient in providing transfer to thepoor. Administrative costs are relatively low, averag-ing about 5 percent of total program costs afterstart-up, compared with food-based programs,whose administrative costs average 36 percent oftotal program costs. However, because they are moredifficult to set up than unconditional programs andmight exclude the neediest where services are scarce,cash transfer programs can be part of an emergencyresponse, for example to high food prices, wherethey are already established.

Care must be taken to ensure that the policy re-sponse to temporary crises is temporary. Although apermanent increase in fiscal space may be justified incountries that have underinvested in adequate safetynet systems, in countries that already had broadlyadequate safety nets a temporary expansion of bene-fits may be best. Permanent changes in the benefitlevels or scope of the transfer program can beavoided by targeting additional benefits at those al-ready qualified for a program; making payments in alump sum or explicitly time-limited fashion.

Source: Grosh, del Ninno, and Tesliuc 2008.

10363_Pg95_140:10363_Pg95_140 11/29/08 7:08 AM Page 126

The international response tohigh commodity prices

The effectiveness of the policy response tothe recent rise in food and fuel prices

will, in the main, depend on the ability of in-dividual governments to put in place well-targeted programs to ameliorate hardshipand to provide the infrastructure, services,and appropriate incentives required to raisefood production and encourage adjustmentsto high food and fuel prices. For the poorestcountries, some form of additional assistancewill be required, while for other countries in-ternational coordination may be required tohelp restore confidence in global food mar-kets and provide emergency assistance forpoor consumers.

The loss of real income from higher foodprices is too great to compensate allconsumersAs discussed earlier, the rise in food and fuelprices substantially reduced the purchasingpower of the poor throughout the developing

world. During such episodes, short-term assis-tance is urgently needed to avoid hardship.However, effective targeting of assistance iscritical. The cost of compensating all con-sumers for the rise in food prices alone sinceJanuary 2007 is impossibly large—perhapsmore than $270 billion annually. Moreover,insulating consumers from the effects of priceincreases (and taxing producers to finance thisassistance) delays the necessary adjustments indemand and supply that will eventually bringprices down.

Even if a program could be devised thatconcentrated aid only on the poor, it wouldcost some $38 billion annually, or about14 percent of all official development aid in2007. Focusing international assistance onthe poorest countries makes sense, in part be-cause higher proportions of their populationsare extremely poor and because their own fis-cal resources are particularly weak. The totalcost of reversing the poverty impact of higherfood prices in IDA-eligible countries wouldbe a more manageable $2.4 billion.


127

Ghana could not continue fuel subsidies as worldoil prices rose in 2004, and the government

launched a poverty and social impact assessment tostudy the situation. Guided by a steering committeeof stakeholders from ministries, academia, and thenational oil company, the assessment was completedin less than a year. By the time the government an-nounced the 50 percent price increases in February2005, it could use the assessment findings to makeits case for liberalizing fuel prices to the public—including the fact that the price subsidies mostlybenefited the better-off.

The minister of finance launched the public rela-tions campaign with a broadcast explaining the needfor the price increases and announcing measures to

Box 3.9 Removing fuel subsidies in Ghanamitigate their impact. A series of interviews withgovernment officials and trade union representativesfollowed. The Energy Ministry used newspaper ad-vertisements with charts to show that Ghana’s fuelprices were the lowest in West Africa, after Nigeria’s.

The mitigation measures, which were transparentand easily monitored by society, included an immedi-ate elimination of fees at government-run primaryand junior secondary schools and a program toimprove public transport. Although the trade unionsremained opposed to the price increases, the publicgenerally accepted them, and no large-scale demon-strations occurred.

Source: Bacon and Kojima 2006.

10363_Pg95_140:10363_Pg95_140 11/29/08 7:08 AM Page 127

The international community has reactedswiftly to the rise in food pricesThe international community has been quickto recognize the serious risks that higher foodprices posed for the poor. The United Nationshas established a Task Force on the GlobalFood Security Crisis to formulate a unified re-sponse to the food crisis (box 3.10).

Donors ramped up existing programs andlaunched new initiatives to speed the provi-sion of food aid to the poor. Examples includethe Food and Agricultural Organization haslaunched the Initiative on Soaring Food Prices,which assists small-holders in critically af-fected countries (beginning with BurkinaFaso, Haiti, Mauritania, and Senegal) to ob-tain seeds, fertilizers, and animal feedstock;the International Fund for Agricultural Devel-opment (IFAD) is making up to $200 millionfrom existing loans and grants available to im-prove poor farmers’ access to seeds and fertil-izer; bilateral donors (for example, the U.S.Agency for International Development andthe U.K. Department for International Devel-opment) are focusing existing programs oncountries most affected by the food crisis; theEuropean Union has committed ⇔€1.0 billionin funds from European farm subsidies that

have not been used (because high prices havereduced the compensatory amounts payableto farmers) to farmers in developing countries,mostly in Africa; and the World Food Pro-gramme has pledged $214 million to provideassistance to vulnerable groups.

For its part, the World Bank has created a$1.2 billion rapid financing facility, the GlobalFood Crisis Response Program (GFRP), to ad-dress immediate needs arising from the foodcrisis. The facility includes $200 million ingrants targeted at vulnerable poor countries,with priority given to the most fragile states.

The GFRP strives to create a balance be-tween short-run food stabilization and mea-sures to ensure that countries are able to copebetter in the medium term. Countries can se-lect measures most relevant to their individualsituations from program components thataddress price policies, social protection andnutrition, and immediate supply response pro-visions for getting seeds and fertilizers tofarmers.

The World Bank is also establishing a mul-tidonor trust fund, with an initial contributionfrom Saudi Arabia, to help the poor respondto high energy and food prices. This fund willoperate in parallel with the GFRP and will


128

The UN secretary-general established a Task Forceon the Global Food Security Crisis aimed at pro-

moting a unified response to the global food pricechallenge. An initial meeting was held in June 2008,attended by 181 countries, and 60 nongovernmentaland civil society organizations.

The summit concluded with a declaration callingon the international community to increase assis-tance for developing counties, in particular the leastdeveloped countries and those that are most nega-tively affected by high food prices. The immediate re-sponse was to call for increased humanitarian assis-tance to those hardest hit by the rise in food prices

Box 3.10 The international response to risingfood prices

through food aid and balance of payments supportto countries. The medium-term response has been toassist countries to put in place revised policies andmeasures to help farmers, particularly small-scaleproducers, to increase production and integrate intolocal, regional, and international markets along withmeasures to moderate the fluctuations in food grainprices through increased stockholding capacity andbetter use of risk management practices. Longer-termresponses have focused on how to increase the re-siliency of food production systems to challengesposed by climate change.

10363_Pg95_140:10363_Pg95_140 11/29/08 7:08 AM Page 128

provide priority assistance to countries whoseeconomies are most severely affected by theincrease in the price of imported fuel, thathave already embraced or are pursuing ener-gies policies that are more fiscally sustainable,and that propose cost-effective social safetynet programs.

Improvements are required in thearchitecture for humanitarian aid tostrengthen the response to the food crisisThe dramatic increase in food prices has un-derlined the importance of improving the effi-ciency of programs to deliver emergency foodaid. Bilateral food aid programs are largelybased on the disposal of surplus commodities.This approach has played an important role ingarnering political support for the provisionof food aid. However, 60–86 percent of theaid is tied, either directly to commodities pro-vided by the donor country or through con-straints on the use of cash donations (FAO2006). As a consequence, the cost of this aidcan be 30–50 percent higher than nontiedsources (OECD 2005). Moreover, tied foodaid of this type slows the delivery of food aid,and reduces supplier incentives in local foodmarkets. 43

Progress is being made in improving theadministration of food aid programs. Somedonors have lifted requirements that food aidbe procured domestically and have shiftedfrom providing commodities to providingcash, making it possible to purchase some foodlocally. Resources have shifted toward the pro-vision of emergency aid, implying an improve-ment in the targeting of food aid (FAO 2006).Additional efforts to provide cash aid andallow the food to be purchased where andfrom whom made most economic sense wouldreduce costs and help make food aid a moreefficient instrument in reducing poverty.

Improvements in food aid management arerequired at the international level as well. Themain multilateral provider of food assistanceis the UN’s World Food Programme, whichdelivers more than half of the humanitarianfood aid in the world. Higher food prices have

made it very expensive for the WFP to pur-chase food on international markets, threaten-ing its capacity to deliver emergency humani-tarian aid in a timely manner.

A strengthening of the financing arrange-ments for the WFP could markedly improve theefficiency of its operations, allowing for an ex-pansion in food aid and a reduction in costs.44

Financing of the WFP depends on voluntarycontributions from donor countries that arelargely tied to assistance for specific countriesor programs on a year-to-year basis.45 As a re-sult, WFP programs can be designed and im-plemented only after financing is committed.Contributions are often based on surplus dis-posal, with provision that the food be trans-ported on the carriers of the donating nation.

These arrangements are major constraintson the WFP’s ability to respond flexibly andefficiently to the need for assistance. The timerequired to obtain donor commitments makesit difficult to respond to unexpected shocks.The timing of commitments also can meanthat food purchases must be made when pricesare at seasonal highs rather than followingharvest when prices are at seasonal lows. Sev-eral donors provide commodities rather thancash, significantly increasing the cost of foodcompared to local purchases. Providing an an-nual dollar budget equivalent to the value ofcurrent commitments would dramatically im-prove the efficiency of WFP operations. Giventhe volatility of food prices, this budget mightbe supplemented by a line of credit uponwhich it could draw in years when eitherprices or needs are unusually high.

Steps to assist the replenishment ofinternational grain stocks would helpThe role that low stocks have played in therise of food prices has raised the issue ofwhether or not an international food stockpileshould be created to help prevent a repetitionof the past year’s high prices, in part by ensur-ing that supply would be available to themarket and by dissuading speculative behav-ior. While an appealing notion, it is not clearthat such a stockpile would be effective—or


129

10363_Pg95_140:10363_Pg95_140 11/29/08 7:08 AM Page 129

needed. To have a significant dampening effecton the market, such a stockpile would have tobe large and would be very expensive to createand maintain. Rough calculations suggest thata stockpile equivalent to 10 percent of globalproduction would cost about $66 billionto create and some $8–10 billion annuallyto maintain.46 Moreover, the creation of thestockpile would add significantly to globalfood demand and price pressures during theperiod in which it was being created. Nor is itclear that a global stockpile would actually in-crease world stocks. The public stock increasemay well be matched by a reduction in privatestocks, thus transferring the costs of keeping astock to the public sector without necessarilyimproving the stability of the market.

A more effective strategy might be to im-prove information flows about stocks and cre-ate mechanisms by which they can be man-aged. Currently most stocks are held by alimited number of major producers and im-porters. It may be possible to create an inter-national agreement that provides for the shar-ing of some of these costs—perhaps along thelines of the International Energy Agencyagreement governing oil reserves. As in thatagreement, the rules for accumulating and dis-tributing grain stocks would need to be clearlydefined to prevent their being used for surplusdisposal or price support and to ensure theyare used for humanitarian purposes.

More multilateral discipline in trade policieswould help mitigate the rise in food pricesA range of multilateral and trade policies(export restrictions, biofuel subsidies, tariffs,mandates, and global protection of agriculturemore generally) have contributed to the risein food prices. Moreover they have reducedconfidence in the international food tradingsystem and interfered with consumer and pro-ducer incentives, reducing supply and increas-ing demand. As a result, the price hike hasbeen larger and longer lasting than it wouldhave been otherwise.

A strengthening of existing internationalrules governing the imposition of export re-

strictions may be desirable. Currently, unlikecountervailing duties, the conditions that mustbe met before export restrictions are intro-duced are ill defined, and although there is arequirement that the World Trade Organiza-tion be notified of their implementation, it isnot enforced.47 Even the enhanced rules pro-posed under the Doha Round should probablybe strengthened.48 Helpful measures might in-clude including stricter (even pre-) notificationrequirements, limits on the allowed durationof restrictions, and possibly a definition of theconditions under which such restrictionsmight be admissible.

Policy makers should also consider phasingout biofuel subsidies and production man-dates, especially where these are coupled withtariffs that restrict imports from lower-costproducers. This step would both reduce pres-sure on food prices and help low-cost andenvironmentally cleaner developing-countrybiofuel producers that are currently shutout of major markets by these rules.49 Thereare indications that a number of developedcountries are beginning to reexamine theirbiofuel policies, but it remains a contentiousissue.

More fundamentally, decades of trade-distorting policies (such as tariffs, quantitativerestrictions, and subsidies) are partly responsi-ble for the current spike in food prices, havingencouraged inefficient agricultural productionin rich countries and discouraged efficientproduction in developing countries (Chauf-four 2008). The kind of agricultural trade bar-rier reductions contemplated in the DohaRound might lead to higher agricultural pricesin the short term, but in the long run, theyshould help establish a more transparent,rules-based, and predictable food trading sys-tem that would stimulate trade and raise in-comes around the world. An ambitious pro-gram could reduce global poverty by as muchas 8 percent (World Bank 2004).50

Moreover, removal of the rules that allowsuch trade restrictions would help ensure that,as prices come down, countries cannot intro-duce new subsidies and restrictions in an effort


130

10363_Pg95_140:10363_Pg95_140 11/29/08 7:08 AM Page 130

to prevent domestic producers’ prices from de-clining as sharply as they would otherwise.

Conclusions

The rise in primary commodity prices since2003 was much larger and more sustained

than those of earlier periods. This boom gen-erated dramatic transfers of income within andamong countries and has imposed severe bur-dens on some consumers. However, it has alsocreated opportunities for producers and these,if managed properly, can provide significantgrowth opportunities. The boom has also ex-posed weaknesses in domestic and interna-tional policies that have contributed to andprolonged the period of high prices and re-duced confidence in international markets.

For commodity producers, commoditydependence need not hurt long-term growth.Although commodity-dependent economieshave, on average, grown more slowly thanmore-diversified economies, for mosteconomies dependence on commodities is theresult of slow growth, not the cause. Toachieve the growth potential inherent in com-modity riches, countries need to implementpolicies that minimize the potential disruptiveimpacts of volatile export revenues, exchangerate appreciation that can erode the competi-tiveness of manufacturing, and incentives forrent seeking and corruption. It would appearthat producing countries have responded tohigher prices in a more prudent manner dur-ing this boom than in the past. Fiscal policyhas been less procyclical than in the past,countries have made greater efforts to savewindfall profits, and rate appreciation hasbeen muted. As a result, they are less likely toendure the major setbacks that characterizedthe 1980s as prices declined. An exception tothis generally welcome response has been theperformance of countries with newfoundcommodity wealth and some newly indepen-dent resource-rich countries that may haverepeated some of the mistakes of the past.

Consumers have faced daunting challengesfrom the commodity price boom. The rise in

food prices has presented the greater challengebecause the poor in developing countries spendas much as half of their incomes on food, whilefuel is a smaller share of their expenditures. Therise in food prices has increased poverty andboosted the cost of many countries’ poorly tar-geted and inefficient subsidy programs, whichby limiting the impact of food and fuel pricesimpede the necessary adjustment to high prices.

The expansion of existing programs and theadoption of emergency measures are under-standable, given the magnitude of the oil andfood price increases, the potentially dire impli-cations for the poor, and the limited time. How-ever, the high cost of this response underlinesthe importance of putting in place well-targetedand efficient safety net programs, so that nexttime countries can address the needs of thepoor without incurring undue fiscal costs. Thisepisode has also shone light on the need forinternational coordination to encourage coun-tries to avoid counterproductive policies andto marshal aid resources to help the poor.

Policies to deal with the rising food andfuel prices have often exacerbated the prob-lem by slowing necessary adjustments. Suchpolicy responses have included price controlsand export bans that have impaired incentivesto reduce consumption and invest in the addi-tional capacity that would help bring pricesdown, while weakening confidence in theinternational trading system.

The dramatic increase in food prices hasunderlined the importance of improving theefficiency of programs to deliver emergencyfood aid and transition these programs fromlargely surplus disposal programs to effectivehumanitarian assistance programs with fewerconstraints on their use. A range of multilat-eral and trade policies (export restrictions,biofuels subsidies, tariffs, mandates, andglobal protection of agriculture more gener-ally) have contributed to the rise in foodprices and need to be reconsidered. The DohaRound, while not likely to lower food pricesin the near term, would provide longer-termdiscipline to agricultural policies and raiseincomes around the world.


131

10363_Pg95_140:10363_Pg95_140 11/29/08 7:08 AM Page 131

Technical Annex: SensitivityAnalysis

The poverty effects of higher food pricesdiscussed in chapter 3 are based on a num-

ber of assumptions. This annex reports thesensitivity of the results (change in the numberof poor and the change in the income gapratio) under different assumptions regardingthe nature of the price shock and the propor-tion of increased food expenditures that accrueto agricultural households.

The results presented in the main text de-flate the increase in food prices by the non-food deflator. More traditionally in high-income countries, where food represents asmall share of total spending, real food pricesare deflated by the overall consumer priceindex. If the whole consumer price index hadbeen used to deflate the increase in foodprices, the overall shock would have beenmuch smaller and hence the estimated povertyeffects would have been milder. Under thisscenario, labeled “real price change” in table3A.1, the total number of poor would bearound half as large as in the central scenario.

Another important assumption driving theestimated poverty effects is the allocation ofthe revenues from higher food prices to differ-ent households. In the central scenario, pro-ducer prices are increased by the same pro-portion as consumer prices. To the extent thatall of the increase in retail food prices is at-tributable to an increase in farmgate prices,then the proportional increase in farmgateprices should have been larger than that expe-rienced by retail prices.51 The other issue ishow the price change affects the incomes ofdifferent households. In the kind of short-termsimulation being conducted here, wages andemployment are normally held constant.Therefore, only the incomes of self-employedagricultural workers or landowners, who sellthe final product, should increase, not those ofagricultural wage laborers. Unfortunately, theGIDD database does not distinguish betweendifferent income sources. Therefore the datain the GIDD is complemented with informa-tion from the Rural Income Generating Activ-ities (RIGA) project. RIGA is an FAO–WorldBank funded project that uses data from 21(household) Living Standards Measurement


132

Table 3A.1 Sensitivity analysis

Real price change Relative price change

Central Scenario:Self-employment Self-employment

All agricultural agricultural incomes All agricultural agricultural incomesRegion incomes affected affected incomes affected affected

Change in number of poor (million)East Asia and the Pacific 52.1 59.9 103.7 114.7Europe and Central Asia 0.0 0.0 0.1 0.1Latin America and the Caribbean 0.4 0.8 0.7 1.3Middle East and North Africa 1.9 3.0 4.6 7.2South Asia 10.8 14.3 16.8 24.4Sub-Saharan Africa 2.0 2.2 5.7 5.9Developing world 67.2 80.3 131.6 153.5

Change in income gap ratio (percent)East Asia and the Pacific 0.36 0.43 0.78 0.93Europe and Central Asia 0.00 0.00 0.00 0.00Latin America and the Caribbean 0.00 0.01 0.01 0.01Middle East and North Africa 0.03 0.06 0.09 0.15South Asia 0.17 0.24 0.28 0.43Sub-Saharan Africa 0.10 0.11 0.30 0.32Developing world 0.16 0.21 0.33 0.41

Source: World Bank.

10363_Pg95_140:10363_Pg95_140 11/29/08 7:08 AM Page 132

Surveys (LSMS) to identify the various incomegenerating activities of rural households.52

The information on total agricultural incomesand self-employment agricultural incomes re-ported in RIGA is used to estimate the econo-metric relationship between this and per-capita household income and consumption,which was then used to impute agricultural in-come shares in all the households included inthe GIDD (De Hoyos and Medvedev 2008).53

If the short-term price increase benefitsonly self-employed landowners, the increaseof self-employment agricultural incomesshould be larger than the increase in retailprices. At the limit, if agricultural wages andemployment are held constant, then all of theadditional income would accrue to landown-ers and none to farm workers.

Mathematically,

Pc1 * Q1 � �1 * SE � W1 * E � other costs,

where Pc1, Q1 are the retail price and quantityconsumed of good 1, respectively. �1, W1 areremunerations of self-employed workers (in-cluding the return to land to self-employedlandowners) and wage earners, respectively.Rearranging:

Pc1 � �1 � W1 � ,

where SE�Q is profits share in total output.We denote these as alpha and those of othercosts as beta, giving us:

Pc1 � �1� � W1(1 � � � �) � � .

Taking the total derivative while holdingwages and other costs constant gives us:

Pc � � �

or

� � Pc.ddt

1�ddt

ddt

ddt

other costQ

SEQ

EAG

Qother costs

Q

Numerically, if the landowner’s share in thevalue of output initially is 50 percent, then thepercent increase in his revenues will be twicethat of the increase in the retail price (assum-ing all the changes in retail price are translatedinto increases in profits).

In the central scenario, all agricultural in-comes are raised by the same amount as retailprices. This is tantamount to assuming thatwages, self-employed profits, and other costsall rise by the same proportion as the increasein consumer food prices.

It is also equivalent to assuming that all ofthe increase in farm incomes accrue tolandowners but that all the farm workerswork for poor landowners.

An alternative assumption is to assume thatonly landowner incomes and other incomesrise in the same proportion as consumerprices. This essentially assumes that none ofthe agricultural workers work for properlandowners. Under this assumption, the head-count poverty rate increases by substantiallymore— 153 million (see results in table 3A.1under the label “self-employed agricultural in-comes affected”).

The lower panel of table 3A.1 reports thechange in the income gap ratio (Foster, Greer,and Thorbecke 1984)—the average differencebetween the per capita income of poor house-holds and the poverty line stated as a percent ofthe poverty line—for the various scenarios. Thedifferences in the income gap ratio between dif-ferent scenarios confirm that larger poverty im-pacts are found when the change in relativeprices is used as the shock and when only self-employment agricultural household incomes areassumed to respond to change in relative prices.

Notes1. The idea that dependence on natural resources

may impede development dates back at least to the de-cline of Spain, a period when it was benefiting fromsubstantial gold inflows from the New World in the17th century (Landes 1999). The idea was forcefullyrestated by development theorists in the decades fol-lowing World War II (such as Prebisch 1950 and Singer1950) and continues to attract attention.


133

10363_Pg95_140:10363_Pg95_140 11/29/08 7:08 AM Page 133

2. Sachs and Warner (1995, 2001) are perhaps themost influential. See also Gylfason, Herbertsson, andZoega (1999); Leite and Weidmann (1999); Auty(1998); and Bravo Ortega and De Gregorio (2005).Gylfason (2001) finds that resource dependence is as-sociated with lower education levels, implying thateconomies dependent on primary commodities havelimited incentives to invest in human capital. Ledermanand Maloney (2007) find that the Sachs and Warner re-sults are not robust to data modifications and changesin estimation techniques.

3. Bevan, Collier, and Gunning (1991) providecase-study evidence of excessive expenditures, debt ac-cumulation, and low-quality investments during com-modity price booms in Sub-Saharan Africa. Cudding-ton (1989) finds that many developing countriesoverspent during and after the 1970s boom.

4. Manzano and Rigobon (2006), for example, findthat the post-boom slowdown in Latin America in the1980s was almost entirely explained by the debt over-hang accumulated during the boom period.

5. The average data presented in this section tend toobscure the great diversity of country experiences, be-cause both the rate of increase in government expendi-tures and in exports (relative to GDP) vary enor-mously. The difference between changes in the ratio ofexports to GDP and changes in the ratio of governmentexpenditures to GDP may be viewed as a rough sum-mary indicator of the fiscal response to primary com-modity booms. In both the 1980s and the 2000s, thisdifference varied by as much as 60 percent of GDPbetween countries.

6. Historically, the very different circumstances fac-ing individual countries were reflected in diverse fiscalresponses to commodity booms. For Sub-SaharanAfrica, see Deaton and Miller (1995); for a more geo-graphically diverse collection of countries, see Collierand Gunning (1994).

7. This analysis includes developing countrieswhere primary commodities accounted for more than70 percent of merchandise exports. Boom periods aredefined as sequential increases in merchandise exportrevenues that average more than 10 percent a year.Thus “booms” do not represent trough-to-peakchanges in prices but simply periods of rapid growthin export revenues in countries dependent on primarycommodities. We report simple averages of the per-centage point change in the ratios of exports and gov-ernment expenditures to GDP.

8. Because of the small number of countries in thesample for fuel exporters during the 1980s (owing tothe lack of government expenditure data for manycountries), these results must be treated with caution.The basic results for nonfuel primary commodity ex-porters remain robust to the exclusion of the two

largest outliers in the sample (São Tomé and Principe,whose government expenditures declined by 45 per-centage points, and Paraguay, whose export revenuesrose by 37 percentage points of GDP).

9. For any given price forecast, countries with 70 or80 years of reserves at current production levels have ahigher permanent income from the oil price rise thancountries with only 10 or 20 years of reserves at cur-rent production. Thus, assuming countries wish tosmooth the revenue flow over an extended period oftime, countries with large reserves relative to produc-tion should spend a larger share of the current revenuesthan countries with smaller reserves.

10. The countries of concern here are mostly oil ex-porters. Based on available data, only one country(Zambia) relies on minerals for more than 70 percentof export revenues. (Botswana’s dependence on dia-monds would be another example, except that a largeshare of diamond exports are counted as processedgoods in trade statistics.)

11. The calculation of the life span of reserves issubject to considerable uncertainty, given that geolo-gists are continually increasing estimates of reserves,and changes in technology and in prices raise the shareof proven reserves that can be exploited profitably (seechapter 2).

12. This calculation does not take into account theshare of the increase in export revenues captured by thegovernment. Most of the high-reserves countries con-trol their oil resources through a state company, buteven so the government may not see the full proceedsfrom the increase in price.

13. A brief discussion of this type of reduction inthe context of the Dutch disease is given in Sachs andWarner (1995). See also the references they cite andTorvik (2001).

14. Comparisons with the experience of the 1980sare difficult to draw because of missing data for oil-exporting countries. Moreover, after initially appreci-ating, the currencies of many non-oil primarycommodity exporters depreciated sharply in real termsin the 1980s in reaction to the debt crisis, so that for-eign exchange was limited, despite the rise in exportearnings.

15. On corruption, see Lane and Tornell (1999),Baland and Francois (2000), Torvik (2002), and Wickand Bulte (2006). On resource wealth and civil wars,see Collier and Hoeffler (2004). On inefficient distrib-ution of rents, see Acemoglu and Robinson (2001).

16. Mehlum, Moene, and Torvik (2006) provideevidence that natural resource abundance has a nega-tive impact on growth only in countries with poor in-stitutions. Murshed (2004) finds that oil and mineralwealth slows growth through impairing institutionaldevelopment.


134

10363_Pg95_140:10363_Pg95_140 11/29/08 7:08 AM Page 134

17. Oil and mineral wealth can be more heavilytaxed than agricultural wealth (see above) and thusgenerates more opportunities for corruption.

18. The relationship between government revenuesand expenditures has been found to be weaker in coun-tries with national revenue funds than in countrieswithout such funds (Davis and others 2001; Crain andDevlin 2003). Analysis of 15 oil-dependent economiesover 30 years indicates that national revenue funds areassociated with reduced volatility of broad money andprices, but the relationship with real exchange volatil-ity is weak (Shabsigh and Ilahi 2007).

19. In countries with strong political institutions (asmeasured by the existence of effective checks and bal-ances in decision making), government consumption isunrelated to changes in oil revenues (that is, it is notprocyclical), but in countries with weak institutions,government consumption is strongly related to oil rev-enues (Humphreys and Standbu 2004).

20. Such deals are, by no means a new phenome-non. Firms from high-income countries have enteredinto such contracts for several decades.

21. Factors such as delivery specifications, contractliquidity, particular industry structures in variouscountries, and transportation differences make defin-ing standardized contracts more difficult.

22. “Indian Fuel Prices, Too Hot to Touch,” Econ-omist, November 29, 2007.

23. Estimating the impact of rising metals prices iseven more difficult, because metals tend to enter intothe consumption basket of households only indirectlyin the form of manufactured goods.

24. The GIDD data set consists of 73 recent house-hold surveys for low- and middle-income countriescomplemented with more aggregate information on in-come distributions for 25 high-income and 22 devel-oping countries, together representing 90 percent ofthe world’s population.

25. According to household surveys in Africa, therelationship between food shares and per capita house-hold incomes is concave, that is, for very low levels ofincome, food shares accelerate as the households be-come richer. The household surveys indicate that in ex-tremely poor households, consumption items such aswood or kerosene are incompressible.

26. The cost is estimated as the change in thepoverty deficit (Atkinson 1987), that is, the variationin financial resources required to eliminate povertyunder a perfect targeting scenario.

27. This share assumes the same poverty line forrural and urban areas. Ravallion, Chen, and Sangrula(2007) use a higher poverty line for urban areas andshow that the rural share of poverty is 75 percent.

28. Real price increases are calculated as the totalincrease in the ratio of the food and nonfood consumer

price index (CPI) over the period January 2005–-December 2007. This differs from the common prac-tice in high-income countries where the numerator isthe level of the overall CPI including food prices. Thedefinition adopted here provides a better measure ofthe relative increase in food prices because food is avery large share of the overall CPI in most develop-ing countries. Were the more usual measure to be em-ployed, the real price increases would be seriouslyunderestimated.

29. For details on this and other reported simula-tions, see De Hoyos and Medvedev (2008).

30. Despite a very different methodology and amuch smaller sample set, Ivanic and Martin (2008)arrive at a similar figure—105 million.

31. In part, this reflects the influence of higher oilprices on nonfood prices—-the numeraire used for cal-culating real food price increases. Unfortunately, toofew countries had information on the actual impact ofhigh fuel prices on the consumer price index to use anonfood non-oil index to deflate the increase in foodprices.

32. An analysis of Argentina suggests that a 10 per-cent increase in prices will increase output by 3.6, 7.1,and 17.8 percent after 5, 10, and 20 years respectively(Cavallo 1988), a result that is consistent with Bin-swanger’s (1989) estimate that long-run effects maytake between 10 and 20 years to play out.

33. The pass-through was defined as the ratio ofabsolute changes since December 2003 in the retailprice of fuel and the local currency price of the relevantfuel import product.

34. Many countries subsidize kerosene, which isused for lighting and cooking fuel by the poor, andunlike gasoline and diesel, whose retail prices roseby more than the international price in 2007, themedian increase in domestic kerosene prices was only85 percent of the international price increase (Mati2008).

35. “Indian Fuel Prices, Too Hot to Touch.” Econ-omist, November 29, 2007.

36. For example, diesel is kept artificially cheap bypreventing state oil companies from raising prices; inreturn these companies issue oil bonds that the govern-ment guarantees.

37. Export bans are not new (the United States im-posed one on soybeans in the 1970s and the EuropeanUnion banned wheat exports in 1995), but their usehas become more common.

38. India’s ban was later replaced by a minimumexport price, which was then replaced by another com-plete ban on exports. Other factors also contributed tothe increase in international rice prices, including thethinness of the international rice market and a simulta-neous decision by consuming countries to increase


135

10363_Pg95_140:10363_Pg95_140 11/29/08 7:08 AM Page 135

demand to build stockpiles. Increased government-to-government rice sales, which are not subject to the ban,have reduced its effectiveness.

39. Reacting to its inability to secure imports of ricein early 2007, the Philippines recently passed policiesaimed at achieving rice self-sufficiency.

40. Although the supply of a single crop may re-spond quickly to an increase in prices, supply is nor-mally achieved through crop switching.

41. Binswanger (1989) estimates the long-termprice elasticity of supply to be approximately 0.2.

42. Self-targeted programs are designed to mini-mize the incentives the nonpoor may have to partici-pate, typically achieved through a mix of rationingbenefits (such as limiting food quantities), imposingphysical requirements (such as manual work for food),and limiting the subsidies to inferior commodities.

43. For example, delivery of emergency food aidprovided under U.S. Title II takes five months, on av-erage (CARE 2006).

44. This discussions is based on “Strengthening theWorld Food Program’s Role in Humanitarian FoodAssistance,” a note prepared by World Bank staff.

45. Fully 93 percent of commitments are tied tospecific operations. A few countries (Canada, theNetherlands, Russia, and the United States) have begunmaking limited three-year pledges.

46. Financing costs (based on a 6 percent interestrate) would be around $4 billion, while storage costswould be around $1.4 billion, based on U.S. storagecosts of $0.29 a bushel or $10.70 a metric ton incurredduring 2004–07 for wheat in the Bill Emerson Human-itarian Trust (pers. comm., Fred Blott, USDA, August11, 2008). Assuming that 3–5 percent of the stockpilespoiled each year (consistent with losses in high-incomecountries), the annual cost would be an additional$3–5 billion.

47. Under existing rules, export restrictions are al-lowed to prevent or relieve critical shortages of food-stuffs or other essential products. The last notification,by Hungary, dates to 1997.

48. The Doha rules, for example, proposed thatnotification be made within 90 days from the entry intoforce of the measure and that it explain the reasons fortheir introduction. The rules also would limit the dura-tion of export restrictions to 12 months unless import-ing members agree to an 18-month period.

49. This need not eliminate the impact of biofuelsproduction on food prices, because at some level allbiofuel production inevitably competes with food foragricultural land, water, and other resources

50. A pro-poor agreement in which rich countriescut tariff peaks to 10 percent in agriculture and 5 per-cent in manufacturing, combined with cuts of 15 and

10 percent in developing countries, respectively, couldyield gains in developing countries of $315 billion over10 years along with gains of $170 billion for rich coun-tries (World Bank 2004).

51. The difference would stem from transport, mer-chandising, and other costs.

52. For more details on the LSMS household surveyssee http://www.worldbank.org/LSMS/. For a completedescription of the RIGA project, including publicationof the first results, see Carletto and others (2007).

53. Notice that given the data restrictions, all ruralhouseholds are assumed to have positive agriculturaland self-employment agricultural income shares, andtherefore a good part of the distribution story behindhigher food prices is lost.

ReferencesAcemoglu, Daron, and James A. Robinson. 2001. “In-

efficient Redistribution.” American Political Sci-ence Review 95: 649.

Aghion, P., G. M. Angeletos, A. Banerjee, and K.Manovia. 2005. “Volatility and Growth: CreditConstraints and Productivity-Enhancing Invest-ments.” NBER Working Paper 1139. NationalBureau of Economic Research, Cambridge, MA.

Aizenman, Joshua, and Nancy Marion. 1996. “Volatil-ity and the Investment Response.” NBER Work-ing Paper 5841. National Bureau of EconomicResearch, Cambridge, MA.

Aksoy, Ataman, and Aylin Isik-Dikmelik. 2008. “AreLow Food Prices Pro-Poor? Net Food Buyers andSellers in Low-Income Countries.” Policy Re-search Working Paper 4642. World Bank, Wash-ington, DC.

Asfaha, Samuel. 2007. “National Revenue Funds:Their Efficacy for Fiscal Stability and Intergener-ational Equity.” Personal RePEc Archive Paper7656. Research Papers in Economics, Munich.

Atkinson, A. B. 1987. “On the Measurement ofPoverty.” Econometrica 55 (4): 749–64.

Auty, Richard M. 1998. Resource Abundance and Eco-nomic Development. Helsinki: UNU World Insti-tute for Development Economics Research.

Baland, Jean-Marie, and Patrick Francois. 2000.“Rent-Seeking and Resource Booms.” Journal ofDevelopment Economics 61: 527–42.

Bacon, Robert, and Masami Kojima. 2006. “Copingwith Higher Oil Prices.” ESMAP Report 323/06.World Bank, Washington, DC.

Barrett, Christopher B., and Emelly Mutambatsere.2008. “Marketing Boards.” In The New PalgraveDictionary of Economics, ed. Lawrence E. Blume


136

10363_Pg95_140:10363_Pg95_140 11/29/08 7:08 AM Page 136

and Steven N. Durlauf, 2d ed. London: PalgraveMacmillan.

Bevan, David, Paul Collier, and Jan Willem Gunning.1991. “Temporary Trade Shocks in DevelopingCountries: Consequences and Policy Responses.”In The Economics of Commodity Markets, ed.David Greenaway and W. Morgan, pp. 64–93.Cheltenham: Edward Elgar.

Binswanger, Hans. 1989. “The Policy Response ofAgriculture.” Proceedings of the World BankAnnual Conference on Development Economics.Washington, DC: World Bank.

Bleaney, Michael. 1996. “Macroeconomic Stability, In-vestment and Growth in Developing Countries.”Journal of Development Economics 48: 461–77.

Bravo Ortega, Claudio, and Jose de Gregorio. 2005.“The Relative Richness of the Poor? NaturalResources, Human Capital, and EconomicGrowth.” Policy Research Working Paper 3484.World Bank, Washington, DC.

Bussolo, M., R. De Hoyos, and D. Medvedev. 2008a.“Economic Growth and Income Distribution:Linking Macroeconomic Models with HouseholdSurvey Data at the Global Level.” World Bank.Washington, DC.

———. Forthcoming. “Global Income Distributionand Poverty in the Absence of AgriculturalDistortions.” In Agricultural Price Distortions,Inequality and Poverty, ed. K. Anderson, J. Cock-burn, and W. A. Martin.

Byerlee, Derek, Robert Myers, and Thomas Jayne.2006. “Managing Food Price Risks and Instabil-ity in an Environment of Market Liberalization.”Agriculture and Rural Development Department,World Bank, Washington, DC.

CARE. 2006. “White Paper on Food Aid Policy.”www.care.org.

Carletto, G., K. Covarrubias, B. Davis, M. Krausova,K. Stamoulis, P. Winters, and A. Zezza. 2007.“Rural Income Generating Activities in Develop-ing Countries: Re-Assessing the Evidence.” Jour-nal of Agricultural and Development Economics4 (1).

Cashin, Paul, Luis Cespedes, and Ratna Sahay. 2002.“Keynes, Cocoa and Copper: In Search of Com-modity Currencies.” IMF Working Paper 02/223.International Monetary Fund, Washington, DC

Caspary, Georg. 2008. “China Eyes Latin AmericanCommodities.” Yale Global. January 18. yale-global.yale.edu.

Cavallo, Domingo. 1988. “Agriculture and EconomicGrowth: The Experience of Argentina 1913–84.”Paper presented at the 20th Conference of Agri-cultural Economists, Buenos Aires, Argentina.

Chauffour, Jean-Pierre. 2008. “Global Food PriceCrisis: Trade Policy Origins and Options.” TradeNote 34, World Bank, Washington, DC.

Chen, S., and M. Ravallion. 2008. “The DevelopingWorld Is Poorer than We Thought, But No LessSuccessful in the Fight against Poverty.” WorldBank Policy Research Working Paper 4703,Washington, DC.

Christiaensen, L., and L. Demery. 2007. Down toEarth: Agriculture and Poverty Reduction inAfrica. Directions in Development series. Wash-ington, DC: World Bank.

Clements, Benedict, Jung Hong-Sang, and SanjeevGupta. 2003. “Real and Distributive Effects ofPetroleum Price Liberalization: The Case of In-donesia.” IMF Working Paper No. 03/204, Inter-national Monetary Fund, Washington, DC.

Coady, David, Moataz El-Said, Robert Gillingham,Kangni Kpodar, Paulo Medas, and David New-house. 2006. “The Magnitude and Distributionof Fuel Subsidies: Evidence from Bolivia, Ghana,Jordan, Mali, and Sri Lanka.” IMF WorkingPaper WP/06/247. International Monetary Fund,Washington, DC.

Coady, David, and David Newhouse. 2005. “Ghana:Evaluation of the Distributional Impacts of Petro-leum Price Reforms.” Technical Assistance Re-port. Unpublished, International Monetary Fund,Washington, DC.

Collier, Paul. 2007. “Managing Commodity Booms:Lessons of International Experience.” Centre forthe Study of African Economies. Oxford Univer-sity, Oxford, U.K.

Collier, Paul, and Anke Hoeffler. 2004. “Greed andGrievance in Civil Wars.” Oxford EconomicPapers 56: 563–95.

Collier, Paul, and Benedikt Goderis. 2007. “Commod-ity Prices, Growth and the Natural ResourceCurse: Reconciling a Conundrum.” Centre for theStudy of African Economics Working PaperSeries/2007-15. Oxford University, Oxford, U.K.

Collier, Paul, and Jan Willem Gunning. 1994. TradeShocks: Consequences and Policy Responses inDeveloping Countries. San Francisco: Interna-tional Center for Economic Growth.

Crain, Mark, and Julia Devlin. 2003. “Non-Renew-able Resource Funds: A Red Herring for FiscalStability?” Paper presented at the annual meetingof the American Political Science Association,August 27, Philadelphia, PA.

Cuddington, John. 1989. “Commodity Export Boomsin Developing Countries.” World Bank ResearchObserver 4 (2):143–65.

Dana, J. 2008, Personal communication, World Bank.


137

10363_Pg95_140:10363_Pg95_140 11/29/08 7:08 AM Page 137

Dana, J., C. J. Gilbert, and E. Shim. 2006. “HedgingGrain Price Risk in the SADC: Case Studies ofMalawi and Zambia.” Food Policy 31: 357–71.

Datt, Gaurav, and Martin Ravallion. 1998. “Farm Pro-ductivity and Rural Poverty in India.” Journal ofDevelopment Studies 34 (4): 58–62.

Davis, Jeffrey M., Rolando Ossowski, James Daniel,and Steven Barnett. 2001. “Stabilization and Sav-ings Funds for Nonrenewable Resources: Experi-ence and Policy Implications.” Occasional Paper205. International Monetary Fund, Washington,DC.

Deaton, Angus, and Ron Miller. 1995. “InternationalCommodity Prices, Macroeconomic Performance,and Politics in Sub-Saharan Africa.” PrincetonStudies in International Finance 79. PrincetonUniversity, Princeton, NJ.

De Hoyos, R. and R. Lessem. 2008. “Food Shares inConsumption: New Evidence Using Engel Curvesfor Developing Countries.” World Bank, Wash-ington, DC.

De Hoyos, R., and D. Medvedev. 2008. “Poverty Effectsof Higher Food Prices: A Global Perspective.”Background paper for Global EconomicProspects 2009. Washington, DC: WorldBank.

Delgado, Christopher L., Jane Hopkins, and Valerie A.Kelly. 1998. “Agricultural Growth Linkages inSub-Saharan Africa.” Research Report 107.International Food Policy Research Institute,Washington, DC.

Dessus, S., S. Herrera, and R. De Hoyos. 2008. “TheImpact of Food Inflation on Urban Povertyand Its Monetary Cost: Some Back-of-the-Envelope Calculations.” Forthcoming, AgriculturalEconomics.

Downs, Erica. 2007. “The Fact and Fiction of Sino-African Energy Relations.” China Security 3 (3):42–68.

Eifert, Benn, Alan Gelb, and Nils Borje Tallroth. 2002.“The Political Economy of Fiscal Policy and Eco-nomic Management in Oil-exporting Countries.”Policy Research Working Paper 2899. WorldBank, Washington, DC.

ESMAP. 2005. “The Impact of Higher Oil Prices onLow Income Countries and on the Poor.” UnitedNations Development Programme/World BankEnergy Sector Management Assistance Pro-gramme, Report 299, March.

———. 2006. “Coping with Higher Oil Prices,ESMAP Report 323/06, Washington DC.

FAO (Food and Agriculture Organization). 2006. TheState of Food and Agriculture 2006. Rome:FAO.

Fasano, Ugo. 2000. “Review of the Experience withOil Stabilization and Savings Funds in SelectedCountries.” Working Paper 00/12, InternationalMonetary Fund, Washington, DC.

Fatas, Antonio, and Ilian Mihov. 2005. “Policy Volatil-ity, Institutions and Economic Growth.” IN-SEAD. Fontainbleau, France.

Foster, J., J. Greer, and E. Thorbecke. 1984. “A Classof Decomposable Poverty Measures.” Economet-rica 52: 761–66.

Gaiha, R. 1993. “Design of Poverty Alleviation Strat-egy in Rural Areas.” FAO Economic and SocialDevelopment Paper 115. FAO, Rome.

Gelb, Alan. 1988. Oil Windfalls: Blessing or Curse?New York: Oxford University Press.

Gilbert, Christopher L. 2007. “International Com-modity Agreements.” Handbook on Trade andPolicy. Edward Elgar Publishing.

Griffith-Jones, Stephanie, and Jose Antonio Ocampo.2008. “Sovereign Wealth Funds: A DevelopingCountry Perspective.” Paper prepared for theworkshop on sovereign wealth funds organizedby the Andean Development Corporation, Febru-ary 18, London.

Grosh, Margaret, Carlo del Ninno, and Emil DanielTesliuc. 2008. “Guidance for Responses from theHuman Development Sectors to Rising Food andFuel Prices.” Human Development Network,World Bank, Washington, DC.

Gylfason, Thorvaldur. 2001. “Natural Resources, Ed-ucation and Economic Development.” EuropeanEconomic Review 45 (4–6): 847–59.

Gylfason, Thorvaldur, Tryggvi Thor Herbertsson, andGylfi Zoega. 1999. “A Mixed Blessing: NaturalResources and Economic Growth.” Macroeco-nomic Dynamics 3: 204–25.

Herrera, Santiago, Tala Khartabil, Gaobo Pang, andStefano Paternostro. 2005. “Preliminary Esti-mates of the Financing, Growth, and PovertyImplications of the Oil Shock on DevelopingNations.” World Bank, Washington, DC.

Herrera, Santiago, and Gaobo Pang. 2006. “Update ofthe Financing, Growth, and Poverty Implicationsof the Recent Oil Price Shock on DevelopingCountries.” World Bank, Washington, DC.

Humphreys, Macartan, and Martin E. Standbu. 2004.“The Political Economy of Natural ResourceFunds.” In Escaping the Resource Curse, ed.Humphreys, Sachs, and Stiglitz. New York:Columbia University.

IEA. 2002. World Energy Outlook. Paris: Interna-tional Energy Agency.

IMF (International Monetary Fund). 2007. “The Roleof Fiscal Institutions in Managing the Oil


138

10363_Pg95_140:10363_Pg95_140 11/29/08 7:08 AM Page 138

Revenue Boom.” Fiscal Affairs Department, IMF,Washington, D.C.

———. 2008a. “Food and Fuel Prices: Recent Devel-opments, Macroeconomic Impacts, and PolicyResponses.” IMF, Washington, DC (June).

———. 2008b. World Economic Outlook. Washing-ton, DC: IMF.

Ivanic, Maros, and Will Martin. 2008. “Implicationsof Higher Global Food Prices for Poverty in Low-Income Countries.” Policy Research WorkingPaper 4594. World Bank, Washington, DC.

Kaufmann, D., A. Kraay, and M. Mastruzzi. 2007.Governance Matters VI: Aggregate and Individ-ual Governance Indicators for 1996–2006.WorldBank Policy Research Working Paper 4280,Washington, DC.

Kojima, Masami, and Robert Bacon. 2006. “PhasingOut Subsidies: Recent Experiences with Fuel inDeveloping Countries.” World Bank, Washington,DC.

Koren, Miklos, and Silvana Tenreyro. 2003. “Diversi-fication and Development.” Working Paper 3-3.Federal Reserve Bank of Boston.

Kpodar, Kangni. 2006. “Distributional Effects of OilPrice Changes: Evidence from Mali.” WorkingPaper 06/91, International Monetary Fund,Washington, DC.

Landes, David. 1999. The Wealth and Poverty ofNations: Why Some Are Rich and Some So Poor.New York: W. W. Norton and Company.

Lane, Philip, and Aaron Tornell. 1999. “The VoracityEffect.” American Economic Review 89:22–46.

Larson, Donald F., Panos Varangis, and Nanae Yabuki.1998. “Commodity Risk Management andDevelopment.” Policy Research Working Paper1963. World Bank, Washington, DC.

Lederman, Daniel, and William Maloney. 2007.“Trade Structure and Growth.” In Natural Re-sources: Neither Curse nor Destiny, ed. DanielLederman and William Maloney. Washington,DC: World Bank.

Leite, Carlos A., and Jens Weidmann. 1999. “DoesMother Nature Corrupt? Natural Resources,Corruption, and Economic Growth.” IMF Work-ing Paper 99/85. IMF, Washington, DC.

Lyman, Princeton. 2005. “China’s Rising Role inAfrica.” Council on Foreign Relations, NewYork. ww.cfr.org/publication/8436.

Manzano, Ozmel, and Roberto Rigobon. 2007. “Re-source Curse or Debt Overhang?” In Natural Re-sources: Neither Curse nor Destiny, ed. DanielLederman and William F. Maloney. Washington,DC: World Bank.

Mati, Amine. 2008. “Managing Surging Oil Prices inthe Developing World.” IMF Survey Magazine:Policy March 20, 2008.

McDonald, Scott, and Melt van Schoor. 2005. “A Com-putable General Equilibrium (CGE) Analysis of theImpact of an Oil Price Increase in South Africa.”PROVIDE Working Paper No. 1, The WesternCape Department of Agriculture, Elsenburg.

Mehlum, Halvor, Karl Moene, and Ragnar Torvik.2006. “Institutions and the Resource Curse.”Economic Journal 116: 1–20.

Murshed, S. Mansoob. 2004 “When Does NaturalResource Abundance Lead to a Resource Curse?”Environmental Economics Program DiscussionPaper 04/01. International Institute for Environ-ment and Development, London.

Ocampo, Jose Antonio. 2003. “Developing Countries’Anti-Cyclical Policies in a Globalized World.” InDevelopment Economics and StructuralistMacroeconomics: Essays in Honour of LanceTaylor, ed. Amitava Dutt and Jaime Ros. Alder-shot, UK: Edward Elgar.

OECD (Organisation for Economic Co-operation andDevelopment). 2005. “The Development Effec-tiveness of Food Aid: Does Tying Matter?”OECD, Paris.

Prebisch, Raul. 1950. The Economic Development ofLatin America and its Principal Problems. NewYork: United Nations.

Ramey, Garey, and Valerie Ramey. 1995. “Cross-Country Evidence on the Link between Volatilityand Growth.” American Economic Review 85:1138–51.

Ravallion, Martin, Shaohua Chen, and Prem San-graula. 2007. “New Evidence on the Urbaniza-tion of Global Poverty.” Policy Research WorkingPaper 4199. World Bank, Washington, DC.

Revenga, A., Q. Wodon, and H. Zaman. 2008. “PolicyResponses to Rising Food Prices.” World Bank,Washington, DC.

Sachs, Jeffrey, and Andrew Warner. 1995. “NaturalResource Abundance and Economic Growth.”NBER Working Paper 5398, National Bureau ofEconomic Research, Cambridge, MA.

———. 2001. “Natural Resources and EconomicDevelopment: The Curse of Natural Resources.”European Economic Review 45: 827–38.

Seshan, Ganesh, and Dina Umali-Deininger. 2007“Agriculture and Import Liberalization andHousehold Welfare in Sri Lanka.” Mimeographed.

Shabsigh, Ghiath, and Nadeem Ilahi. 2007. “Lookingbeyond the Fiscal: Do Oil Funds Bring Macroeco-nomic Stability?” Working Paper 07/96. Interna-tional Monetary Fund. Washington, DC.


139

10363_Pg95_140:10363_Pg95_140 11/29/08 7:08 AM Page 139

Singer, H. 1950. “The Distribution of Gains betweenInvesting and Borrowing Countries.” AmericanEconomic Review 40: 473–85.

Steenblick, Ronald. 2007. “Biofuels—At What Cost?Government Support for Ethanol and Biodiesel inSelected OECD Countries.” Global SubsidiesInitiative, International Institute for SustainableDevelopment, Geneva.

Torvik, Ragnar. 2001. “Learning by Doing and theDutch Disease.” European Economic Review 45:285–306.

———. 2002. “Natural Resources, Rent Seeking, andWelfare.” Journal of Development Economics 67(2): 455–70.

Turnovsky, Stephen J., and Pradip Chattopadhyay. 1998.“Volatility and Growth in Developing Countries:Some Numerical Results and Empirical Evidence.”University of Washington, Seattle.

UNCTAD (UN Conference on Trade and Develop-ment). 2006. “Boosting Africa’s Growth throughRe-injecting Surplus Oil Revenue: An Alternativeto the Traditional Advice to Save and Stabilize.”UNCTAD/DITC/COM/2006/10. UNCTAD, NewYork.

———. 2007. World Investment Report: Transna-tional Corporations, Extractive Industries, andDevelopment. New York: United Nations Con-ference for Trade and Development.

UNDP. 2007. “Overcoming Vulnerability to Rising OilPrices: Options for Asia and the Pacific.” UnitedNations Development Programme, RegionalCenter in Bangkok.

USDA. 2007. “Serbia Grain and Feed Update.” For-eign Agricultural Service GAIN Report No.RB7009, USDA.

Van der Ploeg, Frederick, and Steven Poelhekke. 2007.“Volatility, Financial Development, and theNatural Resource Curse.” Working Paper 2007/36.European University Institute, Florence, Italy.

Wick, Katharina, and Erwin Bulte. 2006. “ContestingResources: Rent Seeking, Conflict, and theNatural Resource Curse.” Public Choice 128(3–4): 457–76.

World Bank. 2003. “Iran: Medium Term Frameworkfor Transition, Converting Oil Wealth to Devel-opment.” Economic Report No. 25848-IRN.World Bank, Washington, DC.

———. 2004. Global Economic Prospects 2004.Washington, DC: World Bank.

———. 2005a. “Egypt: Toward a More Effective So-cial Policy: Subsidies and Social Safety Net.” Re-port 33550-EG. Social and Economic Develop-ment Group, Middle East and North AfricaRegion, World Bank, Washington, DC.

———. 2005b. Extractive Industries and SustainableDevelopment, An Evaluation of World BankGroup Experience. Washington, DC: WorldBank.

———. 2007. World Development Report 2008, Agri-culture for Development. Washington, DC:World Bank.

———. 2008a. “Rising Food and Fuel Prices: Ad-dressing the Risks to Human Capital.” HumanDevelopment Network, World Bank, Washington,DC.

———. 2008b. “Multiple Shocks, Poverty and Coun-try Vulnerability.” Presentation for the October2008 Annual Meetings, Washington, DC.


140

10363_Pg95_140:10363_Pg95_140 11/29/08 7:08 AM Page 140

141

AppendixRegional Economic Prospects

East Asia and the PacificRecent Developments

Substantial headwinds buffeted the economiesof East Asia and the Pacific during 2008,

causing GDP growth to slow sharply, from the10.5 percent pace of 2007 to 8.5 percent inthe year. The surge and relapse of crude oiland non-energy commodity prices affected alarge and diverse set of countries in the region,from the hydrocarbon-exporting countries ofIndonesia, Malaysia, Papua New Guinea, andVietnam, to food and agricultural raw materi-als exporters, Thailand, the Philippines, andagain Indonesia and Malaysia. The fall tonegative ground in U.S. and Japanese importdemand—under way for more than two yearsin the case of the United States—began to takea toll on the region’s export growth and todampen the earlier buoyancy of intra-regiontrade.1

What began in August 2007 as financial dif-ficulties in the United States tied to subprimemortgage-based securities had turned into aglobal financial crisis as of October 2008, rais-ing risk perceptions for several economies inEast Asia. Equity markets were hard hit,spreads on international sovereign- and espe-cially corporate debt increased sharply, ex-change rates depreciated rapidly, and grosscapital flows to the region fell by half duringthe first 9 months of 2008. Slower investmentgrowth in East Asia is now expected to spillover into still weaker production, employment,household spending, and GDP growth.

Growth outturns were fairly diverse acrossthe region in 2008. China registered a dimin-ished 9.4 percent advance, down from 11.9percent during 2007, on a slowdown in in-vestment and smaller positive contributions togrowth from net exports. The larger membersof the Association of South Eastern AsianNations (ASEAN)—Indonesia, Malaysia, andThailand—grew 5.2 percent in the year, downfrom 6.1 percent during 2007. Growth inVietnam dropped by 2 full percentage pointsto 6.5, in part as oil and non-energy com-modities prices slumped, while a group ofsmaller economies saw a pick-up in growth to5.1 percent from 3.7 percent, on the back ofrecovery in Fiji and continued strong growthin Papua New Guinea, powered by oil exports(table A1).

Even before the financial crisis intensified,there were signs of slowing growth. In China,GDP in the third quarter of 2008 eased to again of 9 percent (year-over-year) from 11.2percent in the final quarter of 2007, marking afifth consecutive quarter of slowing growth(figure A1). Thailand and Malaysia witnesseda larger falloff, with Thailand dropping to 2.9percent in the second quarter from 7.1 (saar) inthe fourth quarter of 2007, and Malaysiafalling to 4.2 percent from 6.7 percent, onsofter exports and private consumption. Incontrast, growth in Indonesia accelerated,boosted by public spending financed from in-creases in windfall revenues thanks to highprices for hydrocarbons, fats, and oils.

10363_Pg141-180:10363_Pg141-180 11/29/08 7:13 AM Page 141

umes are expected to decline from 8.3 percentin 2008 to 2.6 percent; investment to ease to7 percent (still relatively strong due to devel-opments in China), and net trade to contributeno impetus to regional growth for the firsttime in some years.

Commodity prices plummet;export-market growth contractsEast Asia (excluding China), along with theLatin America region, has benefitted from highfood and fuel prices from 2005 through mid-2008. During this period, terms of tradeimproved by a cumulative 10.3 percent in Viet-nam, 4 percent in Indonesia, and 4.8 percent inMalaysia. In contrast, the terms of trademoved against China by a substantial 11.4 per-cent, with little effect, however, on the currentaccount surplus. The steep decline in commod-ity prices since mid-2008 should benefit Chinaand other oil importers in the region, helpingto improve East Asia’s aggregate terms of trade


142

4

2

6

8

10

12

GDP growth (percentage change, saar)

Figure A1 GDP growth eases in several EastAsian economies

Source: World Bank and national agencies.

a. China 5 year-on-year.

Chinaa Thailand Malaysia Indonesia

Q4, 2007 Q1, 2008 Q2, 2008 Q3, 2008

Regional GDP is projected to slow to 6.7percent in 2009, the weakest since the dot.comrecession of 2001, and prior to that, the EastAsia crisis of 1997–98. Regional export vol-

Table A1 East Asia and Pacific forecast summary(annual percent change unless indicated otherwise)

1991–2000a 2005 2006 2007 2008e 2009f 2010f

GDP at market prices (2000 US$)b 8.4 9.1 10.1 10.5 8.5 6.7 7.8GDP per capita (units in US$) 7.1 8.2 9.2 9.7 7.6 5.9 7.0PPP GDPc — 9.1 10.0 10.5 8.4 6.7 7.8

Private consumption 7.3 7.5 2.6 3.4 5.6 6.7 7.9Public consumption 9.0 10.9 9.5 11.8 13.0 13.4 10.4Fixed investment 10.3 12.6 12.6 12.9 10.5 6.9 8.4Exports, GNFSd 11.7 18.5 18.6 15.4 8.3 2.6 9.7Imports, GNFSd 11.2 11.0 11.6 10.9 10.8 3.4 11.7

Net exports, contribution to growth 0.3 4.1 4.6 3.8 0.2 0.0 0.5Current account balance/GDP (%) 0.1 5.8 8.6 10.5 9.0 8.7 7.7GDP deflator (median, LCU) 6.7 6.5 5.8 4.0 7.5 6.6 4.9Fiscal balance/GDP (%) �0.7 �1.1 �0.6 0.2 �0.9 �1.4 �1.5

Memo items: GDPEast Asia excluding China 4.8 5.4 5.7 6.2 5.3 4.0 5.3China 10.4 10.4 11.6 11.9 9.4 7.5 8.5Indonesia 4.2 5.7 5.5 6.3 6.0 4.4 6.0Thailand 4.5 4.5 5.1 4.8 4.6 3.6 5.0

Source: World Bank.

Notes: — � not available.a. Growth rates over intervals are compound averages; growth contributions, ratios, and the GDP deflator are averages.b. GDP measured in constant 2000 U.S. dollars.c. GDP measured at PPP exchange rates.d. Exports and imports of goods and nonfactor services.e. Estimate.f. Forecast.

10363_Pg141-180:10363_Pg141-180 11/29/08 7:13 AM Page 142

by 3.5 percent in 2009, with China’s pickingup 5.5 percent.

Sharply higher food and fuel prices andoverheating in several economies acceleratedinflation in the region, from a median 5.7 per-cent increase during 2007 to 11.9 percent byJuly 2008 (year-over-year). September figures(8.2 percent) suggest that favorable inflationresponses are coming in step with the falloff incommodity prices and improved terms oftrade since mid-2008. Headline consumerprice inflation has eased substantially inChina, for example, from a peak of 8.5 per-cent in April to 4 percent by October 2008;but Indonesia and the Philippines continue towitness building price pressures, stoked in theformer by still strong consumer demand.

The spread of technical recession from theUnited States to Japan and the Euro Area dur-ing the second half of 2008 has begun to makea dent in export performance for the region,with China’s outbound shipments (in dollarterms) easing below 20 percent growth (year-over-year) in October from the 30 percent paceof early 2007 (figure A2). Growth of exportsfrom Hong Kong, China, reflecting in largepart transshipments from the mainland, havehalved to 5 percent. And the falloff in exportperformance is particularly acute in Singaporeand Taiwan, China, where exports are now de-clining, affected in particular by a sharp drop

in demand for high-tech products. Exportgrowth is also slowing in Malaysia and Thai-land, which are experiencing sluggish manu-factures shipments as well as the effects ofcommodity price declines on the dollar valueof oil and agricultural exports. As recessiondeepens across the countries of the Organisa-tion for Economic Co-operation and Devel-opment (OECD) during the course of 2009,East Asian export volumes are likely to fallsharply—to negative territory for manycountries—with China seeing a modest ad-vance of some 4.2 percent, down from the10.1 percent gain of 2008.

Ripples of the financial crisis arereaching East AsiaThe region was spared significant fallout dur-ing the early stages of the financial crisis in2007, because, outside of China, holdings ofsecurities backed by subprime U.S. mortgageswere quite small. But with the intensificationof the crisis, effects within the region arespreading. A sharp increase in risk aversion atthe global level, plus a process of deleveragingby firms and banks that have suffered largelosses in both high-income and developingcountries, resulted in a heavy sell-off of global,including East Asian, equities. The benchmarkMSCI Asia-Pacific Index plummeted by a cu-mulative 50 percent from January throughOctober 2008, while China’s ‘B’ share marketin Shanghai is off a full 75 percent. The pro-ceeds of these sales have been converted out oflocal currencies, resulting in a sharp deprecia-tion for many regional currencies against boththe dollar and the yen (figure A3). The Philip-pine peso, for example, has given up some18 percent against the dollar over 2008 todate and 30 percent versus the yen. These de-velopments have sharply increased the cost ofcapital for regional firms and escalated thelocal currency cost of international debt ser-vicing, both factors likely to dampen privateinvestment outlays in the coming months.

In international debt markets, sovereignspreads for East Asia jumped by some 610basis points since the spring of 2008, reaching

R E G I O N A L E C O N O M I C P R O S P E C T S

143

Export values (US$, percentage change, 3-month movingaverage year-on-year)

Source: Haver Analytics.

0

5

10

15

20

25

30

35

Figure A2 Export growth in East Asia turnsdown on falling OECD demand

Jan.2007

May2007

May2008

Sep.2008

Sep.2007

Jan.2008

Hong Kong,ChinaMalaysia

China

Thailand

10363_Pg141-180:10363_Pg141-180 11/29/08 7:13 AM Page 143

825 basis points as of late October, well abovethe high of 450 basis points at the peak of theEast Asian crisis in 1998. But as conditions ininternational markets began to unfreeze, andmore and more countries announced fiscalstimulus packages to underpin theireconomies, spreads narrowed once more to560 basis points during the first week of No-vember. As of November 7, 2008, spreadswere up by a modest 50 basis points forChina, 300 points for Malaysia, 250 points inthe Philippines, but a more-substantial 570points in Indonesia. Spreads for corporate bor-rowers have increased by far more, and thosefor noninvestment grade corporations—themajority of private sector issuance in theregion—have skyrocketed (see chapter 1). Forseveral countries in East Asia, the hike inspreads has become problematic, effectivelyshutting down bond issuance as a cost-effectivemeans of finance.

Given the process of deleveraging nowunder way among high-income financial insti-tutions, the retreat from regional equity mar-kets should be viewed together with a sub-stantial falloff in capital flows to the regionover the course of 2008. Gross capital flows toEast Asia and the Pacific, not including for-eign direct investment (FDI), dropped from$100 billion to $60 billion from Januarythrough August 2008, down 40 percent from

the same period in 2007. The bulk of thefalloff may be traced to sharp contractions inthe issuance of initial public equity offerings(IPOs), largely from China, which were off65 percent, from $56 billion to $19 billion inthe year, in line with the deterioration of con-ditions in international markets. But bankingflows also dropped 12.5 percent to $35 bil-lion, and bond issuance eased by 7.5 percentto $7 billion (figure A4).

In contrast, FDI flows to the region surgedby some two-thirds to a fresh record $175 bil-lion in 2007, with FDI to China picking up75 percent to near $140 billion, and with ad-vances of 40 percent in Malaysia to $8.5 bil-lion. Estimates for 2008 suggest a modest in-crease in overall FDI flows, showing someresilience in the face of the crisis. Measured bythe extent to which sovereign spreads haveincreased, equity markets declined and ex-change rates depreciated since September 15,together with the sharp falloff in capital flowsin the last year, East Asian economies hit hard-est by the crisis to date include Fiji, Indonesia,the Philippines, Thailand, and Vanuatu.

Difficult policy decisionsThe general stance of policy in the region ismoving from a tightening posture—initiated


144

0

10

20

30

Bond issuance Equity issuance Bank borrowing

40

50

60

US$ billions

Figure A4 Gross capital inflows to East Asiacontracted 40 percent in 2008

Source: World Bank.

Jan.–Aug. 2007 Jan.–Aug. 2008

Local currency unit/US$ index (Jan. 01, 2008 5 100)

Source: Thomson/Datastream.

Note: Increase implies weaker local currency.

80

90

100

110

120

Figure A3 Exchange rates decline sharplyas carry trades unwind

Jan. 12008

Mar. 12008

Jul. 12008

Sep. 12008

Japanese yenThai baht

Philippine peso

Nov. 12008

May 12008

Malaysian ringgit

10363_Pg141-180:10363_Pg141-180 11/29/08 7:13 AM Page 144

to deal with rising inflation—to a more re-laxed one; large efforts have been made to freeup liquidity to support banking systems fromthe contagion of financial stress from the high-income countries.Moreover, measures to under-pin growth at a time of downside risks havealso come to the fore. In China, bank rateswere raised to 7.47 percent in January to helpdampen inflation, then reduced to 6.66 per-cent on October 28, as the risk of financialdisruptions and loss of liquidity in the bankingsystem increased in importance. Furtheractions undertaken by China to prop up eco-nomic activity have included the announce-ment of a massive $586 billion stimulus pro-gram to focus on infrastructure, housing andincome support, and increasing export tax re-bates. In contrast, the Philippines first reducedpolicy rates to 7 percent to stimulate growth,then raised rates in four steps of 25 basispoints to 8 percent to help stem a ramp-up ininflation.

Medium-term outlookAs always, developments in China will play akey role in shaping the region’s growth profilethrough 2010. China’s buffers against thefinancial crisis are impressive: $1.6 trillion ininternational reserves; a fiscal surplus of 1 per-cent of GDP; and a current account surplus ofalmost $400 billion or 10.4 percent of GDP in2008. Policy efforts to underpin exports andhousehold spending—to maintain GDPgrowth at rates near 9 percent in 2009 andforward—should carry positive effects. But anextreme falloff in export volume growth to4.2 percent, on the back of recession in high-income countries, and slippage in investmentto 8 percent in the year is projected to slowGDP growth to 7.5 percent in 2009, from the9.4 percent pace of 2008 (table A2). A step-down in China’s import growth to 6.5 percentwill dampen the momentum of intraregionaltrade, causing exports for East Asia in aggre-gate to slide to 2.6 percent from the 8.3 per-cent advance of 2008.

Growth among the larger ASEAN coun-tries is expected to ease to 3.8 percent from


145

5.2 percent in 2008, as export volumes declineby a percentage point, and the squeeze oncommercial credit hits fixed investment, drop-ping it from growth of 8.8 percent in 2008to 3.6 percent. Lower commodity prices andweaker import demand are projected to im-prove the group’s current account surplus to$58 billion in 2009 from $55 billion in 2008.Among smaller countries, including Fiji, theLao People’s Democratic Republic, and PapuaNew Guinea, output growth is projected toslow to 3.4 percent from 5.1 percent in 2008,on the back of a sharp 5 percent decline inexports.

Recovery in regional growth during 2010is anticipated to be fairly swift. The downturnin investment should be relatively short-lived,as credit and capital flows begin to thaw, andexpectations for stronger domestic and exter-nal demand underpins a revival in regionalcapital spending to 8.4 percent (see table A1).Export growth is expected to rebound to9.7 percent in the region (to 10.7 percent forChina), as OECD and regional demand re-turn to positive territory. Moreover, a moder-ation in East Asian inflation, as the surge incommodity prices passes out of calculation,will help to restore purchasing power tohouseholds and support a renewal in spend-ing. Inflation as measured by the median GDPdeflator for the region is expected to declinefrom 7.5 percent in 2008 to 4.9 percent by2010.

Under these conditions, aggregate GDP forthe region is anticipated to grow 7.8 percent in2010, underpinned by 8.5 percent growth inChina. For East Asia excluding China, GDP isexpected to grow 5.3 percent in 2010, up from4 percent. Current account positions are pro-jected to vary across countries, easing to8.8 percent of GDP in China, to 4.4 percent inthe larger ASEAN members to minus 5.7 per-cent among the smaller countries of the region.

RisksThe favorable external environment that cameto benefit the region in the past five yearshas shifted dramatically to the downside. Given

10363_Pg141-180:10363_Pg141-180 11/29/08 7:13 AM Page 145

the sensitivity of regional GDP growth totrade, the possibility of a more extended pe-riod of recession, or only sluggish activityamong the OECD countries, represents one ofthe primary risks to growth in East Asia. Sucha scenario would be predicated on a more pro-longed period than projected for the financialsector in the high-income countries to redresstheir balance sheets and for lending to resume.A second area of risk relates to continued

adverse developments in financial markets.Should sovereign and especially corporatespreads not retreat from current levels, theregion could face difficulty financing new in-vestment and sustaining current projects. As aresult, investment activity would continue tobe depressed and the recession deeper; in thatcase, the risk that a country in the region couldsuffer significant exchange rate pressure or abalance of payments crisis cannot be ruled out.


146

Table A2 East Asia and Pacific country forecasts(annual percent change unless indicated otherwise)

1991–2000a 2005 2006 2007 2008c 2009d 2010d

CambodiaGDP at market prices (2000 US$)b — 13.5 10.8 10.2 6.7 4.9 6.0Current account balance/GDP (%) — �5.7 �4.7 �6.0 �15.3 �11.2 �8.0ChinaGDP at market prices (2000 US$)b 10.4 10.4 11.6 11.9 9.4 7.5 8.5Current account bal/GDP (%) 1.5 7.2 9.9 12.2 10.7 10.2 8.8FijiGDP at market prices (2000 US$)b 2.1 0.7 3.6 �6.6 1.7 �1.0 3.0Current account bal/GDP (%) �3.7 �13.3 �24.1 �15.6 �22.6 �23.6 �22.5IndonesiaGDP at market prices (2000 US$)b 4.2 5.7 5.5 6.3 6.0 4.4 6.0Current account bal/GDP (%) �0.4 0.1 2.9 2.6 0.8 �0.1 �0.4Lao PDRGDP at market prices (2000 US$)b 6.3 7.1 8.1 7.9 6.8 4.5 7.5Current account bal/GDP (%) �12.5 �19.3 �9.7 �16.4 �16.0 �17.2 �16.8MalaysiaGDP at market prices (2000 US$)b 7.1 5.0 5.8 6.4 5.5 3.7 4.6Current account bal/GDP (%) �0.4 14.6 17.2 16.7 22.0 17.5 16.4Papua New GuineaGDP at market prices (2000 US$)b 4.8 3.3 2.6 6.2 5.5 4.5 5.5Current account bal/GDP (%) 2.4 8.5 17.5 21.6 24.1 9.8 7.7PhilippinesGDP at market prices (2000 US$)b 3.0 4.9 5.4 7.2 4.0 3.0 4.1Current account bal/GDP (%) �3.1 2.0 4.0 4.1 0.4 3.6 3.5ThailandGDP at market prices (2000 US$)b 4.5 4.5 5.1 4.8 4.6 3.6 5.0Current account bal/GDP (%) �1.2 �4.3 1.1 6.3 2.2 5.2 5.0VanuatuGDP at market prices (2000 US$)b 4.1 6.5 7.2 5.0 4.5 3.0 5.2Current account bal/GDP (%) �8.2 �14.3 �8.1 �9.8 �14.4 �7.1 �5.3VietnamGDP at market prices (2000 US$)b 7.6 8.4 8.2 8.5 6.5 6.5 7.5Current account bal/GDP (%) �5.1 0.2 1.2 �0.2 �8.5 �3.4 �2.6

Source: World Bank.

Note: — � not available. Growth and current account figures presented here are World Bank projections and may differ fromtargets contained in other World Bank documents. American Samoa; Micronesia; Federated States of Kiribati; Marshall Islands;Myanmar; Mongolia; N. Mariana Islands; Palau; Korea, Dem. Rep. Of; Solomon Islands; Timor-Leste; and Tonga are notforecast because of data limitations.

a. Growth rates over intervals are compound averages; growth contributions, ratios, and the GDP deflator are averages.b. GDP measured in constant 2000 U.S. dollars.c. Estimate.d. Forecast.

10363_Pg141-180:10363_Pg141-180 11/29/08 7:13 AM Page 146


147

Europe and Central AsiaRecent developments

The rapid GDP growth in Europe and Cen-tral Asia of the past 20 years, which largely

reflected the enormous reform efforts under-taken by countries in the region (includingthose associated with accession to the Euro-pean Union), eased in 2008 and is expected togive way to a sharp slowdown in 2009. Theglobal financial crisis is expected to cut heavilyinto capital inflows and investment in the re-gion. Moreover, a number of countries are par-ticularly vulnerable because of high current ac-count deficits that in many instances have beenreliant on short-term capital inflows for theirfinancing.

Regional GDP growth fell almost 2 per-centage points to 5.3 percent in 2008, moder-ating from 7.1 percent in 2007, tied largely toa sharp falloff in growth during the secondhalf of the year. The financial crisis and asso-ciated growth slowdown outside of the regionis eroding macroeconomic buffers, includinginternational reserves, and is placing bankingsectors in several countries (notably, Hungary,the Russian Federation, and Ukraine) undersevere stress. Even economies with little directexposure to troubled U.S. financial assets arelikely to be hit hard by direct and indirectspillover effects from the financial crisis.

The region exhibits diverse performanceGDP growth slowed across the region during2008. The group of Central and EasternEuropean countries (CEE), (including Bulgaria,Poland, Romania, and the middle-incomeBaltic states but excluding Turkey), sawgrowth ease from 6.6 percent to 5.5 percent inthe year. Slowing demand in the Euro Areadampened export performance, while over-heating in several countries required a mix offiscal and monetary tightening to stem infla-tionary pressures. Growth in the Baltic stateshas come close to a standstill, with Estoniaand Latvia falling into recession and Lithuaniafaring little better. The global financial crisis

disrupted Hungary’s slow recovery of domesticdemand and led the country to accept an emer-gency €15 billion Standby Arrangement withthe International Monetary Fund. Growth inTurkey eased from 4.6 percent to 3 percent in2008, as financial and exchange rate pressurespicked up in the second half of the year(table A3).

GDP among the Commonwealth of Inde-pendent States (CIS) slid from the robust 8.6percent registered in 2007—grounded in a surgein activity across hydrocarbon exporters—to6.4 percent in 2008, reflecting reduced in-comes as oil prices declined, and the effects ofthe banking crisis in Russia (growth in Russiaeased from 8.1 percent to 6.0 percent). Exclud-ing Russia and Kazakhstan (where growthslowed sharply from 8.5 to 4 percent), GDPdeclined less dramatically in the remainingCIS states, falling from 10.4 to 8.5 percent inthe year.

The commodity price surge of 2006through mid-2008 contributed directly tohigh inflation across almost all countries ofthe region. Most countries tightened mone-tary policy to stem second-round effectsfrom the initial price hikes, while substantialcurrency appreciation (against the dollar)helped to mitigate inflation pressures to adegree. Romania posted the highest interestrates in the European Union, while Turkeyscored the highest across all developing andadvanced economies in Europe. The globalfood crisis had not caused the serious socialtensions witnessed in other regions, becausealmost all countries in Europe and CentralAsia have more or less adequate social safetynets in place. The World Bank is currentlyhelping to finance seed purchases and nutri-tional programs for the Kyrgyz Republic,Moldova, and Tajikistan. And with threemajor grain exporters (Kazakhstan, Russia,and Ukraine) relaxing previously imposedexport bans amid the region’s best harvest ina decade, food prices are expected to moder-ate, helping to ease the earlier jump in head-line inflation.

10363_Pg141-180:10363_Pg141-180 11/29/08 7:13 AM Page 147

Intensification of global crisis beginsto exact tollThe sudden deepening of the financial crisis inthe United States during September andOctober, and the accompanying start ofdeleveraging across financial institutions world-wide, triggered a wave of sell-offs in emergingmarket assets across the globe. Widening sov-ereign spreads, sharp currency depreciation,and a halving of domestic equity prices havebeen witnessed across emerging markets. Themagnitude and extent of these developmentsin Europe and Central Asia are of concern(figure A5).

Recent spikes in sovereign spreads for anumber of countries in the region havedwarfed those witnessed in earlier periods offlare-up since the start of financial turmoil in2007. Except for Kazakhstan and Russia,where massive central bank intervention hastaken place, other regional currencies have


148

260

Russia

n

Fede

ratio

nPola

nd

250

240

230

220

210

20 0

60

120

180

240

300

360

Percent Basis points (inverted scale)

Figure A5 Deepening global financial crisisaffects Europe and Central Asia

Source: World Bank.

Ukrain

e

Turk

ey

Kazak

hsta

n

Bulgar

ia

Lithu

ania

Latvi

a

Roman

ia

% change in stock market inSept. and Oct., 2008, local currency

% change in exchange rate versusUS$ in Sept. and Oct., 2008, (-) depreciation

% change in capital flows, Jan.–Aug. 2008vs. Jan.–Aug. 2007

increase in JPMorgan sovereignbond spread in Sept. and Oct., 2008 (right axis)

Table A3 Europe and Central Asia forecast summary(annual percent change unless indicated otherwise)

1991–2000a 2005 2006 2007 2008e 2009f 2010f

GDP at market prices (2000 US$)b �1.1 6.4 7.5 7.1 5.3 2.7 5.0GDP per capita (units in US$) �1.3 6.3 7.4 7.0 5.3 2.7 5.0PPP GDPc �1.2 6.3 7.7 7.4 5.7 2.6 5.1

Private consumption 0.6 7.0 8.2 8.5 8.4 5.3 6.2Public consumption 0.0 3.6 5.2 5.5 4.9 3.3 4.0Fixed investment �7.0 11.0 14.9 15.4 10.0 �0.7 7.2Exports, GNFSd 0.3 5.6 8.0 7.8 9.4 5.4 10.1Imports, GNFSd �2.8 10.6 15.5 18.8 14.7 6.3 11.0

Net exports, contribution to growth 1.1 �2.0 �3.4 �5.5 �3.6 �1.2 �1.8Current account balance/GDP (%) �0.7 2.6 1.5 �0.6 �0.8 �4.1 �4.5GDP deflator (median, LCU) — 6.8 5.8 7.5 10.9 8.9 6.8Fiscal balance/GDP (%) �5.0 2.6 2.9 2.4 1.9 1.1 1.1

Memo items: GDPTransition countries 2.3 6.1 6.7 5.7 4.4 2.6 4.8Central and Eastern Europe 1.4 4.3 6.6 6.6 5.5 3.2 4.7Commonwealth of Independent States �4.3 6.8 8.4 8.6 6.4 2.9 5.2

Russian Federation �3.9 6.4 7.4 8.1 6.0 3.0 5.0Turkey 3.7 8.4 6.9 4.6 3.0 1.7 4.9Poland 3.8 3.6 6.2 6.6 5.4 4.0 4.7

Source: World Bank.Note: — � not available.a. Growth rates over intervals are compound averages; growth contributions, ratios, and the GDP deflator are averages.b. GDP measured in constant 2000 U.S. dollars.c. GDP measured at PPP exchange rates.d. Exports and imports of goods and nonfactor services.e. Estimate.f. Forecast.

10363_Pg141-180:10363_Pg141-180 11/29/08 7:13 AM Page 148

depreciated quite sharply, reversing almost allthe gains of the last two years. Moreover,gross capital inflows to the region (equityIPOs, bond issuance, and bank lending) de-clined to $123 billion from January throughAugust 2008, from $187 billion in the likeperiod of 2007, a drop of some 34 percent.These developments underscore the swiftspread of effects from the deterioration ininternational financial markets and point tomore difficult financing conditions ahead,with funding for fixed investment in theregion—a primary driver for growth—underparticular uncertainty.

Activity in Russia already showed signs ofslowing before fall 2008, when the financialcrisis entered a more intense phase. Industrialproduction over the first eight months of2008 declined by 2.3 points to 4.9 percent,compared with the same period in 2007, andgrowth in fixed capital investment almosthalved. Gross capital inflows did halve to$74 billion in the January–August period,compared with $150 billion for all of 2007.Moreover, the credit crunch appeared to bedraining domestic liquidity from the economyeither directly (given that Russia is Europe’sthird largest bank borrower) or indirectlythrough the interbank and corporate sectors.

The Russian stock market crisis forcedmultiple suspensions of trading, and the gov-ernment has taken all possible measures tomitigate growing financial and economic dif-ficulties. These include but are not limited tocutting banks’ reserve requirements and oilcompanies’ export duties several times; inject-ing liquidity (more than $200 billion in fed-eral budget fund deposits, subordinatedloans, and the like), increasing coverage of re-tail bank deposit insurance by 75 percent; in-tervening in the foreign exchange market, ev-idenced in a decline of more than $100 billionin reserves between August and October;committing an additional $50 billion of re-serves to solve refinancing difficulties inbanks and companies (estimated to hold $80billion-90 billion in debt service due in 2009);and using another $20 billion from its

national wealth fund to boost domestic stockmarkets directly.

As in Russia, Ukraine’s banks have reliedon foreign bank- and other loans to fund do-mestic lending. And about $1 billion to ser-vice foreign debts is due during the finalmonths of 2008. Facing rating agencies’downgrades, and massive withdrawals fromthe banking system during the first threeweeks of October (amounting to $3 billion orabout 4 percent of total deposits), the centralbank banned preterm withdrawals, injectedfurther liquidity, and imposed exchangecontrols. On the real side of the economy,Ukraine is starting to see decline in the metal-lurgy industry and in exports of these prod-ucts (which provide 40 percent of exportrevenues), as global production and metalprices cool. These negative developments haveprompted Ukraine to seek an IMF loan of$16.4 billion. Turkey’s second-quarter GDPdeteriorated sharply to 1.9 percent year-over-year from 6.7 percent a quarter earlier. Andgiven Turkey’s traditional reliance on short-term debt and external financing, the debtrollover situation is no better for Turkey thanfor Russia and Ukraine; Turkey holds morethan $280 billion of foreign debt, of whichone-sixth is short term.

Based on credit-default swap prices,Kazakhstan stands second in the globalleague of economies as riskiest for severebanking disruptions—after Iceland. The gov-ernment has $15 billion dollars ($10 billionof which from its oil fund) available to stabi-lize the banking situation. Many other coun-tries in Central and Eastern Europe carrysimilar vulnerabilities in terms of bankingexposures, external deficits, and reliance onforeign capital flows, and governments havereacted to address them while trying to reas-sure investors and depositors. In Bulgaria,Poland, and Romania, guarantees on individ-ual bank deposits have been raised in linewith EU levels; Hungary, the Slovak Repub-lic, and Slovenia have all enacted unlimitedgovernment guarantees on private bankdeposits.


149

10363_Pg141-180:10363_Pg141-180 11/29/08 7:13 AM Page 149

Medium-term outlookThe outlook for 2009 appears fairly soberingat this juncture. Slower growth in the region’smain trading partners, the EU, (and for theCIS countries) Russia and China, will limit ex-port opportunities. For example, the auto fab-rication and export–industry, which had beenperforming well in Turkey and some CEEcountries, will be put to a difficult test. De-clines in equity markets will tend to raise thecost of capital for domestic firms and coulddelay privatization plans. Moreover, in coun-tries where foreign banks have a dominantpresence, local subsidiaries may feel the pinchfrom headquarters in the high-income coun-tries, further escalating difficulties in domesticcredit markets and contributing to a slow-down in economic activity.

Table A3 shows that still-robust gains in in-vestment continued during 2008—an advanceof 10 percent for the region; Russia gained16 percent, other CIS countries grew capitalspending 14 percent, with the CEE countries up10.5 percent. But a flattening in domestic andforeign demand and much more difficult fi-nancing conditions are expected to cause realinvestment to stagnate in 2009, with related de-clines in orders, production, and employment.

Signs of the slowdown have already begunto emerge. In Russia, for example, Sberbankand Gazprom, the leading state bank and statecompany, both plan to cut back on workforceand investment; the third-largest steelmaker,Magnitogorsk, is reducing workforce levels by3,000; truck manufacturer KamAZ plans tocurtail production by 20 percent; and car-maker GAZ also foresees substantially less do-mestic and export demand. Russia’s GDP isanticipated to drop to 3 percent in 2009, fromthe 6 percent pace of 2008 (table A4). How-ever, financial support policies enacted to date,plus the substantial amount of internationalreserves held by the country, should help Rus-sia weather the depth of global crisis in 2009and rebound to growth of 5 percent by 2010.2

Deterioration in the external environment—and the fragile set of current conditions in alarge number of European and Central Asian

countries suggests the potential for a sharpslowdown in regional GDP growth to 2.7 per-cent in 2009 from the 5.3 percent advance of2008. But under assumptions that global creditmarkets begin to function once more by earlyto mid-2009, and that growth in OECD centersstarts to pull-up at the same time, regionalgrowth is anticipated to firm to 5 percent by2010. CIS countries outside of Russia are ex-pected to realize a rebound in exports and apick-up in consumer spending, as growth re-covers from 2.8 percent in 2009 to 5.7 percent.And gradual revival in Euro Area demand helpsCEE exports pick-up from 2.5 percent gains in2009 to 7.6 percent by 2010, supporting amove in GDP from 3.2 percent to 4.7 percent.Lower oil prices will help alleviate a portion ofthe current account burden in oil importingcountries, especially Turkey, and a large num-ber of Central European countries.

RisksIn the short term, the financial system will betested. In Russia, for example, the largestbanks have enjoyed generous government sup-port, but private and smaller banks may faceliquidity shortages and possibly large-scalewithdrawals should the situation worsen.Russia currently is home to 1,100 banks, ofwhich the 20 largest account for 70 percent ofhousehold deposits and corporate loans. Out-side Russia, the financial sector in a number ofcountries is dominated by banks fromWesternEurope, carrying the potential risk of conta-gion from difficulties being experienced bytheir home-country institutions.

In the medium-term, divergent perfor-mance in 2008 should not belie either thecommon factors underlying growth in Europeand Central Asia or the associated commonrisks. Recent growth has been supported bydomestic demand and enabled by easy accessto external financing in bank lending, bondissuance, and FDI, while net exports continueto offer a substantial drag on growth. Rapidcredit expansion and accommodative wagepolicies have been widespread, while domesticsaving is insufficient, while pro-cyclical fiscal


150

10363_Pg141-180:10363_Pg141-180 11/29/08 7:13 AM Page 150


151

Table A4 Europe and Central Asia country forecasts(annual percent change unless indicated otherwise)

1991–2000a 2005 2006 2007 2008c 2009d 2010d

AlbaniaGDP at market prices (2000 US$)b 1.4 5.5 5.0 6.0 6.0 5.0 5.5Current account balance/GDP (%) �5.6 �6.8 �7.3 �10.0 �11.2 �5.3 �4.7ArmeniaGDP at market prices (2000 US$)b �3.8 13.9 13.3 13.7 9.0 6.4 6.7Current account balance/GDP (%) �12.0 �1.1 �1.8 �6.2 �7.6 �4.3 �4.3AzerbaijanGDP at market prices (2000 US$)b �5.2 26.2 34.5 25.0 17.7 10.4 7.8Current account balance/GDP (%) �15.8 1.3 17.7 30.7 41.6 30.7 28.4BelarusGDP at market prices (2000 US$)b �1.2 9.4 9.9 8.2 9.2 5.0 5.8Current account balance/GDP (%) — 1.4 �4.1 �6.4 �5.5 �6.2 �6.4BulgariaGDP at market prices (2000 US$)b �1.7 6.2 6.3 6.2 6.0 2.4 6.0Current account balance/GDP (%) �2.3 �12.3 �15.7 �21.6 �24.3 �15.6 �13.6CroatiaGDP at market prices (2000 US$)b �1.5 4.3 4.8 5.6 3.5 2.3 5.1Current account balance/GDP (%) 1.0 �6.6 �7.6 �8.6 �9.9 �4.2 �3.2GeorgiaGDP at market prices (2000 US$)b �9.3 9.6 9.4 12.4 3.5 4.0 6.0Current account balance/GDP (%) — �11.9 �16.2 �21.5 �21.9 �20.7 �22.0KazakhstanGDP at market prices (2000 US$)b �3.6 9.7 10.7 8.5 4.0 1.9 6.2Current account balance/GDP (%) �2.1 �1.9 �2.2 �6.9 0.1 �7.0 �7.2Kyrgyz RepublicGDP at market prices (2000 US$)b �4.0 �0.2 2.7 8.2 6.6 4.2 5.6Current account balance/GDP (%) �10.6 �2.4 �10.6 �7.2 �10.6 �5.6 �2.4LithuaniaGDP at market prices (2000 US$)b �3.3 7.9 7.7 8.8 4.0 �0.3 2.0Current account balance/GDP (%) �5.8 �7.1 �10.7 �13.6 �13.9 �12.2 �10.9LatviaGDP at market prices (2000 US$)b �2.8 10.6 12.2 10.3 �0.8 �3.5 0.7Current account balance/GDP (%) �1.6 �12.4 �22.7 �22.8 �15.2 �10.5 �8.2MoldovaGDP at market prices (2000 US$)b �9.8 7.5 4.0 3.0 6.5 4.0 4.0Current account balance/GDP (%) — �8.3 �11.5 �15.8 �17.7 �4.4 �5.8Macedonia, FYRGDP at market prices (2000 US$)b �0.9 4.1 3.0 5.1 5.5 4.8 5.6Current account balance/GDP (%) — �1.4 �0.4 �3.4 �9.8 �4.4 �3.5PolandGDP at market prices (2000 US$)b 3.8 3.6 6.2 6.6 5.4 4.0 4.7Current account balance/GDP (%) �3.5 �1.2 �2.7 �3.8 �5.4 �6.2 �5.6RomaniaGDP at market prices (2000 US$)b �1.7 4.1 7.9 6.0 8.6 3.2 5.8Current account balance/GDP (%) �4.8 �8.7 �10.5 �13.7 �15.5 �8.6 �7.4Russian FederationGDP at market prices (2000 US$)b �3.9 6.4 7.4 8.1 6.0 3.0 5.0Current account balance/GDP (%) — 11.1 9.6 6.1 6.0 �3.4 �5.0TurkeyGDP at market prices (2000 US$)b 3.7 8.4 6.9 4.6 3.0 1.7 4.9Current account balance/GDP (%) �1.1 �4.7 �6.0 �5.7 �8.4 �3.9 �3.1

(continued)

10363_Pg141-180:10363_Pg141-180 11/29/08 7:13 AM Page 151

policy is underway in a number of countries,such as Belarus, Romania, Russia and Ukraine.

The potential for second-round inflationeffects remains a problem in the region. Thereversal in commodity prices since mid-2008has been reflected in a flattening or decline ininflation trends in at least 12 countries amidsome indications of a falloff in core inflation(figure A6).3 However, because domestic fac-tors such as government spending and strongwage growth also drive prices, inflation ex-pectations remain high, and the potential fora wage spiral is notable. Moreover, recentsharp currency declines and loosening ofmonetary policy, together with other aggres-sive measures to resist the economic down-turn, may drive up inflation and endangerfiscal positions, causing problems in thelonger run.

For many small and poorer countries thatrely on remittances as an important source offinancing, a downturn in neighboring coun-tries in Western Europe and the CIS impliesless in remittance flows from migrantsabroad, raising the need for financing fromother sources and potentially exacerbatingpoverty. This said, historical evidence showsremittances tend to be relatively resilient dur-ing a downturn, and should help cushion theslowdown.

Beyond the set of immediate challenges, alonger-term concern is the set of substantialbottlenecks to growth that have been reachedin infrastructure and labor markets in devel-oping countries in general, and in a large num-ber of European and Central Asian countriesin particular. Faster GDP growth in the futureis likely only if countries can take the neces-sary steps to improve the supply of essential


152

Table A4 (continued )(annual percent change unless indicated otherwise)

1991–2000a 2005 2006 2007 2008c 2009d 2010d

UkraineGDP at market prices (2000 US$)b �8.0 2.7 7.9 7.7 6.0 �3.0 4.4Current account balance/GDP (%) — 2.9 �1.5 �4.2 �6.5 �2.2 �1.3UzbekistanGDP at market prices (2000 US$)b �0.2 7.0 7.3 9.5 8.0 7.0 6.5Current account balance/GDP (%) �0.9 13.1 14.3 18.8 20.6 14.7 13.1

Source: World Bank.

Note: — � not available.Growth and Current Account figures presented here are World Bank projections and may differ from targets contained in otherBank documents. Bosnia and Herzegovina, Montenegro, Serbia, Tajikistan, Turkmenistan, and Yugoslavia are not forecastbecause of data limitations.


May2006

Sep.2006

Jan.2007

May2007

Sep.2007

Jan.2008

May2008

Sep.2008

0

4

8

12

16

CPI inflation (percentage change, year-on-year)

Figure A6 Core inflation is rising in severalcountries of Europe and Central Asia

Source: World Bank.

Jan.2006

Croatia Kazakhstan

Lithuania

Belarus

Poland Turkey

Latvia

RussianFederation

10363_Pg141-180:10363_Pg141-180 11/29/08 7:13 AM Page 152

utilities and upgrade transport, communica-tions, and other key infrastructure. Such im-provement, together with a diminishing of in-stitutional and structural inefficiencies couldhelp alleviate current constraints on growth inthe longer run.

Latin America andthe CaribbeanRecent developments

The global financial crisis has come to affectLatin America and the Caribbean after a

period of exceptional GDP growth. The regiongrew at an annual rate of 5.3 percent over2004–08, the strongest pace in the last threedecades. GDP gains were led by RepúblicaBolivariana de Venezuela, which advanced at a10.5 percent clip; Argentina at 8.4 percent;and Peru at 7.4 percent. Growth was alsobroad-based during this period, with theCaribbean countries gaining 5.9 percent annu-ally and Central American countries growing3.7 percent. The oil-exporting economies ofthe region saw GDP pick up at a 5.7 percentrate, and oil importers also grew briskly at5.3 percent. Only two countries grew slowerthan 3 percent per year over the period—Jamaica at 1.6 percent and Haiti at 1.4 per-cent. The last period of strong region-widegrowth occurred in 1991–94 when GDP ad-vanced 4.2 percent annually (figure A7).

A favorable external environment of highcommodity prices and strong import demandin high-income countries supported the re-gion’s recent growth performance. The role ofthe external environment is emphasized inIzquierdo and others 2008; Calvo and Talvi2007; and Österholm and Zettelmeyer 2007.However, the region has also made genuineprogress in maintaining independent mone-tary policy and increasing the credibility ofcentral banks, introducing exchange rate flex-ibility, deepening local currency debt markets,and providing supportive fiscal policy (WorldBank 2008c). Because of the improvements inpolicy and in the external environment, the re-gion is in better macroeconomic and fiscal

health than it was five years ago, or at the endof the previous growth spurt. But this healthystarting position will be seriously tested by theglobal crisis, which has already led to a with-drawal of funds from regional equity marketsby international investors, sharply depreciat-ing currencies and soaring sovereign- and cor-porate bond spreads. The U.S. and Europeanrecessions and the turnaround to decline inglobal commodity prices further darken theexternal environment for the region.

During 2008, Latin American GDP ad-vanced 4.4 percent, still robust, albeit downfrom the strong 5.7 percent pace of the previ-ous year. Buffers in the form of large levels ofreserves and current account surpluses miti-gated the impact of slowing exports to theUnited States to a degree. Latin America’s ex-ports lost momentum, however, growing only1.7 percent in 2008 compared with 5 percent in2007, while the region’s current account posi-tion dropped from a surplus of 0.5 percent ofGDP to a deficit of the same magnitude.

Output gains were quite differentiatedacross key countries and sub regions in LatinAmerica and the Caribbean. The outrightdecline in U.S. imports adversely affected


153

24

22

0

2

4

6

8

Percentage points

Figure A7 Contributions to GDP growth in Latin America and the Caribbean

Source: World Bank.

GDP

Investments

Net exports

Government consumption

Private consumption

1992

1993

1994

2000

2001

2002

2003

2004

2005

2006

2007

2008

10363_Pg141-180:10363_Pg141-180 11/29/08 7:13 AM Page 153

Mexico’s exports, sending them from growthof 3.3 percent in 2007 to contraction of0.9 percent in 2008, and contributing to aslowdown in GDP growth from 3.2 percent to2 percent over the period. Argentina’s growthperformance also slipped, from 8.7 percent in2007 to 6.6 percent in 2008, on the back ofslowing consumer spending and exports. Incontrast, Brazil maintained GDP gains at astill-robust 5.2 percent pace, with its economygrounded in stronger consumer outlays andinvestment, further supported by favorableterms-of-trade developments during the firsthalf of the year.

GDP growth eased in the Caribbean, de-clining from 6 percent in 2007 to 4.6 percentin 2008. The falloff was linked in part tohurricane damage but also to weaker exportsand a negative contribution of trade to GDP.And Central American GDP slowed by morethan a percentage point to 2.2 percent from

3.6 percent, largely because of a downshift inexports tied to the slowdown in U.S. demand(table A5).

Although not yet visible in GDP figures, alarge number of countries in the region are al-ready subject to adverse spillover effects of thefinancial crisis. Between September 15, whenLehman Brothers announced bankruptcy, andthe end of October, equity markets lost half oftheir dollar values; currencies, especially thoseof Brazil, Chile, and Mexico depreciated pre-cipitously against the dollar; the cost of cor-porate and government borrowing on interna-tional bond markets surged; investmentspending appeared to be slowing, and theavailability of trade finance tightened. Thesedevelopments have added to the region’s con-cerns regarding falling commodity prices (onthe upside of which food and oil exportersbenefited greatly), slowing remittance inflowsand rising inflation.


154

Table A5 Latin America and the Caribbean forecast summary(annual percent change unless indicated otherwise)

1991–2000a 2005 2006 2007 2008e 2009f 2010f

GDP at market prices (2000 US$)b 3.3 4.6 5.6 5.7 4.4 2.1 4.0GDP per capita (units in US$) 1.6 3.3 4.2 4.4 3.1 0.9 2.8PPP GDPc 4.2 4.6 5.5 5.7 4.4 2.2 4.1

Private consumption — 5.8 6.3 6.9 5.4 3.1 4.6Public consumption — 3.0 4.6 4.0 4.5 2.4 2.6Fixed investment 4.7 11.3 14.6 12.2 14.6 �4.1 8.8Exports, GNFSd 8.1 8.1 7.7 5.0 1.7 �2.1 2.4Imports, GNFSd 10.9 11.9 14.3 11.9 12.3 �3.9 6.9

Net exports, contribution to growth �0.4 �0.8 �1.6 �1.9 �2.9 0.6 �1.4Current account balance/GDP (%) �2.8 1.4 1.6 0.5 �0.6 �0.3 0.0GDP deflator (median, LCU) 11.3 5.7 8.0 7.5 10.2 6.7 5.5Fiscal balance/GDP (%) — 1.2 1.4 1.3 0.9 0.6 0.4

Memo items: GDPLatin America excluding Argentina 3.1 3.9 5.1 5.2 4.1 2.2 4.1Central America 3.6 3.0 5.1 3.6 2.2 1.4 3.3Caribbean 3.6 6.7 8.7 6.0 4.6 3.3 4.7

Brazil 2.5 2.9 3.8 5.4 5.2 2.8 4.6Mexico 3.5 2.8 4.9 3.2 2.0 1.1 3.1Argentina 4.5 9.2 8.5 8.7 6.6 1.5 4.0

Source: World Bank.

Note: — � not available.

a. Growth rates over intervals are compound averages; growth contributions, ratios, and the GDP deflator are averages.b. GDP measured in constant 2000 U.S. dollars.c. GDP measured at PPP exchange rates.d. Exports and imports of goods and nonfactor services.e. Estimate.f. Forecast.

10363_Pg141-180:10363_Pg141-180 11/29/08 7:13 AM Page 154

Credit conditions tighten, capital flowsplummetSovereign spreads, as measured by JPMorgan-Chase Emerging Market Price Index (EMBI),have increased rapidly since mid-Septemberthroughout the region, with the largest rise(over 1,000 basis points) for Argentina andRepública Bolivariana de Venezuela (figureA8). The corporate bond market is seeing asimilar trend, with soaring corporate spreads.Moreover, gross capital inflows to the regionhalved over January–August 2008 comparedwith the like period in 2007. Bond issuancedropped 46 percent to $18.5 billion; equityIPO issues virtually vanished in the hostile cli-mate of 2008 (down 75 percent); and bankborrowing dropped one-third to $36 billionover the year to date.

Tighter financing conditions and expecta-tions of weaker demand growth have led cor-porations and governments alike to reviewinvestment plans. The Republic of Korea’sHyundai, India’s Reliance, and Brazil’s Petro-bras have either announced or postponeddecisions on investment plans in Brazil.Petroleos de Venezuela has postponed severalrefining projects across the Caribbean andCentral America. The Mexican airline Aladia


155

filed for bankruptcy protection in Octoberbecause of financing difficulties. And Contro-ladora Comercial Mexicana, a large super-market chain, also filed for bankruptcy aftersustaining losses from derivatives trading.

As a result of falling equity markets andrepatriation of foreign funds to home curren-cies, many of the region’s currencies haveexperienced sharp depreciation since mid-September, a situation that runs the risk ofreigniting inflation, even as commodity pricesdecline. After several years of appreciation, theBrazilian real started to decline in early July(figure A9). The central banks of Argentina,Brazil, Chile, and Mexico sold dollars on thespot market during October to prevent theircurrencies from sliding further. Mexico offereddirect financing to commercial banks. Brazilrelaxed reserve requirements, eliminated taxeson foreign investment, authorized state-ownedbanks to buy stakes in financial institutions,and allowed the central bank to enter intocurrency swaps with other central banks.

Another consequence of tightened creditconditions has been vanishing export credit

0

100

200

300

400

500

700

600


Figure A8 Bond spreads have increasedsharply for many Latin American countries


Jan. 1, 2008 Nov. 10, 2008

Brazil

Chile

Colom

bia

Mex

icoPer

u

Argen

tina

Venez

uela,

R. B. d

e

14751710

80

90

100

110

120

130

Local currency unit/US$ index (Jan. 1, 2008 5 100)

Figure A9 Exchange rates in Latin America and the Caribbean fell sharply against the dollar in late 2008

Source: Thomson/Datastream.

Note: Increase = weaker local currency.

Jan. 12008

Feb. 202008

Apr. 102008

May 302008

Jul. 192008

Sep. 072008

Argentina

Chile

Mexico

Venezuela, R. B. de

Brazil

Colombia

Peru

10363_Pg141-180:10363_Pg141-180 11/29/08 7:13 AM Page 155

lines, which allow exporters to purchasegoods and services they need to support theirexport sales. Exporters now face a double hit,with slowing import demand in high-incomecountries on the one hand, and more expen-sive credit to support export operations on theother. Anecdotal evidence from Brazil suggeststhat the fall of Lehman Brothers precipitateda collapse in export credit in Brazil, leadingforeign investors and companies to repatriatebillions of dollars from Brazil. Shrinking ex-port credit could lead to difficult conditionsfor businesses that supply inputs to exporters,with ripple effects to the rest of the economy.

Remittance inflows are slowingWorker remittances are an important source ofincome for many Latin American countries. Theregion has sent 28.3 million workers abroad—5.1 percent of the region’s population—whosend back $60 billion, on average, to their homecountries (World Bank 2008a). The UnitedStates is the primary recipient of the region’semigrants, followed by Spain and Italy. In eightLatin American countries, remittances accountfor more than 10 percent of GDP. Mexico is thelargest recipient with $25 billion in receipts. Butthe slowdown in the U.S. housing market andthe resulting loss of construction jobs led to a4.2 percent decline in remittances to Mexicoover January–August 2008, compared with thesame period of 2007. No evidence of similarlarge-scale decline has yet come to light in otherregional economies.

Inflation remains high, notably in foodRapid economic growth in Latin America alsobrought with it a ramp-up in inflation, whichin 2008—abetted by the price surge for oil andfood traded internationally—was at its highestlevel in a decade. Food prices rose substan-tially faster than the overall consumer priceindex for most countries (World Bank 2008b),and fundamental changes in global food dy-namics appear to be under way. High energyprices, climate change, and rising biofuelproduction have driven the rise in food price

inflation. Although Latin America has thelargest surplus in food trade of all developingregions, food price inflation adversely affectsmost of the population, because householdsare net buyers of food. Poor people are alsoaffected disproportionately because theyspend a larger share of their income on food.

Two developments have occurred in recentmonths that are likely to help ease the pressureof food inflation. Central governments acrossthe region raised interest rates in the first halfof 2008 to stem inflationary pressures. Moreimportantly, commodity prices began toplummet after reaching historically high levelsin mid-2008. The latest inflation numbers(September 2008) for the region were gener-ally lower than their peak levels in the preced-ing months (figure A10).

Medium-term outlookGDP growth in the region is expected to falloff sharply to 2.1 percent in 2009 from4.4 percent in 2008, driven by a sharp declinein capital spending—from robust growth of


156

0

3

6

9

12

Consumer price inflation (percentage change year-on-year)

Figure A10 Inflation still high in LatinAmerica and the Caribbean despite sharpfalloff in food and fuel prices

Source: World Bank.

Jan.2006

May2006

Sep.2006

Jan.2007

May2007

Sep.2007

Jan.2008

May2008

Sep.2008

Argentina

Mexico

Latin America& Caribbean Chile

Brazil

Colombia

Peru

10363_Pg141-180:10363_Pg141-180 11/29/08 7:13 AM Page 156


157

14.6 percent in 2008 to a decline of 4 percent.Increasing cost of capital for business, chan-neled through falling domestic equity markets,widening spreads on international corporatebonds, and depreciating exchange rates is an-ticipated to combine with expectations for asharp falloff in both domestic and overseassales growth, leading to a retrenchment in pri-vate capital outlays (see table A5, earlier).

Falling investment is expected to lead tosimilar declines in regional imports, becausethe import content of investment tends to bequite high in Latin America. With imports de-clining almost 4 percent in 2009 and exportsfalling 2 percent, the contribution of trade togrowth will shift to positive 0.6 points ofgrowth for the first time in 20 years. But thedrop in investment and in export revenuesalso carry multiplier effects through the re-gional economy, with real household spendingeasing to a 3.1 pace from 5.4 percent in 2008,and GDP growth slowing to 2.1 percent. GDPcould rebound fairly quickly to 4 percentgains by 2010, should global credit marketsthaw, risk aversion subside, and OECD coun-tries revive on the back of renewed vigor inconsumer spending—in step with the antici-pated remission of inflation pressures. Thesedevelopments represent a substantial changefrom recent global forecasts prepared in June2008, when the region was expected to grow4.3 percent in 2009 and 4.2 percent in 2010(see Global Development Finance 2008,World Bank 2007d).

Growth in Brazil is expected to slow from5.2 percent in 2008 to 2.8 percent in 2009. In-flation has already started to level off, in partas a result of the Central Bank of Brazil’s rais-ing policy interest rates, amid falling com-modity prices. Consumer price inflation is ex-pected to diminish from 6.3 percent in 2008 to4.8 percent in 2009. Brazil is likely to witnessits first current account deficit since 2002, tiedto developments in the income accounts, asrepatriation of profits by foreign companies isunder way. However, a decline in the importsof capital goods is expected to help improve

the current account in 2009 and 2010(table A6).

Mexico’s close economic ties to the U.S.economy are expected to slow its growthsharply in 2009. Export volume growth—which was already in negative territory in2008—is projected to drop by 5 percent in2009. Argentina will perhaps see the sharpestgrowth falloff in the region as it experiencesdeclines in export market demand, commod-ity prices, and investment. Peru, Panama, andthe Dominican Republic will also slow aftervery high growth averaging 8 percent for thelast four years.

RisksWith the onset of the financial crisis, skyrock-eting costs of capital, or an outright shutdownin credit flows, are the primary risks faced bythe region. Should sovereign and corporatespreads not retreat from current levels, LatinAmerica could have difficulty financing newinvestment projects and sustaining currentprojects. Although central banks worldwidehave undertaken steps to inject liquidity intobanking systems, a marked thawing of inter-bank rates and revival of credit flows has yetto be seen.

The favorable external environment thatbenefited the region in the past five years hasalmost vanished. Both high-income anddeveloping-country GDP growth is slowing,diminishing demand for Latin America’s com-modities, manufactures, and services exports.Although inflation is still much higher thanin early 2007, increases in headline inflationappear to have peaked in July or August2008 in the seven largest economies in the re-gion. Although inflation will likely continueto ease given declining commodity prices, apotential revival of inflation remains a con-cern, given depreciating currencies, a movetoward monetary accommodation (mitigat-ing a portion of the economic downturn),and the potential for second-round inflationeffects.

10363_Pg141-180:10363_Pg141-180 11/29/08 7:13 AM Page 157


158

Table A6 Latin America and the Caribbean country forecasts(annual percent change unless indicated otherwise)

1991–2000a 2005 2006 2007 2008c 2009d 2010d

ArgentinaGDP at market prices (2000 US$)b 4.5 9.2 8.5 8.7 6.6 1.5 4.0Current account balance/GDP (%) �3.1 2.8 3.7 3.0 0.4 �3.2 �2.3BelizeGDP at market prices (2000 US$)b 5.9 3.1 5.6 3.0 2.8 2.1 2.9Current account balance/GDP (%) �7.3 �13.6 �1.2 �3.0 �3.7 �3.7 �1.6BoliviaGDP at market prices (2000 US$)b 3.8 4.4 4.8 4.6 4.1 3.6 4.3Current account balance/GDP (%) �6.1 6.5 11.5 13.4 13.3 9.9 8.3BrazilGDP at market prices (2000 US$)b 2.5 2.9 3.8 5.4 5.2 2.8 4.6Current account balance/GDP (%) �2.0 1.7 1.3 0.1 �1.3 0.6 1.1ChileGDP at market prices (2000 US$)b 6.4 5.7 4.3 5.1 4.2 3.4 4.7Current account balance/GDP (%) �2.7 1.2 5.0 4.3 �0.8 �0.8 0.0ColombiaGDP at market prices (2000 US$)b 2.5 4.7 6.8 8.2 3.7 2.6 4.7Current account balance/GDP (%) �1.9 �1.6 �2.9 �2.6 �3.0 �1.5 �0.6Costa RicaGDP at market prices (2000 US$)b 5.2 5.9 8.8 6.8 4.0 3.9 4.9Current account balance/GDP (%) �3.6 �4.9 �1.9 �8.7 �2.2 �3.3 �6.9DominicaGDP at market prices (2000 US$)b 1.8 3.1 4.0 3.2 3.1 �1.5 3.3Current account balance/GDP (%) �16.9 �32.6 �0.3 �0.4 0.4 6.2 6.9Dominican RepublicGDP at market prices (2000 US$)b 6.0 9.3 10.7 8.5 5.2 2.6 4.5Current account balance/GDP (%) �3.2 �1.9 �3.7 �5.7 �9.5 �8.0 �3.8EcuadorGDP at market prices (2000 US$)b 1.8 6.0 3.9 1.9 2.5 0.8 2.1Current account balance/GDP (%) �2.3 0.8 3.5 2.3 5.2 5.4 4.0El SalvadorGDP at market prices (2000 US$)b 4.6 3.1 4.2 4.2 2.0 2.6 2.9Current account balance/GDP (%) �2.0 �5.3 �4.7 �6.0 �8.4 �5.5 �5.2GuatemalaGDP at market prices (2000 US$)b 4.1 3.2 4.5 5.7 2.8 3.1 3.3Current account balance/GDP (%) �4.6 �4.5 �4.4 �5.1 �7.5 �5.3 �4.3GuyanaGDP at market prices (2000 US$)b 4.9 �2.2 4.8 5.5 4.8 4.0 3.1Current account balance/GDP (%) �15.4 �12.1 �19.7 �15.4 �18.2 �16.6 �15.1HondurasGDP at market prices (2000 US$)b 3.3 6.1 6.3 6.3 3.1 4.0 4.8Current account balance/GDP (%) �7.7 �3.0 �4.7 �10.6 �14.7 �9.6 �8.3HaitiGDP at market prices (2000 US$)b �1.3 1.8 2.3 3.5 3.0 3.8 3.9Current account balance/GDP (%) �1.7 �6.4 �7.6 �1.8 �11.9 �12.1 �13.1JamaicaGDP at market prices (2000 US$)b 1.9 1.8 2.5 1.2 0.9 0.8 2.3Current account balance/GDP (%) �2.7 �11.4 �10.9 �11.7 �17.0 �12.8 �10.8MexicoGDP at market prices (2000 US$)b 3.5 2.8 4.9 3.2 2.0 1.1 3.1Current account balance/GDP (%) �3.7 �0.7 �0.3 �0.6 �1.0 �1.7 �1.7NicaraguaGDP at market prices (2000 US$)b 3.4 4.3 3.7 3.5 2.2 1.5 2.9Current account balance/GDP (%) �28.7 �15.3 �16.4 �17.7 �20.9 �19.0 �15.4

(continued)

10363_Pg141-180:10363_Pg141-180 11/29/08 7:13 AM Page 158

Middle East and North AfricaRecent developments

The Middle East and North Africa regionhas been affected dramatically by develop-

ments in global commodity markets over thelast three years, notably in 2008.4 As a resultthere have been substantial up- and downshiftsin terms of trade, current account positions,and external financing requirements. Theseshifts have occurred at the same time as the ex-ternal environment for growth and for interna-tional finance deteriorated markedly. Still, re-gional GDP held up well through 2008, withdomestic demand, notably investment financedin large part by FDI, providing impetus forgrowth. The pace of GDP growth for the de-veloping countries of the region was un-changed in 2008 from the strong 5.8 percentregistered in 2007. A falloff in the Islamic Re-public of Iran’s hydrocarbon sector eased GDPgrowth among oil-dominant economies from6.4 percent in 2007 to 5.8 percent. And growth

among the more diversified economies pickedup from 3.8 percent in 2007 to 5.7 percent, ledby a strong recovery from drought in Morocco(table A7).

Commodity price changes carry extremeeffects across the regionThe region has undergone tortuous changelinked to global commodity prices through thelast years—from gradual increases to a surge incrude oil, food (especially grains), and raw ma-terials prices from 2005 through mid-2008—toa sudden and forceful unwinding of the bubbleduring the second half of 2008. On the upsideof the commodity run, the developing oil ex-porters—Algeria, the Arab Republic of Egypt(though a more diversified economy), the Is-lamic Republic of Iran, the Syrian Arab Repub-lic, and the Republic of Yemen—accumulated$82 billion in additional revenues over2003–07, with receipts coming to stand at$130 billion in the latter year. During the first


159

1991–2000a 2005 2006 2007 2008c 2009d 2010d

PanamaGDP at market prices (2000 US$)b 5.1 7.2 8.5 11.5 7.8 3.3 6.2Current account balance/GDP (%) �4.8 �3.1 �7.2 �5.4 �7.6 �9.4 �9.1PeruGDP at market prices (2000 US$)b 4.0 6.4 7.6 9.0 8.5 5.2 6.6Current account balance/GDP (%) �5.5 1.6 3.0 1.3 �2.2 �1.6 �1.6ParaguayGDP at market prices (2000 US$)b 1.8 2.9 6.0 6.8 4.2 3.0 3.8Current account balance/GDP (%) �2.2 0.5 2.0 0.8 0.0 �1.0 �0.8St. LuciaGDP at market prices (2000 US$)b 3.1 7.3 4.5 4.0 4.4 4.8 5.0Current account balance/GDP (%) �11.6 �17.4 �23.4 �21.4 �21.8 �20.7 �19.8St. Vincent and the GrenadinesGDP at market prices (2000 US$)b 3.1 1.5 4.5 5.5 6.3 �0.6 5.6Current account balance/GDP (%) �18.8 �26.3 �25.9 �24.5 �24.7 �24.2 �20.0UruguayGDP at market prices (2000 US$)b 3.0 6.6 7.0 7.4 4.7 2.8 3.0Current account balance/GDP (%) �1.5 0.1 �2.3 �0.7 �1.7 �1.4 �0.9Venezuela, R. B. deGDP at market prices (2000 US$)b 2.1 10.3 10.3 8.4 5.3 1.0 3.2Current account balance/GDP (%) 2.6 17.5 14.8 8.8 8.7 9.0 8.0

Source: World Bank.

Note: Growth and current account figures presented here are World Bank projections and may differ from targets contained inother World Bank documents. Barbados, Cuba, Grenada, and Suriname are not forecast because of data limitations.


10363_Pg141-180:10363_Pg141-180 11/29/08 7:13 AM Page 159

half of 2008, revenues jumped a further50 percent to nearly $200 billion. Since then,however, the financial crisis and expectationsof much lower global growth have caused oilprices to plunge from peaks of nearly$150/bbl in early July to near $65/bbl byend-October 2008. As a result, regional oilexporters are now experiencing a substantialdownshift in hydrocarbon receipts, terms oftrade, and current account surplus posi-tions that will manifest more clearly in 2009(figure A11).

The oil exporters’ current account surplusincreased from 17.2 percent of GDP in 2007only moderately to 18.7 percent in 2008, butglobal economic recession in 2009 will pres-sure oil prices lower and yield a sizable addi-tional falloff in world oil demand. The group’s

current surplus position is projected to dropsteeply to 8 percent of GDP during 2009 andto 5.4 percent by 2010. Real-side growth willbe affected as revenue declines are likely to re-sult in downsizing of ambitious investmentprojects or postponement of planned pro-grams. At the same time, the Organization ofPetroleum Exporting Countries (OPEC) willlikely attempt to set limits on the decline in oilprices by constraining oil production, whichwill depress the oil sector in many economies,with ripple effects to the non-oil economy andthe private sector.

The diversified economies of the region,including Jordan, Lebanon, Morocco, andTunisia, are to varying degrees highly depen-dent on imports of oil and refined petroleumproducts, as well as on food and feedstuffs,


160

Table A7 Middle East and North Africa forecast summary(annual percent change unless indicated otherwise)

1991–2000a 2005 2006 2007 2008* 2009† 2010†


Private consumption 3.9 5.0 6.2 6.1 7.0 4.2 6.0Public consumption 4.2 5.6 4.2 1.8 8.7 5.4 5.4Fixed investment 3.9 7.8 4.8 16.8 18.9 7.0 10.5Exports, GNFSd 3.1 9.5 6.7 6.0 10.1 �2.1 4.9Imports, GNFSd 1.4 14.0 7.6 14.3 19.8 1.7 8.8

Net exports, contribution to growth 0.4 �1.7 �0.6 �3.1 �4.3 �1.3 �2.2Current account balance/GDP (%) �0.3 10.9 14.9 12.8 13.5 6.0 4.1GDP deflator (median, LCU) 9.1 6.3 8.0 5.3 14.1 6.6 7.2Fiscal balance/GDP (%) 4.0 5.5 0.7 1.3 2.0 0.0 �1.0

Memo items: GDPMiddle East and North Africae 3.4 5.1 4.9 5.1 5.9 4.1 5.5Resource poor and labor abundantf 4.2 3.8 6.3 5.6 6.5 4.3 5.9Resource rich and labor abundantg 3.3 4.6 4.5 6.1 5.1 3.6 4.5Resource rich and labor importingh 3.0 6.5 4.2 4.1 6.0 4.3 6.0

Egypt, Arab Rep. 4.3 4.4 6.8 7.1 7.2 4.5 6.0Iran, Islamic Rep. 3.7 4.3 5.9 7.8 5.6 3.5 4.2Algeria 1.7 5.3 1.8 3.1 4.9 3.8 5.4

Source: World Bank.* Estimate.† Forecast.

a. Growth rates over intervals are compound averages; growth contributions, ratios, and the GDP deflator are averages.b. GDP measured in constant 2000 U.S. dollars.c. GDP measured at PPP exchange rates.d. Exports and imports of goods and nonfactor services.e. Geographic region includes these high-income countries: Bahrain, Kuwait, and Saudi Arabia.f. Arab Rep. of Egypt, Jordan, Lebanon, Morocco, and Tunisia.g. Algeria, Islamic Rep. of Iran, Syrian Arab Rep., and Republic of Yemen.h. Bahrain, Kuwait, Oman, and Saudi Arabia.

10363_Pg141-180:10363_Pg141-180 11/29/08 7:13 AM Page 160

notably wheat and coarse grains. Their termsof trade worsened by 4.2 percent in 2008,pushing the group’s current account deficit to7.3 percent of GDP (not witnessed since theAsia crisis of 1997) from 3.6 percent in 2007(see figure A11). Looking forward, theseeconomies will benefit from lower commodityprices through 2010, and current accountdeficits are projected to decline to 0.7 and 0percent of GDP in 2009 and 2010 respectively.

Effects of financial crisis are fairly muted,but several countries are vulnerableTo date, the direct effects of the financial cri-sis experienced by most developing economiesin the region have been relatively mild. Banksand investment companies in the Middle Eastand North Africa were not large holders ofsubprime mortgage-backed securities, or“toxic assets” (though there may be questionsconcerning portfolios of sovereign wealthfunds in the Gulf States). Indirect effects, how-ever, have become evident. Following the an-nouncement of the U.S. financial rescue planin early October 2008, spreads on sovereigndebt increased 170 basis points for Lebanonand 100 points for Egypt; but these increasescontrasted well with the average rise of 250basis points for all developing economies atthat time. With country-specific developments

set against the background of subsequent con-certed policy rate reductions across the OECDcountries, a step-up in economic stimulusplans and the beginnings of a thaw in creditmarkets, spreads in Egypt eased to 350 points,but those in Lebanon escalated to 730 basispoints by early November.

Equity markets across the region echoedthe sharp declines seen by emerging marketsgenerally, as international (and domestic) in-vestors withdrew funds from the asset class.From peak levels in the spring through early-November 2008, the Egyptian bourse dropped54 percent, Morocco’s exchange fell 33 per-cent, and the Gulf Cooperation Council (GCC)markets in aggregate declined 50 percent. Thiscontrasts with a 54 percent decline in theMSCI index which covers all emerging mar-kets (figure A12).

Gross capital flows to countries in the re-gion have also declined, and may be expectedto weaken further. Bond issuance dropped bytwo-thirds from $4.6 billion to $1.5 billionbetween January and August 2007 and thelike period of 2008. Equity issuance declinedfrom $2.1 billion to $750 million or 65 per-cent in the period as well. But a surge in bankborrowing from $4 billion to $14 billion inthe period offset the downturn in other


161

50

75

100

125

200

175

150

MSCI equity market price indexes (percentage change, US$)

Figure A12 Markets in the Middle East andNorth Africa are hard hit by financial crisis

Source: Morgan-Stanley.

Jan. 12007

May 12007

Sep. 12007

Jan. 12008

May 12008

Sep. 12008

Morocco

GCC market average

MSCI total

Egypt,ArabRep. of

Note: Gulf Cooperation Council (GCC).

Middle East &North Africa

Oilexporters

Diversifiedeconomies

25

210

0

5

20

15

10

Figure A11 Current account positions in the Middle East and North Africa set to shift dramatically

Source: World Bank.

Current account balance as a share of GDP (%)

2006 2007 2008

2009 2010

10363_Pg141-180:10363_Pg141-180 11/29/08 7:13 AM Page 161

finance components. Moreover, preliminarydata for September show declines across allsegments of flows to the region. The processof deleveraging across high-income financialinstitutions appears to have raised the possi-bility for a potentially sharp reduction in cap-ital flows, particularly syndicated bank lend-ing and to a lesser degree bond issuance, forthe region. This is likely to carry adverse ef-fects across countries, but with highly differ-entiated outcomes.5

Several countries stand exposed to the riskof adverse developments in international fi-nancial markets, which could negatively af-fect investment spending and growth. Thesecountries may suffer from fragilities in macro-economic structure (for example, a string ofsubstantial current account- or fiscal deficits)or from the presence of stress points madeclearer by the heightening of investor riskaversion.

The vulnerability index presented in chap-ter 1 (a weighted measure of the exposure of acountry to developments in sovereign spreads,equity markets, and exchange rates, as well asin gross capital inflows) suggests that Lebanon,Syria, Jordan, and Egypt have been among themore affected countries in the region, thoughthe vulnerability of these economies is low incontrast with the average exposure for otherregions. Under a global scenario in which fi-nancial markets require a prolonged period toreturn to balance, these countries might findthemselves at risk of adverse capital move-ments, pressures on equity markets, exchangerates, and eventually investment and growth.

Production is mixed; inflation rampshigher, denting budgets across the regionIndustrial production for the diversifiedeconomies of the region tailed off in late 2008,shifting from gains of 8 percent (on a GDP-weighted basis) during the first quarter to 4.5percent by the third quarter (year-over-year).This decline reflects the increasingly sluggishperformance of exports to key European andU.S. markets as well as emerging softness in do-mestic demand. In contrast, production among


162

the developing oil exporters picked up fromnegative ground in the first quarter to 3 per-cent in the third, as growth in the non-oil sec-tors in both Algeria and the Islamic Republicof Iran well outpaced sluggish conditions inhydrocarbons. Output among the high-incomeGCC economies continued to grow quickly,underpinned by a continued rapid pace ofcommercial and residential real estate develop-ment in Bahrain, Qatar, Saudi Arabia, and theUnited Arab Emirates (UAE). Production gainsfor the group jumped from 3 percent to 10 per-cent by the third quarter, with Qatar up 12 per-cent, Saudi Arabia climbing 11 percent, andthe UAE moving to 20 percent growth in theperiod.

The surge in global prices for crude oil,food, and feed grains (50 percent or moreduring the first half of 2008), together withoverheated domestic demand in severaleconomies in the region (notably Egypt, theIslamic Republic of Iran, and a number ofGCC countries), led to a sharp rise in con-sumer price inflation across the Middle Eastand North Africa (figure A13). Consumerprices for the diversified economies (GDP-weighted) accelerated to 7 percent in August2008 from a trough nearer 1.5 percent inmid-2007; much of the increase was centered

0

4

8

24

20

16

12

Figure A13 Inflation rises across the MiddleEast and North Africa


Note: Gulf Cooperation Council (GCC).

Jan.2006

Jul.2006

Jan.2007

Jul.2007

Jan.2008

Jul.2008

Consumer price inflation (percentage change, year-over-year)

GCC

Diversified

Egypt, Arab Rep. of

Developingoil exporters

10363_Pg141-180:10363_Pg141-180 11/29/08 7:13 AM Page 162

in Jordan, where inflation reached 20 percent.Aggregate inflation for developing oil exportersalso breached 20 percent, mainly reflecting de-velopments in the Islamic Republic of Iran,where substantial monetary stimulus led tooverheating, pushing inflation to 27 percent;consumer prices in Egypt rose to a 24 percentpace, pushed up primarily by rising foodprices and expanding domestic demand fueledby monetary growth.

Inflation remains a key challenge for the re-gion. Although extensive reliance on fuel andfood subsidies helps limit inflationary pres-sures, it comes at a very high fiscal cost. Notonly do such steps reduce fiscal space to ad-dress other priorities, as discussed in chapter 3,they tend to be very inefficient mechanisms foralleviating poverty. Iranian energy subsidies ex-ceeded 20 percent of GDP in 2007—08. Foodand energy subsidies in Egypt increased to 1.9and 6.9 percent of GDP in fiscal 2008, up from1.3 and 5.5 percent, respectively, in fiscal 2007.Second-round inflationary effects have beenboosted in several countries that responded tohigh food prices by increasing wages of selectgroups to help mitigate the worst of the impacton living standards.

Domestic demand, underpinnedby substantial FDI, drives growthin the regionStrong gains in consumer spending, and espe-cially in fixed investment, have been the keyfactors supporting growth across the region in2008—all the more so as exports of the oil-dominant economies have been restrained in aneffort to prop up crude oil prices, and those ofthe diversified economies have been increas-ingly affected by the slowdown in export mar-ket demand. Investment in the region grew al-most 20 percent in 2008, accounting for 3.4points of the region’s 5.8 percent growth in theyear, while consumer spending grew 7 percent(see table A7, earlier). Large infrastructure in-vestment projects, such as the ProgrammeComplémentaire de Soutien à la Croissance(PCSC) in Algeria, find counterparts in new realestate, commercial, and industrial developments

in countries such as Egypt, Jordan, Morocco,and Tunisia, funded in large part by direct in-vestment flows from the GCC.

Sectors benefiting from FDI have diversi-fied from real estate and tourism-related prop-erties toward industrial and infrastructureprojects in the past few years. FDI to the de-veloping countries of the region increasedmore than five-fold from $4.7 billion in 2000to $26.4 billion in 2006; preliminary estimatesfor 2007 suggest a moderate downshift to$21.5 billion, reflecting diminishing levels offlows to Egypt, Jordan, and Tunisia.

Recent examples of FDI-driven develop-ments include the Mediterranean Gate pro-ject, which aims to turn Tunis into a regionalhub for finance, business, and technology.Work has begun on the first $1 billion of the$25 billion project, which is slated to house2,500 international firms and provide350,000 jobs over a 20-year period. In Jordan,Aqaba Development signed a $100 millionagreement to develop an industrial port atAqaba to handle potash exports. And inMorocco, construction is under way on a newRenault/Nissan production site—the plant willsponsor about 6,000 direct jobs, with 90 per-cent of production exported.

Among the region’s oil exporters, Algeriaexperienced a fillip to growth in 2008, to 4.9percent from 3.1 in 2007, as gains continued ata rapid 6 percent clip in the non-oil sector, no-tably in construction and services linked to in-frastructure projects (table A8). Algeria standsin fair stead to weather financial spillovers fromthe global crisis; at end-September 2008, reservesstood at $140 billion, up $30 billion from end-2007. A falloff in the oil sector to 2 percent pres-sured growth in the Islamic Republic of Iranfrom 7.8 percent in 2007 to 5.6 percent. Over-all GDP was supported by industry, whichadvanced 7.4 percent, services (6.8 percent),and agriculture (6.2 percent). Growth is beingpowered by a highly expansionary fiscal policy,which has pushed inflation toward 30 percent;and public spending is anticipated to movehigher still ahead of presidential elections slatedfor 2009.


163

10363_Pg141-180:10363_Pg141-180 11/29/08 7:13 AM Page 163

Growth in Egypt continued its strong mo-mentum into 2008, moving from GDP gains of7.1 percent in 2007 to 7.2 percent, as invest-ment advanced by more than 30 percentfunded in large part by FDI. Egypt has been thelargest recipient of FDI in the region, attracting$13 billion (8.4 percent of GDP) in fiscal 2008,up from $11 billion in the previous fiscal year.But the country’s equity markets have beenhard hit by recent financial turmoil, with theCASE index off more than 50 percent sinceMay 2008. Moody’s and Fitch had earlier low-ered the outlook on the country’s Ba1 rating tonegative from stable, citing inflation and thefiscal deficit as primary concerns. Egypt ap-pears among the more exposed economies in

the Middle East and North Africa to potentialrepercussions from developments in interna-tional financial markets.

In Syria, domestic demand, supported bystrong output gains in transport, communica-tions, finance and real estate, and public ad-ministration, drove growth of 3.7 percent dur-ing 2008. Declining oil production is the keychallenge facing the economy. Oil outputdropped 23 percent between 2003 and 2007,increasing pressure to expand the scope forprivate non-oil activities. Similar falloffs in oilproduction in the Republic of Yemen continueto plague the economy, restraining GDPgrowth there to 2.7 percent in 2008; though acoming online of large natural gas facilities in


164

Table A8 Middle East and North Africa country forecasts(annual percent change unless indicated otherwise)

1991–2000a 2005 2006 2007 2008c 2009d 2010d

AlgeriaGDP at market prices (2000 US$)b 1.7 5.3 1.8 3.1 4.9 3.8 5.4Current account balance/GDP (%) 3.2 21.6 24.7 21.2 23.3 12.2 9.8Egypt, Arab Rep.GDP at market prices (2000 US$)b 4.3 4.4 6.8 7.1 7.2 4.5 6.0Current account balance/GDP (%) 0.9 2.3 2.4 0.3 �5.1 �2.7 �1.3Iran, Islamic Rep.GDP at market prices (2000 US$)b 3.7 4.3 5.9 7.8 5.6 3.5 4.2Current account balance/GDP (%) 1.2 20.4 27.6 28.5 36.3 17.9 12.1JordanGDP at market prices (2000 US$)b 5.1 7.3 6.3 6.0 5.5 4.2 6.0Current account balance/GDP (%) �4.3 �17.7 �8.1 �13.9 �14.9 �0.8 0.0LebanonGDP at market prices (2000 US$)b 7.2 1.0 0.0 2.0 5.5 4.0 4.5Current account balance/GDP (%) — �12.8 �5.3 �8.7 �16.4 �6.5 �5.9MoroccoGDP at market prices (2000 US$)b 2.4 2.4 7.8 2.7 6.2 4.0 6.0Current account balance/GDP (%) �1.4 1.7 2.0 �0.3 �4.5 1.0 1.7Syrian Arab Rep.GDP at market prices (2000 US$)b 5.1 4.5 5.1 6.6 3.7 2.5 4.2Current account balance/GDP (%) 1.0 1.0 2.7 1.2 2.4 �2.0 �3.3TunisiaGDP at market prices (2000 US$)b 4.7 4.2 5.7 6.3 5.1 3.7 5.8Current account balance/GDP (%) �4.3 �1.1 �2.0 �2.6 �3.9 0.0 0.8Yemen, Rep.GDP at market prices (2000 US$)b 5.5 4.6 3.2 2.8 2.7 5.7 4.0Current account balance/GDP (%) �4.3 3.7 8.1 �0.5 1.3 1.3 2.5

Source: World Bank.

Note: Growth and current account figures presented here are World Bank projections and may differ from targets contained inother World Bank documents. Djibouti, Iraq, Libya, and West Bank and Gaza are not forecast because of data limitations.


10363_Pg141-180:10363_Pg141-180 11/29/08 7:13 AM Page 164

2009 should yield a fillip to growth by morethan 3 percentage points at that time.

Among the more-diversified economies,growth in Morocco recouped sharply to 6.2percent in 2008 from the drought-inflicted 2.7percent outturn of 2007. Vigor in non-agricultural sectors, especially in telecommu-nications, financial services, and construction,has driven growth. Policies to control domes-tic prices––food and fuel subsidies, temporarywaivers on customs duties for cereals, and ac-tions to fight price speculation––have helpedmaintain overall inflation at relatively low lev-els compared with many countries in the re-gion. But subsidies have tripled in two years,reaching close to 6 percent of GDP in 2008. InTunisia, GDP eased to 5.1 percent growth in2008, from 6.3 percent in 2007, largely be-cause of deterioration in the external environ-ment, in particular the economic slowdown inthe EU. Remaining import tariffs on EU goodswere dismantled in January within the frame-work of the EU-Tunisia Association Agree-ment, and steps have been taken in the financialsector to reduce unsound and nonperformingloans by improving credit risk appraisals.Over the first seven months of 2008, foreigninvestment in industry increased 47.2 percent,widening from the earlier focus of FDI intourism.

Jordan’s growth slipped to 5.5 percentfrom 6 percent in 2007, on the back of still-buoyant domestic demand, financed in part bylarge capital inflows. Heavy public sector out-lays in 2008 (and anticipated in the draft 2009budget) suggest that fiscal and financing pres-sures will continue in the short term. The risein fuel and food prices, together with expan-sionary policy, has pushed inflation above 22percent as of August, and the current accountdeficit widened to almost 15 percent of GDPin the year. These circumstances place Jordanat some risk of interruption in private capitalflows in the short term, but official develop-ment assistance and worker remittances mayhelp the country bridge the potential financinggap. Finally, in Lebanon, GDP picked up to a5.5 percent pace in 2008, from 2 percent

during 2007, on a strong rise in consumerspending. At the same time, inflation loftedinto double digits on the back of food and fuelprices, as well as high public sector wagesettlements. Lebanon managed to finance itslarge trade deficit through stronger exports ofservices and higher net inflows from abroad.

The medium-term outlookThe global downturn and financial crisis willexact a toll on growth in the Middle East andNorth Africa, but one that will be less dramaticthan, for example, in Europe and Central Asiaor South Asia, where country exposure andfragility of initial conditions are considerablymore pronounced. As world oil demand fallssharply, any decisions by the region’s oil ex-porters to curtail output to set a “floor” underoil prices—will play a large role in shapinggrowth profiles. And the shift from windfallrevenue gains to current account surplus posi-tions of less than 6 percent of GDP by 2010,much weaker oil revenues, tighter credit con-ditions, and weaker demand for the region’sexports (including tourism) are expected tocause investment to decelerate sharply, risingby 7 percent in 2009 after growing 18.9 per-cent in 2008.

As a result, the region’s GDP is anticipatedto slow from 5.8 percent in 2008 to 3.9 per-cent in 2009. Growth among the oil exportersas well as the diversified economies is antici-pated to fall to about 4 percent in 2009 (figureA14). Recovery in 2010, predicated upon aquick resolution of the financial crisis in high-income countries and a moderate revival ofOECD growth, would see GDP pick-up to 5.2percent, led by a return to 5.7 percent growthamong the diversified economies. A very grad-ual buildup in global oil demand is likely torestrain GDP gains among the oil-exportingcountries to 5 percent in 2010. Mainly reflect-ing cuts in oil production, export volumes areprojected to decline 2.1 percent in 2009, whilethe regional current account surplus falls to6 percent of GDP, from 13.5 percent in 2008.

Recovery for the region in 2010 hinges ona pickup in exports and a moderate upturn in


165

10363_Pg141-180:10363_Pg141-180 11/29/08 7:13 AM Page 165

investment, but primarily on a 1.8 percentagepoint pickup in household outlays to growthof 6 percent, as the earlier run-up in commod-ity prices and consumer price inflation moder-ates, giving way to gradual stabilization andto a pick-up in consumer purchasing power.The region’s current account position shouldcontinue to narrow to some 4 percent of GDP,providing a new set of “initial conditions”from which developments into the nextdecade are likely to spring.

RisksUncertainty surrounding the medium-termpath for oil prices is probably the element ofgreatest risk confronting the region. Wherethe global price of oil settles, grounded in thefundamentals—as well as by pressures exertedby OPEC—will determine the potentialgrowth path for the oil-dominant economiesof the region. The “base case” view positsworld crude oil prices remaining within a $65to $75/bbl range through 2010, moving to-ward a real equilibrium price of $60/bbl in2007 dollars by 2015. But substantial down-sides to this price forecast can be envisionedshould the slowdown in developing-countryGDP growth fall much below the 4.5 percentposited for 2009. Although a repeat of1985–86, when oil prices tumbled to $10/bbl

is unlikely, prices below $50/bbl could be inthe cards, with attendant adjustments requiredby the region’s exporters.

A second element of concern for the regionis the potential for unrest among the populaceunder the potentially harsh conditions of aglobal recession. A slowdown in remittanceinflows would carry direct effects to poor fam-ilies in need of income to sustain householdconsumption. And government budgets willremain under pressure, in part to maintainsubsidies for basic goods.

South AsiaRecent Developments

GDPgrowth in South Asia slowed markedlyin 2008 to 6.3 percent from 8.4 percent in

2007.6 The onset of the financial crisis in theUnited States and Europe in mid-September2008—which led to severe financial turmoil inemerging markets, including in many SouthAsian countries—ushered in a downshift in ac-tivity that started to take hold in late-2008.Growth had already begun to wane in the re-gion prior to the onset of the global crisis, asrising inflationary pressures and tight creditconditions had started to take a toll on domes-tic activity, while already slowing external de-mand and high international commodity pricesled to a deterioration in external positions.

The initial effects of the global financial cri-sis in South Asia were sharp corrections in re-gional equity markets. Bourses in India, Pak-istan, and Sri Lanka dropped 57 percent, 39percent, and 35 percent, respectively, over theyear through mid-November (and 66, 50, and39 percent, when measured in U.S. dollars).Notably in Pakistan, curbs on the sale of equi-ties were imposed in August, effectively pre-venting the exit of existing investors and dis-couraging potential new investors.

Equity sell-offs and ‘flight-to-quality’ con-tributed to significant currency depreciation insome countries, with local currencies in India,Pakistan, and Nepal7 falling by 21 percent,30 percent, and 21 percent, respectively,against the U.S. dollar, over the year through


166

0

1

2

3

4

5

6

7

Figure A14 Notable slowing of growth acrossthe Middle East and North Africa in 2009

Source: World Bank.

GDP growth (%)

Oil exporters Diversified economies

GCC countries

2007 2008 2009 2010

10363_Pg141-180:10363_Pg141-180 11/29/08 7:13 AM Page 166

mid-November. The Sri Lankan rupee depreci-ated by nearly 2 percent when the CentralBank allowed the peg against the U.S. dollarto adjust at end-October 2008. In contrast, theBangladeshi taka appreciated slightly (2 per-cent) over the same period.

Notably, the region’s banking sectors havebeen largely insulated from the crisis, givenvery limited exposures to the toxic debt instru-ments tied to U.S. sub-prime mortgages. Withrespect to the associated impacts of the finan-cial crisis on the real economy—as financingfor corporations, loans for households, andtrade credit for exporters have become signifi-cantly more difficult to obtain—indications ofa fall-off in external and domestic demand havebegun to trickle in. For example, India’s goodsexports contracted 12 percent in October (year-over-year). This comes on the heels of a sub-stantial deceleration in export growth to 10 per-cent in September from 27 percent in Augustdespite the marked weakening of the rupee. SriLanka’s exports also declined, falling 9.4 per-cent in September, contrasted with growth of16.6 percent and 24.1 percent in August andJuly, respectively. Further, consumer confidencein India has deteriorated, with the index relatedto consumer spending down for a fourth con-secutive month in October.8

Weaker conditions in South Asia were evi-dent in the region prior to the onset of theglobal financial crisis, and were marked by anincreasingly challenging global environment.In particular, sharply negative terms-of-tradeeffects from the rise in oil and global non-energy commodity prices, which peaked in mid-2008, acted as a drag on regional growth andcontributed to a doubling of the regional cur-rent account deficit in the year. Rising interna-tional prices also contributed to a pronouncedbuildup in South Asia’s inflation pressures.Higher prices, particularly for food and fuel,undermined real household incomes—withthe poorest households generally affected themost—thus crimping household expenditures.Several governments attempted to offset theinternational price hikes with domestic subsi-dies, placing strains on fiscal balances. But

with substantial and sustained increasesthrough the middle of 2008, a greater degreeof feed-through of higher food and fuel pricesto households in these countries became in-evitable. Tighter credit conditions, moderatingdemand, higher prices, and diminishing levelsof confidence weighed on consumer and busi-ness spending alike. And investment growthdecelerated to single digits from the robustgrowth witnessed during recent years.

The slowdown in growth during 2008reflected increasing weakness in the region’stwo largest economies, India and Pakistan(table A9). In India, growth slowed across allsectors, with tighter monetary policy, rising in-flationary pressures, and mounting fiscal andcurrent account deficits weighing down eco-nomic activity. The more recent onset of theglobal financial crisis resulted in sharp lossesin India’s equity markets and drove down thevalue of the rupee. Foreign institutional in-vestors pulled out of India to cover losses inhigh-income countries and as risk aversionheightened across the globe.

In Pakistan, the economy deterioratedsharply over the course of 2008, as headline in-flation surged, and the current account and fis-cal deficits jumped on the back of rising oil andfood prices. Political turmoil and ongoingsecurity concerns have also taken a toll onPakistan’s economy, while the global financialcrisis added substantial downward pressureson its financial markets. Prior to reaching anagreement with the IMF for standby credit inmid-November, Pakistan came close to a full-blown balance of payments crisis. In neigh-boring Afghanistan, the economy has beenhurt by a decline in agricultural output causedby poor precipitation, a sharp rise in interna-tional food prices, and the wheat export re-strictions imposed by Pakistan, in addition tothe disruptive effects of the spreading insur-gency. And while GDP growth in Bhutan re-mained vibrant at 14.4 percent in 2008, itmoderated from the 17 percent expansion of2007, stemming from the initial boost from thefirst full year of operation of the immense Talahydropower project.


167

10363_Pg141-180:10363_Pg141-180 11/29/08 7:13 AM Page 167

In contrast with broadly declining activityin the region, growth in Bangladesh heldsteady, with domestic demand buoyed by asharp increase in remittance inflows and by ro-bust garment exports recorded in the first halfof 2008. With relatively thin capital markets,Bangladesh’s equities experienced much moremuted declines than those experienced in otherregional markets. Growth in Sri Lanka hasalso proven resilient in 2008, primarily be-cause of a marked rise in agricultural produc-tion and a boom in tea exports, which helpedto offset slower growth in garment exports.While its equity markets also suffered sharpcorrections with the global financial crisis, theSri Lankan rupee, which was pegged to theU.S. dollar early in 2008, remained stableagainst that currency. Partially in consequence,real appreciation of the Sri Lankan rupee hascontributed to substantial widening of the cur-rent account deficit. In Nepal, growth firmedin 2008, helped by higher agricultural output

and rising remittance inflows. These factorssupported an increase in household incomesand private consumption, despite a buildup ininflation pressures.

The general deterioration in regional tradebalances has been offset by large remittanceinflows, which represent a sizable, and gener-ally increasing share of GDP: during 2007,14 percent in Nepal, 8 percent in Bangladeshand Sri Lanka, 4 percent in Pakistan, and 3percent in India. FDI inflows remained strongthrough the first half of 2008, helping to easeexternal financing requirements. In India, FDIsurged to 3 percent of GDP in 2008, up from1.4 percent in 2007. FDI inflows to Pakistanremained relatively steady through the sum-mer of 2008—on course to match the 3.7 per-cent of GDP recorded in 2007—but the ex-treme financial and economic difficultiesencountered during the second half of the yearwere likely to have changed that for the worse.In 2007, FDI inflows to Sri Lanka and


168

Table A9 South Asia forecast summary(annual percent change unless indicated otherwise)

1991–2000a 2005 2006 2007 2008e 2009f 2010f



Net exports, contribution to growth �0.1 �1.0 �0.8 0.0 �0.5 0.2 0.0Current account balance/GDP (%) �1.6 �1.2 �1.5 �1.6 �3.5 �2.0 �1.9GDP deflator (median, LCU) 8.2 6.5 9.2 8.4 9.7 8.0 6.0Fiscal balance/GDP (%) �7.7 �5.9 �6.1 �6.4 �8.1 �8.6 �8.0

Memo items: GDPSouth Asia excluding India 4.4 6.7 6.4 6.1 6.1 4.0 5.2India 5.5 9.2 9.7 9.0 6.3 5.8 7.7Pakistan 3.9 7.7 6.2 6.0 6.0 3.0 4.5Bangladesh 4.8 6.0 6.6 6.4 6.2 5.7 6.2

Source: World Bank.

Note: To simplify presentation across countries and with other regions, annual national income and product account data forSouth Asia are reported in calendar years, although official country data are originally reported by fiscal year for Bangladesh,India, Pakistan, and Nepal.

a. Growth rates over intervals are compound average; growth contributions, ratios, and the GDP deflator are averages.b. GDP measured in constant 2000 U.S. dollars.c. GDP measured at PPP exchange rates.d. Exports and imports of goods and nonfactor services.e. Estimate.f. Forecast.

10363_Pg141-180:10363_Pg141-180 11/29/08 7:13 AM Page 168

Bangladesh reached 1.7 and 1 percent of GDPrespectively (figure A15).

In contrast, net portfolio flows to the regionturned sharply negative during the first half of2008, shifting from vibrant inflows of recentyears. In India, where portfolio inflows surgedto 3 percent of GDP in 2007, outflows areprojected to exceed 1 percent of GDP in 2008.With the increase in global risk aversion andrebalancing of portfolio holdings in the high-income countries, gross capital flows deceler-ated in 2008, with an especially sharp falloff inequity and bond issuance and a somewhat lesspronounced decline in bank borrowing. Falter-ing investor confidence led to higher interna-tional bond spreads, with those for Pakistanand Sri Lanka spiking to prohibitive levels inSeptember and October. Hard currency re-serves were drawn down to varying degrees, asinvestors pulled out of regional markets and ascentral banks sought to shore up currencies.

Fiscal policy across South Asia is broadlyexpansionary, with deficits generally exceed-ing 4.5 percent of GDP—they are projected toreach 8.5 percent in India, 7.5 percent in Pak-istan and Sri Lanka, and 4.7 percent inBangladesh in 2008. Nepal is an exception,where the 2008 deficit is projected at 2.8 per-cent of GDP, although that would be doublethe 2007 deficit. In 2008, budget deficits roseacross the region—or remained high—as

price subsidies for food, fuel, and fertilizercontributed to higher fiscal outlays. In somecases (India, Bangladesh, Pakistan), the subsi-dies contributed to a reversal in the generaltrend toward fiscal consolidation in recentyears. Downward pressure on the revenuestream, resulting from the deceleration ingrowth, also played a role. As a consequence,a number of regional governments had begunto cut development spending.

With low, or in many cases, negative real in-terest rates, monetary policy is also broadly ex-pansionary in South Asia. Prior to the onset ofthe global financial market crash in September2008, some countries had tightened monetaryconditions through interest rate hikes (India)or slower credit growth (Sri Lanka) in an effortto curtail rising inflationary pressures. Later inthe year, however, as the credit crunch becamemanifest, regional monetary authoritiesquickly responded by injecting liquidity intobanking systems through various measures, in-cluding lowering required reserve ratios andreducing policy interest rates.

Medium-term outlookThe outlook for regional growth is highly un-certain, because of the sustained degree ofvolatility and synchronized nature of the slow-down across countries—and because the fullextent of financial disruption on both the re-gional and global economies remains unclear.South Asian GDP growth is projected to stepdown to 5.4 percent in 2009 from 6.3 percentin 2008. Continued financial sector volatilityand balance sheet weakness will translate intoongoing risk aversion. That is expected to leadto a further contraction in portfolio inflowsand mute the prospects for FDI, primarily af-fecting India and Pakistan, which receive thelion’s share of the region’s inflows. In turn,these factors are projected to lead to a sharpfalloff in private investment growth. Equityprice declines are expected to generate negativewealth effects, especially in the case of India,where market capitalization reached 160 per-cent of GDP in 2007, up from 90 percent in


169

India Pakistan Sri Lanka Bangladesh

3

0

6

9

Figure A15 Key international finance links in South Asia, 2007

Source: World Bank.

Share of GDP (%)

Inwardremittances

Portfolioinflows

FDIinflows

10363_Pg141-180:10363_Pg141-180 11/29/08 7:13 AM Page 169

2006 and where the housing boom has begunto lose steam (table A10).

Weakening foreign demand is expected tolead to a significant slowing in regional exportgrowth, including services. In particular, the in-formation technology and communications sec-tor is considered vulnerable to shifts in financialsector activity, and clothing and tourism rev-enues are vulnerable to shifts in discretionaryspending. Potential mitigating factors includecost-cutting measures by companies in high-income countries to the benefit of outsourcingsuppliers (such as India) and shifts in spendingto low-priced retailers, such as Wal-Mart, to thebenefit of their suppliers (such as Bangladesh).Recession in high-income countries and a slow-down in growth among the Gulf oil exportersare expected to depress remittances inflows.

However, the set of unfavorable globalconditions are anticipated to lead to lowercommodity prices, which will not only pro-vide a fillip to real household incomes, butalso provide governments with greater scopefor fiscal stimulus. Falling commodity priceswill reduce the import bill and boost the re-gion’s terms of trade. At the regional level, the

current account deficit is expected to narrowsubstantially. Additionally, the recent sharpdepreciation of local currencies against thedollar for India, Pakistan, and Nepal will helpboost export competitiveness. This shouldhelp offset partially the negative effects of thecoming contraction in world trade.

To help cushion the downturn related to thefinancial crisis, South Asian governments areseen to pursue countercyclical measures, al-though fiscal space is limited. Thus, monetarypolicy measures will often be the key mecha-nism for response to the crisis, although reduc-tions in policy rates should be undertaken withcare where there is pressure on the exchangerate. Even with tight budget envelopes, re-gional governments can improve the efficiencyof public outlays by more directly targetingsafety net programs to the benefit of the poor.In addition, and particularly where both fiscaland monetary policy responses are con-strained, expansion of structural reformsshould be pursued to stimulate growth in thenear term and improve prospects for themedium and longer terms. Examples includeimproving governance and management of


170

Table A10 South Asia country forecasts(annual percent change unless indicated otherwise)

1991–2000a 2005 2006 2007 2008c 2009d 2010d

BangladeshGDP at market prices (2000 US$)b 4.8 6.0 6.6 6.4 6.2 5.7 6.2Current account balance/GDP (%) �0.4 �0.3 2.0 1.2 0.8 0.7 0.7IndiaGDP at market prices (2000 US$)b 5.5 9.2 9.7 9.0 6.3 5.8 7.7Current account balance/GDP (%) �1.2 �1.0 �1.0 �1.2 �3.1 �1.7 �1.9NepalGDP at market prices (2000 US$)b 5.0 3.1 3.7 2.6 5.5 3.8 4.9Current account balance/GDP (%) �6.3 0.0 �0.1 �1.2 1.2 1.0 0.8PakistanGDP at market prices (2000 US$)b 3.9 7.7 6.2 6.0 6.0 3.0 4.5Current account balance/GDP (%) �3.7 �3.3 �5.4 �5.8 �8.1 �4.6 �3.2Sri LankaGDP at market prices (2000 US$)b 5.2 6.0 7.7 6.8 6.3 4.0 5.5Current account balance/GDP (%) �4.6 �3.2 �5.3 �4.4 �7.5 �5.7 �5.5

Source: World Bank.

Note: Growth and Current Account figures presented here are World Bank projections and may differ from targets contained inother World Bank documents. Afghanistan, Bhutan, and Maldives are not forecast because of data limitations.

a. Growth rates over intervals are compound average; growth contributions, ratios and the GDP deflator are averages.b. GDP measured in constant 2000 U.S. dollars.c. Estimate.d. Forecast.

10363_Pg141-180:10363_Pg141-180 11/29/08 7:13 AM Page 170

public sector firms, rationalizing governmentfinances, enhancing openness where it couldimprove stability (such as FDI), and improvingthe quality of physical and financial infrastruc-ture. Moving forward with existing plannedinvestment programs, which often take yearsto develop, will support current activity andbuild capacity for eventual recovery.

Given strong underlying growth dynamicsin South Asia, the negative feedback effects ofthe global financial crisis are expected to betemporary. A relatively rapid rebound is ex-pected in 2010, with a projected revival ofGDP growth to 7.2 percent by 2010. Privateconsumption and investment growth are fore-cast to gain steam, supported by strengtheningglobal demand and a rebound in consumerand business confidence. Commodity price de-creases are projected, which will support a re-duction of inflationary pressures within theSouth Asian economies. The region’s medianinflation rate will have peaked in 2008 at 9.7percent, although sustained pipeline pressureswill prevent a rapid easing, with inflationmoderating incrementally to 8 percent in 2009and 6 percent by 2010.

RisksGiven the synchronized and widespread natureof the current crisis, downside risks to thebaseline are pronounced. A more prolongedand pervasive credit crunch than envisioned inthe baseline would lead to a deeper global re-cession. That in turn would likely lead to out-right contraction of South Asia’s private fixedinvestment (compared with the sharp slowingof growth found in the baseline), driven in partby a crimping of FDI inflows. South Asia’s ex-ports would also likely contract instead ofslow, and remittances could compress sharply,especially were destination countries to sendmigrants home. With the growth slowdown,household incomes would decline and unem-ployment rise. Progress in poverty alleviationcould slow markedly. While the region’s bank-ing sector has not been exposed to the toxicdebt instruments that have plagued many high-income countries, in a downside scenario of

protracted sluggish growth and risk aversion,weaknesses in the region’s financial sectorcould emerge.

Whereas India holds sizable foreign ex-change reserves (despite recent draw-downs)to help weather more negative than expectedgrowth dynamics, other countries in the re-gion have seen widening trade deficits andcapital outflows reduce their reserve holdings,increasing their vulnerability to sustainedpressure on currencies. Countries holding sub-stantial short-term debt obligations would bemore vulnerable. In the Maldives, the rapidbuildup of debt obligations with the construc-tion boom following the tsunami is of con-cern. Sri Lanka has a large public debt (equiv-alent to 83 percent of GDP in 2007), with44 percent of the debt external, albeit primarilyconcessional; the country’s fiscal position isthus vulnerable to higher interest rates and ex-change rate depreciation. The Sri Lankan cur-rency peg against the dollar could come underpressure, because the foreign reserve cover isrelatively low. Bhutan also holds significantexternal debt obligations, but these are heldprimarily by India for the development ofhydroelectric power, which Bhutan is in turnexporting to India. Should a deeper crisis leadto a falloff in foreign assistance, countries sig-nificantly reliant on aid (such as Afghanistan)would be more adversely affected.

In contrast, should the current global fi-nancial crisis be resolved relatively quickly,and growth dynamics prove more favorablethan projected, policy makers would face verydifferent challenges. Inflationary pressurescould return to the forefront—as counter-cyclical measures could become effectivelypro-cyclical—leading to higher internal and ex-ternal deficits, hindering investment (throughcrowding out), and acting as a drag on growth.

Sub-Saharan AfricaRecent Developments

Sub-Saharan Africa’s economy expanded5.4 percent in 2008, the first time in more

than 45 years that growth exceeded 5 percent


171

10363_Pg141-180:10363_Pg141-180 11/29/08 7:13 AM Page 171

for five years in succession—this despite sub-stantial deterioration in the external environ-ment during the year. GDP gains have beenbroad-based and less volatile, even in oil-im-porting economies, as strong commodity ex-port revenues and capital inflows underpinneddomestic demand. Another notable and en-couraging feature of the recent growth spurt isthe sustained contribution of fixed investmentto growth, which carries positive implicationsfor long-term potential growth.

Strong external demand, high commodityprices, and relatively robust private capital in-flows invigorated growth across a large spec-trum of economies, whether resource rich orresource poor. Oil-importing economies, out-side of South Africa, grew 5.2 percent in 2008,down from 5.8 percent in 2007, while oil-ex-porting countries grew by more than 7.5 per-cent for a second consecutive year. However,several years of above-trend economic expan-sion have pushed a larger number of Africaneconomies up against capacity constraints

stemming from inadequate investment in en-ergy, roads, railways, and ports over the pastdecades. This constraint along with high foodand fuel prices has contributed to the upturnin inflation witnessed across the subcontinentduring the year (table A11).

South African growth easesGrowth in the Republic of South Africatrailed growth in other African economies in2008, slowing markedly to an estimated 3.4percent from 5.1 percent in 2007. Power out-ages plagued output growth in the mining sec-tor, and household consumption slowedsharply, undercut by slower growth of credit,falling asset prices, and higher food and fuelprices. The region’s largest economy has feltthe repercussions of the intensification of thefinancial crisis since September 15. Increasedrisk aversion vis-à-vis emerging marketscaused asset prices in South Africa to plum-met, putting pressure on the rand, which hasdepreciated nearly 25 percent in nominal


172

Table A11 Sub-Saharan Africa forecast summary(annual percent change unless indicated otherwise)

1991–2000a 2005 2006 2007 2008e 2009f 2010f

GDP at market prices (2000 US$)b 2.3 5.9 5.9 6.3 5.4 4.6 5.8GDP per capita (units in US$) �0.5 3.4 3.4 4.3 3.4 2.7 3.8PPP GDPc 3.2 6.2 6.1 6.7 5.7 4.9 6.1


Net exports, contribution to growth 0.1 �2.3 �3.1 �2.9 �1.2 �0.8 �1.6Current account balance/GDP (%) �2.0 2.4 0.7 �0.3 1.0 �3.5 �3.7GDP deflator (median, LCU) 10.2 7.2 7.3 6.3 8.6 6.5 4.1Fiscal balance/GDP (%) �4.7 0.2 1.0 �1.9 �0.6 �1.3 �1.5

Memo items: GDPSub-Saharan Africa excluding South Africa 2.6 6.4 6.2 7.0 6.6 5.7 6.6Oil exporters 2.0 7.5 6.8 8.2 7.8 6.6 7.3CFA countries 2.5 4.0 2.4 3.4 4.5 4.3 5.0

South Africa 1.8 5.0 5.4 5.1 3.4 2.8 4.4Nigeria 2.8 7.2 5.2 6.5 6.3 5.8 6.2Kenya 1.9 5.7 6.1 7.1 3.3 3.7 5.9

Source: World Bank.

a. Growth rates over intervals are compound averages; growth contributions, ratios, and the GDP deflator are averages.b. GDP measured in constant 2000 U.S. dollars.c. GDP measured at PPP exchange rates.d. Exports and imports of goods and nonfactor services.e. Estimate.f. Forecast.

10363_Pg141-180:10363_Pg141-180 11/29/08 7:13 AM Page 172

effective terms since the beginning of 2008.Like most emerging markets, South Africa sawspreads on its sovereign bonds surge, by morethan 450 basis points between the beginning ofSeptember and mid November 2008; equityprices plummeted 40 percent in dollar termsover the same period (in local currency the losswas 21.7 percent) (figure A16).

Helped by higher exports of gold and plat-inum, South Africa’s current account deficitretreated to 7.3 percent of GDP in the secondquarter of 2008, from 8.9 percent in the pre-vious quarter. Lower prices for its main ex-ports, together with weaker external demand,will cause the current account deficit to rise to8 percent of GDP this year from 7.3 percent in2007. With portfolio inflows financing three-fourths of South Africa’s current accountdeficit in 2007, and with the increased volatil-ity of these inflows, South Africa may find itdifficult to finance its large current accountdeficit, especially if FDI inflows are alsofalling. Meanwhile, fiscal financing require-ments are much more modest; South Africa isexpected to run a small budget deficit in 2009,after being almost in-balance in 2008. Fur-thermore, government indebtedness remainslow, with debt at 23.3 percent of GDP as ofMarch 2008, and foreign debt accounting forless than 20 percent of total. This means thatthe government has room to borrow domesti-cally to finance countercyclical policies.

Higher fiscal spending, a slowdown in FDI,and drying up of credit suggests that privateinvestment growth, which has already been af-fected by tighter monetary policy, will slowfurther. The intensification of political ten-sions within the ruling African National Con-gress Party, which led to the resignation ofPresident Thabo Mbeki several months beforethe end of his term, is likely to have a minimaldirect impact on the economy but will add touncertainties faced by investors now worriedabout a possible shift in economic policy.

Outside of South Africa, large commoditywindfalls have fueled growth in resource-richcountries. Encouragingly, growth is spillingover to other sectors outside oil and mining, aspart of the windfall is spent. In Nigeria, thenon-oil economy is booming, despite contin-ued unrest in the Niger Delta that caused oiloutput to drop 11.2 percent in the secondquarter of 2008. Despite underperformance inthe oil sector, second-quarter growth acceler-ated to 6.7 percent, from 5.5 percent in thefirst, as output in non-oil industries picked upto 8.5 percent, mainly on strong growth inagriculture, trade, and telecoms, which to-gether accounted for 95 percent of non-oilgrowth. In Angola, GDP growth remained ro-bust in the first half of the year, and growth inthe non-oil sector will approach 20 percentthis year, marginally down from 21.5 percentin 2007, as the construction, agriculture, andcommunication sectors continue to expand atan impressive pace.

High energy and agricultural prices andlower agriculture output caused by unfavor-able weather conditions have affected indus-trial output in some countries. Indeed, in manyWest African countries the food processing sec-tor has contracted due to lower agriculturaloutput and higher input costs. Surges in foodand fuel prices have pushed headline consumerprice inflation into double digits in almost halfof the countries in Sub-Saharan Africa, withmedian inflation moving rapidly to nearly 13percent as of September 2008; median food in-flation increased to more than 17.7 percent(figure A17). For example in Ethiopia, headline


173



200Sep. 12008

Sep. 212008

Sep. 112008

Oct. 12008

Oct. 112008

Oct. 212008

Oct. 312008

400

600

800

1000

1200

1400

Figure A16 Bond spreads for African countries jumped after mid-September

Gabon

South Africa

Ghana

10363_Pg141-180:10363_Pg141-180 11/29/08 7:13 AM Page 173

inflation surged as high as 64 percent in 2008,as food inflation breached 80 percent. In somecases, core inflation also accelerated, as second-round inflation effects through wage settle-ments incorporating expectations of higher in-flation were concluded. For example, in SouthAfrica, unit labor costs increased 10.5 percentin the second quarter of 2008, spurred by av-erage wage increases of 9.6 percent in the firstnine months of the year.

High import prices in conjunction, in somecases, with strong investment demand (invest-ment carries high import content in Africa) ledcurrent account balances to deteriorate inmore than one of every two countries during2008 relative to 2007. Thirteen of 44 countriesexperienced a worsening in excess of 2 percentof GDP, and 19 of 44 countries registered cur-rent account deficits in excess of 10 percent ofGDP. In Ghana, for example, the trade deficitbreached 26.2 percent of GDP in the secondquarter of 2008 and is expected to reach morethan 30 percent in 2009. The deficit excludingofficial transfers is likely to rise to more than17 percent of GDP.

Political instability can still derail growth,as it did in Kenya, where output contracted0.8 percent in the first quarter of 2008 (year-over-year). Political tensions caused a sharpcontraction in tourism arrivals and in theagriculture sector. Other sectors were also af-

fected to varying degrees, with transport con-tracting 2.2 percent; strong performance inmining and construction prevented a moredismal growth outcome. A resumption ofconflict is also threatening growth prospectsin the Democratic Republic of Congo.

Medium-term outlookThe rapid and marked deterioration in the ex-ternal environment will cause growth in Sub-Saharan Africa to slow, coming in at 4.6 per-cent in 2009, a pace below 5 percent for onlythe first time in five years (see table A11, ear-lier). Direct effects of the global financial andeconomic crisis are likely to be much more lim-ited than in other regions, because Africaneconomies are less integrated into the interna-tional financial system and rely relatively lesson international capital and bond markets tofinance investment.

For Africa, weaker external demand andlower commodity prices will be the majormechanisms through which the financial crisiswill be transmitted. Declines in demand in keyexternal markets will take a toll on exports,and the contribution of trade to GDP growthis likely to be negative in 2009. Perhaps moreimportantly, export revenues will be affectedby markedly lower commodity prices nextyear, eroding government and corporate fi-nances and affecting farmers’ incomes ad-versely. Additional adverse factors coming toaffect Sub-Saharan Africa, potentially withsome lag, are a slowing pace of worker remit-tance receipts, and importantly for many low-income countries, possible moderation in Of-ficial Development Assistance (ODA) flows.

More of an issue for commodity-rich coun-tries, gross portfolio flows to the region are ex-pected to fall markedly as credit becomes scarceand more expensive and investors’ risk aversionintensifies. Official aid may also be squeezed byreduced fiscal space in donor countries as theytackle financial crises at home. As a result, frag-ile countries that rely heavily on aid are facedwith a potential deterioration in growthprospects. Moreover, recession in high-incomecountries will undermine tourism arrivals and


174

Median CPI inflation (year-on-year)

Source: World Bank.

4

8

12

20

16

Figure A17 African headline inflation jumpsas food prices skyrocket

Jul.2005

Jul.2006

Jul.2007

Jan.2005

Jan.2006

Jul.2008

Jan.2007

Jan.2008

Headline consumer price inflation

Food

10363_Pg141-180:10363_Pg141-180 11/29/08 7:13 AM Page 174

revenues, as well as remittances, which repre-sent a significant share of GDP for Cape Verde,the Gambia, Kenya, Liberia, Lesotho, and theSeychelles, among other countries. However,for countries where currencies depreciatedheavily with respect to donor country curren-cies, receipts in local currency terms could stillincrease.

Countries with very large current accountdeficits, including Burundi, Eritrea, the Gambia,Ghana, Madagascar, Malawi, Rwanda, Togo,and the Seychelles will need to adjust domesticdemand to lessen import growth as financingexternal deficits becomes more difficult, andas export revenues and transfers are dimin-ished by slower global growth. Many of theseeconomies are especially vulnerable becausethey have low levels of international reserves,in many cases covering less than three monthsof imports.

While oil exporters hold sufficient re-sources to weather the global economic down-turn, many oil-importing economies havebeen hit hard by higher food and fuel pricesand are less well equipped for the comingdownturn. In a number of cases, increasedsubsidies have limited fiscal space for counter-cyclical spending. Over the past year, as foodprices surged, many governments removed orsuspended tariffs on imported foods, whichundercut tariff revenues. In addition, a less-than-full pass-through of higher internationaloil prices led to a large increase in fuel subsi-dies in some countries, further reducing fiscalroom. While the movement to lower food andfuel prices will bring some relief, countries re-main in a weakened state.

Overall, aggregate GDP growth in Sub-Saharan Africa is projected to decline to 4.6percent in 2009 from 5.4 percent in 2008, onthe back of weaker investment outlays, falter-ing export performance, and softer privateconsumption. As external demand graduallyrecovers over the second half of 2009 and into2010, growth should firm to 5.8 percent bythe latter year. Excluding South Africa andNigeria, growth is projected to ease by a fullpercentage point to 5.6 percent in 2009 and to

bounce back to 6.7 percent by 2010. In oil-exporting economies, growth will slow bymore than a full percentage point to 6.6 per-cent, but the exporters will remain the fastest-growing group of countries in the region,while growth in oil-importing countries out-side South Africa is projected to ease to 4.6 per-cent, a rate still above the historical trend(table A12).

South Africa’s economic growth is likely toweaken further in 2009, falling below 3 per-cent for the first time in almost a decade, astighter monetary policy and high inflationcauses household consumption to falter. Pri-vate investment growth will continue to decel-erate, pushed down by tighter credit marketsand as demand in main export markets con-tracts (figure A18). Large asset price declinesand associated negative wealth effects, alongwith slower credit creation will underminehousehold consumption, and together withweaker external demand will cut into manu-facturing output. Although investment growthin South Africa is projected to ease in 2009, itwill still remain one of the engines of growthfor the country, as the South African govern-ment continues to bring forth large projects inthe energy sector to address the chronic elec-tricity deficit and in infrastructure ahead ofthe 2010 World Football Cup. The falloff inSouth Africa’s GDP growth will carry reper-cussions for neighboring economies that tradeheavily with South Africa and receive remit-tances from expatriate workers in SouthAfrica.

RisksWith the world economy at a crossroads, risksfacing Sub-Saharan Africa have intensified. Ifthe concerted efforts of policy makers aroundthe globe fail to re-establish trust in the inter-national financial system, the world economyrisks a deeper and more prolonged recession.As a result, Sub-Saharan Africa’s growthwould drop more sharply than envisaged inthe base forecast.

Among African countries, South Africa isprobably the country most directly exposed to


175

10363_Pg141-180:10363_Pg141-180 11/29/08 7:13 AM Page 175


176

Table A12 Sub-Saharan Africa country forecasts(annual percent change unless indicated otherwise)

1991–2000a 2005 2006 2007 2008e 2009f 2010f

AngolaGDP at market prices (2000 US$)b 0.8 20.6 18.6 24.7 15.8 11.1 11.3Current account balance/GDP (%) �6.1 16.8 20.9 14.1 17.8 8.9 9.2BeninGDP at market prices (2000 US$)b 4.8 2.9 3.8 4.7 5.0 5.1 5.7Current account balance/GDP (%) �6.8 �6.3 �7.3 �7.0 �8.6 �9.5 �9.5BotswanaGDP at market prices (2000 US$)b 6.2 4.0 2.1 6.0 4.7 4.1 4.5Current account balance/GDP (%) 8.1 15.3 18.2 19.3 9.5 7.4 5.2Burkina FasoGDP at market prices (2000 US$)b 4.0 7.1 5.5 3.6 4.1 4.6 5.5Current account balance/GDP (%) �5.6 �12.4 �11.7 �13.1 �14.2 �12.8 �12.2BurundiGDP at market prices (2000 US$) b �1.7 0.9 5.1 3.4 4.4 4.0 5.1Current account bal/GDP (%) �3.4 �28.4 �36.0 �37.6 �38.9 �36.2 �36.7Cape VerdeGDP at market prices (2000 US$)b 5.8 11.9 10.8 6.5 6.7 5.1 6.3Current account bal/GDP (%) �8.3 �3.5 �9.6 �15.5 �13.9 �11.9 �12.8CameroonGDP at market prices (2000 US$)b 1.4 2.0 3.2 3.4 3.9 4.0 4.4Current account bal/GDP (%) �2.9 �2.4 �2.1 �2.3 �0.5 �4.3 �4.6Central African RepublicGDP at market prices (2000 US$)b 1.6 2.2 4.1 3.8 3.4 4.2 4.8Current account balance/GDP (%) �4.3 �7.1 �7.4 �7.3 �7.0 �6.7 �7.1ChadGDP at market prices (2000 US$)b 2.3 7.9 �0.5 0.7 1.6 2.8 3.0Current account balance/GDP (%) �5.5 �6.3 �7.3 �8.2 �3.8 �7.0 �8.8ComorosGDP at market prices (2000 US$)b 1.1 4.2 0.5 1.8 0.6 1.2 2.5Current account balance/GDP (%) �6.8 �4.6 �5.5 �4.8 �7.2 �6.3 �6.8Congo, Dem. Rep.GDP at market prices (2000 US$)b �5.6 6.5 5.6 6.3 10.7 8.3 11.9Current account balance/GDP (%) 2.0 �10.0 �9.8 �12.2 �11.9 �13.4 �10.1Congo, Rep.GDP at market prices (2000 US$)b 1.4 7.7 6.2 �1.4 9.1 7.4 9.7Current account balance/GDP (%) �16.5 15.1 �3.9 �23.0 6.8 2.6 11.6Cote d’IvoireGDP at market prices (2000 US$)b 2.3 1.2 0.9 1.5 2.6 3.1 4.9Current account balance/GDP (%) �4.0 0.2 3.4 �0.2 0.7 �3.2 �3.8EritreaGDP at market prices (2000 US$)b — 4.8 �1.0 1.3 1.2 2.0 4.2Current account balance/GDP (%) — �26.1 �29.5 �30.2 �32.8 �22.1 �19.0EthiopiaGDP at market prices (2000 US$)b 2.9 10.2 11.5 11.1 8.8 6.0 7.3Current account balance/GDP (%) �0.8 �13.7 �12.3 �11.8 �11.6 �8.9 �8.6GabonGDP at market prices (2000 US$)b 1.7 3.0 1.3 5.4 3.7 4.2 4.0Current account balance/GDP (%) 5.6 15.9 15.5 15.4 19.6 10.3 8.1Gambia, TheGDP at market prices (2000 US$)b 3.3 5.0 6.5 6.4 5.3 4.5 5.4Current account balance/GDP (%) �1.6 �10.9 �13.8 �16.9 �19.8 �17.4 �16.3GhanaGDP at market prices (2000 US$)b 4.3 5.9 6.2 6.5 6.0 5.6 6.0Current account balance/GDP (%) �6.5 �10.3 �12.6 �14.2 �17.1 �15.6 �16.0

(continued)

10363_Pg141-180:10363_Pg141-180 11/29/08 7:13 AM Page 176


177

GuineaGDP at market prices (2000 US$)b 4.1 3.3 2.8 1.8 4.3 3.7 4.6Current account balance/GDP (%) �5.6 �4.9 �1.7 �2.4 �4.9 �7.1 �7.2Guinea-BissauGDP at market prices (2000 US$)b 1.5 3.5 4.2 2.7 2.9 2.8 3.4Current account balance/GDP (%) �24.0 �7.2 �19.3 �15.9 �10.9 �12.1 �10.8KenyaGDP at market prices (2000 US$)b 1.9 5.7 6.1 7.1 3.3 3.7 5.9Current account balance/GDP (%) �1.6 �1.4 �2.3 �3.7 �8.0 �6.4 �5.5LesothoGDP at market prices (2000 US$)b 3.5 2.9 7.2 4.9 4.1 3.2 4.2Current account balance/GDP (%) �13.4 �6.9 0.7 �0.2 1.8 4.6 4.5MadagascarGDP at market prices (2000 US$)b 1.7 4.6 4.9 6.3 6.8 6.0 10.4Current account balance/GDP (%) �7.8 �12.4 �9.4 �14.2 �21.0 �15.7 �4.8MalawiGDP at market prices (2000 US$)b 3.4 2.7 8.2 8.4 7.9 6.5 7.9Current account balance/GDP (%) �8.5 �11.8 �19.3 �17.9 �20.7 �19.8 �19.6MaliGDP at market prices (2000 US$)b 4.0 6.1 5.3 3.1 5.1 3.9 5.1Current account balance/GDP (%) �8.9 �8.2 �6.4 �9.9 �9.1 �10.5 �10.4MauritaniaGDP at market prices (2000 US$)b 2.9 5.4 11.6 0.9 2.1 5.9 6.4Current account balance/GDP (%) �0.3 �49.0 �2.8 �4.9 �4.4 �9.4 �10.7MauritiusGDP at market prices (2000 US$)b 5.3 4.6 3.5 5.4 5.0 3.8 5.3Current account balance/GDP (%) �1.6 �5.0 �10.0 �8.4 �10.0 �8.2 �8.3MozambiqueGDP at market prices (2000 US$)b 5.0 8.4 8.7 7.0 6.0 6.3 6.4Current account balance/GDP (%) �16.4 �11.6 �9.0 �15.8 �17.0 �16.0 �16.3NamibiaGDP at market prices (2000 US$)b 4.2 4.7 4.1 3.8 3.6 3.1 4.5Current account balance/GDP (%) 3.1 4.3 18.8 20.9 23.0 21.3 20.1NigerGDP at market prices (2000 US$)b 1.8 7.2 5.1 3.2 4.9 3.6 4.9Current account balance/GDP (%) �6.9 �9.1 �9.2 �10.9 �14.5 �13.7 �15.4NigeriaGDP at market prices (2000 US$)b 2.8 7.2 5.2 6.5 6.3 5.8 6.2Current account balance/GDP (%) �0.8 31.5 22.5 21.8 20.3 7.0 4.6RwandaGDP at market prices (2000 US$)b 0.2 6.0 5.5 6.0 8.0 5.0 5.5Current account balance/GDP (%) �1.2 �3.9 �15.7 �15.6 �21.7 �15.7 �16.7SenegalGDP at market prices (2000 US$)b 3.1 5.6 2.3 4.6 4.5 4.7 5.9Current account balance/GDP (%) �5.7 �6.5 �9.3 �11.7 �14.4 �12.7 �12.8SeychellesGDP at market prices (2000 US$)b 4.5 1.2 5.3 7.3 2.3 0.5 3.0Current account balance/GDP (%) �7.4 �29.0 �22.6 �31.4 �30.2 �27.4 �23.8Sierra LeoneGDP at market prices (2000 US$)b �4.7 7.3 7.4 6.4 5.8 5.1 6.5Current account balance/GDP (%) �9.0 �14.3 �8.8 �7.2 �8.3 �9.6 �9.7South AfricaGDP at market prices (2000 US$)b 1.8 5.0 5.4 5.1 3.4 2.8 4.4Current account balance/GDP (%) �0.2 �4.0 �6.5 �7.3 �8.0 �8.1 �8.3SudanGDP at market prices (2000 US$)b 5.8 8.6 11.8 10.1 10.3 8.0 8.1Current account balance/GDP (%) �8.2 �10.8 �13.6 �10.7 �6.8 �8.9 �9.0

1991–2000a 2005 2006 2007 2008e 2009f 2010f

(continued)

10363_Pg141-180:10363_Pg141-180 11/29/08 7:13 AM Page 177


178

SwazilandGDP at market prices (2000 US$)b 3.1 2.3 2.1 3.2 2.2 1.8 1.9Current account balance/GDP (%) �2.4 3.3 5.9 9.5 10.7 10.7 9.3TanzaniaGDP at market prices (2000 US$)b 2.9 6.8 6.2 7.1 7.2 6.3 7.0Current account balance/GDP (%) �12.5 �7.0 �12.9 �12.7 �14.2 �12.7 �12.9TogoGDP at market prices (2000 US$)b 2.2 1.2 4.1 2.3 0.8 2.4 3.3Current account balance/GDP (%) �8.5 �21.8 �17.6 �16.0 �22.1 �16.9 �16.9UgandaGDP at market prices (2000 US$)b 6.8 6.7 8.4 8.9 7.9 5.9 7.6Current account balance/GDP (%) �7.0 �4.8 �7.6 �8.0 �8.8 �9.3 �9.9ZambiaGDP at market prices (2000 US$)b 0.7 5.2 6.2 6.2 6.1 4.6 6.0Current account balance/GDP (%) �10.6 �10.0 �7.3 �7.2 �5.5 �8.1 �9.5ZimbabweGDP at market prices (2000 US$)b 0.9 �5.3 �4.2 �6.3 �4.9 �2.1 �2.1Current account balance/GDP (%) �7.5 28.5 30.7 36.6 40.2 18.1 18.6

Source: World Bank.

Note: — � not available.

a. Growth rates over intervals are compound averages; growth contributions, ratios, and the GDP deflator are averages.b. GDP measured in constant 2000 U.S. dollars.c. Growth and current account figures presented here are World Bank projections and may differ from targets contained in otherWorld Bank documents.

d. Liberia, Somalia, Sao Tomé and Príncipe, are not forecast because of data limitations. Current account balances excludeofficial transfers. In SACU members they include nonduty SACU transfers.

e. Estimate.f. Forecast.

Table A12 (continued )(annual percent change unless indicated otherwise)

1991–2000a 2005 2006 2007 2008e 2009f 2010f

the current global financial turmoil. SouthAfrica’s risk of financial contagion, however,is limited by low exposure to “toxic” assetsand foreign currency risks. However, a “flightto quality” could cause large portfolio out-flows, which would imperil the country’s abil-ity to finance its large current account deficit,which in turn could trigger sharp depreciationof the rand and higher inflation.

Sub-Saharan African countries that are lessintegrated with international financial andcapital markets would suffer more from lowerexternal demand, dwindling tourism revenues,remittances, or aid. Commodity prices wouldfall further in such a scenario, causing exportrevenues in many countries to fall sharply anderoding fiscal positions, corporate profitabil-ity, and incomes. Vulnerability to externalshocks, including terms-of-trade shocks, has

Figure A18 Economic growth to slow inSub-Saharan Africa

5

6

7

8

9

0

1

2

3

4

2009 2010

Oil importers, excluding South Africa

Oil exporters

South Africa

2005 2006

GDP growth (%)

2007 2008

Source: World Bank.

10363_Pg141-180:10363_Pg141-180 11/29/08 7:13 AM Page 178


179

increased over the past couple of years, as ex-ternal and fiscal balances have deteriorated inmany countries. Several countries havereached a point where external imbalances areunsustainable and a shut-off of financing, orlarge negative terms-of-trade shocks, couldlead to balance of payments and currencycrises, with adverse consequences for hard-won macroeconomic stability and long-termgrowth. Moreover, fragile economies that relyheavily on external aid and face daunting re-construction and stabilization challenges willsee their efforts to normalize the situation de-railed by lack of sufficient external financing.

A sharp deceleration in growth would havesignificant consequences for poverty reductionin Africa. According to Arbache and Page(2007), had the region avoided some of thesharpest declines in per capita GDP growth,overall growth would have been 1 percentagepoint faster every year for the past threedecades. Another risk is that the large-scale in-jection of liquidity into the global financialsystem comes to fuel inflation if monetary au-thorities fail to reverse polices at the first signsof a turnaround.

Notes1. A downturn in the global high-tech cycle, in

which East Asia plays a key role in the production andexport of higher-tech goods, also contributed to theslowing of trade growth.

2. Russia currently holds somewhat less than$500 billion in reserves, along with two large oil funds(about $140 billion and $50 billion, respectively, as ofSeptember 2008); it also has ample fiscal and currentaccount surpluses (the latter, $90 billion, or 8 percentof GDP as of September 2008).

3. Core inflation is calculated as headline CPI net offood, household energy, and transport fuels. Data forBelarus, Russia, and Turkey, where detailed sub-indexes are not available, are from official sources di-rectly, which may use different definitions and calcula-tion methods.

4. This report covers the developing (that is, low-and middle-income) countries of the Middle East andNorth Africa region, and thus excludes high-incomeeconomies Bahrain, Kuwait, Qatar, Saudi Arabia, andthe United Arab Emirates. In addition, for a number of

middle- and high-income countries, the availability ofeconomic data is insufficient for inclusion in the report;these include Djibuti, Iraq, Libya, and the West Bankand Gaza. For recent developments and the outlook fora broader range of Middle Eastern and North Africaneconomies, see Economic Developments and Prospects,2008, The Middle East and North Africa RegionWorld Bank, 2008.

5. Several GCC countries have responded forcefullyto head off financial contagion. Qatar on October 13launched a $5.3 billion plan to purchase up to 20 per-cent of shares in banks listed on the Doha stock ex-change. This followed by one day an announcementthat United Arab Emirates would guarantee all de-posits and savings in national banks, as well as all in-terbank operations in the Emirates. And Saudi Arabiacut interest rates in line with the Federal Reserve, whileindicating that some $40 billion would be made avail-able to local banks.

6. National income and product account figures arepresented in calendar years, although originally re-ported in fiscal years by Bangladesh, India, Nepal, andPakistan. For example, fiscal year 2007/08 is reportedas calendar year 2007 for India, and as 2008 forBangladesh, Nepal, and Pakistan, due to differences inthe timing of their fiscal years.

7. Nepal’s currency is pegged to the Indian rupee.8. Source: Boston Analytics Consumer Sentiment

Index (BACSI), which is based on a monthly surveytargeting Indian consumers across 11 cities (Delhi,Mumbai, Kolkata, Chennai, Hyderabad, Bengaluru,Nagpur, Kochi, Lucknow, Chandigarh, and Jaipur).See http://www.bostonanalytics.com/news.html.

ReferencesArbache, Jorge Saba, and Page, John (2007). “More

Growth or Fewer Collapses? A New Look atLong Run Growth in Sub-Saharan Africa.” PolicyResearch Working Paper 4384, World Bank.

Calvo, Guillermo, and Ernesto Talvi. 2007. “CurrentAccount Surplus in Latin America: Recipe againstCapital Market Crises.” http://www.rgemonitor.com/latam-blog/58/current_account_surplus_in_latin_america_recipe_against_capital_market_crise.

DJF (Dow Jones Factiva). 2008. http://www.factiva.com.Fajnzylber, Pablo, and J. Humberto Lopez. 2008. “The

Development Impact of Remittances in LatinAmerica.” In Remittances and Development:Lessons from Latin America, ed. Pablo Fajnzylberand J. Humberto Lopez. Washington, DC: WorldBank.

ISI (ISI Emerging Markets). 2008. http://www.securities.com.

10363_Pg141-180:10363_Pg141-180 11/29/08 7:13 AM Page 179

Izquierdo, Alejandro, Randall Romero, and ErnestoTalvi. 2008. “Booms and Busts in Latin America:The Role of External Factors.” Research Depart-ment Working Paper 631. Inter-American Devel-opment Bank, Washington, DC. http://www.imf.org/external/np/seminars/eng/2007/whd/pdf/session1-2.pdf.

Österholm, Pär, and Jeromin Zettelmeyer. 2007. “TheEffect of External Conditions on Growth in LatinAmerica.” Working Paper 07/176. InternationalMonetary Fund, Washington, DC. http://www.imf.org/external/pubs/ft/wp/2007/wp07176.pdf.

World Bank. 2008a. Migration and Remittances Fact-book 2008. Washington, DC: world Bank.

______. 2008b. “Rising Food Prices: The World Bank’sLatin America and Caribbean Region PositionPaper.” Washington, DC.

______. 2008c. “Shockwaves from the North: LatinAmerica and the External Deterioration.” ChiefEconomist Office, Latin America and the CaribbeanRegion. http://siteresources.worldbank.org/EXTLACOFFICEOFCE/Resources/870892-1197314973189/Shockwaves.pdf.


180

10363_Pg141-180:10363_Pg141-180 11/29/08 7:13 AM Page 180

Eco-AuditEnvironmental Benefits Statement

The World Bank is committed to preserv-ing endangered forests and natural re-sources. The Office of the Publisherhas chosen to print Global EconomicProspects 2009 on recycled paper with30 percent post-consumer waste, in accor-dance with the recommended standardsfor paper usage set by the Green PressInitiative, a nonprofit program sup-porting publishers in using fiber that isnot sourced from endangered forests.For more information, visit www.greenpressinitiative.org.

Saved:• 31 trees• 1,466 pounds ofsolid waste

• 11,419 gallonsof wastewater

• 2,751 pounds ofnet greenhousegases

• 22 millionBritish thermalunits of totalenergy

10363_green_page:10363_green_page 11/29/08 7:14 AM Page 1

The eruption of the worldwide financial crisis has radically recast prospects for the world economy. Global Economic Prospects 2009: Commodities at the Crossroads analyzes the implications of

the crisis for low- and middle-income countries, including an in-depth look at long-term prospects for global commodity markets and the policies of both commodity producing and consuming nations.

Developing countries face sharply higher borrowing costs and reduced access to capital. This will cut into their capacity to finance investment spending—ending a five-year stretch of developing-country growth in excess of 6 percent annually. The looming recession presents new risks, coming as it does on the heels of the recent food and fuel crisis.

Commodity markets, meanwhile, are at a crossroads. Following decades of low prices and weak investment in supply capacity, commodity prices first spiked—spurred on by five years of very fast developing-country growth—and have now plummeted in response to the financial crisis.

In the longer run, commodities are not expected to be in short supply. Prices should be higher than they were in the 1990s but much lower than in the recent past. These higher prices should provide producers with sufficient incentive to discover new supplies, improve output from existing resources, and promote greater conservation and substitution with more abundant alternatives. At the same time, slower population growth will ease the pace at which commodity demand grows. Policies to limit carbon emissions and boost agricultural investment, along with the dissemination of efficient techniques, should also contribute to this long-term outcome.

This year’s Global Economic Prospects also looks at government responses to the recent price boom. Producing-country governments have saved more of their windfall revenues, and are therefore less likely to be forced to cut into spending now that prices have declined. The spike in food prices tipped more people into poverty, which led governments to expand social assistance programs. These programs need to be better targeted to the needs of the very poor so that governments can respond effectively the next time there is a crisis.

For additional information, please visit www.worldbank.org/prospects. An online companion to the prospects section of this report, including access to additional data and analysis not reported here, is also available at www.worldbank.org/globaloutlook.

“While developing countries entered this tumultuous

period with much improved fundamentals, this crisis is

expected to test severely both them and the international

financial system. In the longer run, even after developing-

country growth recovers, commodity supply should keep

pace with demand, but policy will need to foster conservation

efforts and technological progress. In particular, if poor

countries are to maintain domestic food self-sufficiency,

governments will need to strengthen investment in rural

infrastructure, agricultural research, and technological

outreach.”

—Justin Yifu LinSenior Vice President and


2009Commodities at the Crossroads 2009



SKU 17799

ISBN 978-0-8213-7799-4

Global E

conomic Prospects

2009

10363_Cover 1 11/29/08 7:15:41 AM

Global Global Economic Prospects Economic...

Documents

Transcript of Global Global Economic Prospects Economic...