Global Economic Prospects 2008

Technology Diffusion in the Developing World

Global Economic Prospects

Global Economic Prospects

20082008

GlobalEconomicProspects

GlobalEconomicProspects

Technology Diffusion in the Developing World

2008

© 2008 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 11 10 09 08

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: 978-0-8213-7365-1eISBN: 978-0-8213-7366-8DOI: 10.1596/978-0-8213-7365-1

ISSN: 1014-8906

Cover photos: Irrigation by Chris Stowers/Panos; Man with Cell Phone by JacobSilberberg/Panos; Train by Qilai Shen/Panos; Map Projection by Chris Stowers/Panos; andResearcher by Jenny Matthews/Panos.Cover design: Critical Stages

The cutoff date for the data used in this report was December 12, 2007. Dollars are currentU.S. dollars unless otherwise indicated.

Foreword xi

Acknowledgments xiii

Abbreviations xv

Overview 1Technological achievement and diffusion in developing countries 2Some policy directions 13Note 15References 15

Chapter 1 Prospects for Developing Countries 17Growth outlook 17Risks 18Financial markets: Needed correction or major disruption? 18Global growth 21World trade 33Inflation and commodity markets 36Risks and uncertainties: Danger of a banking crisis and a U.S. recession 41Long-term prospects and poverty forecasts 43Notes 48References 49

Chapter 2 Technology and Technological Diffusion in Developing Countries 51The role of technology in development 53Measuring technology in developing countries 58Evaluating overall technological progress 78Technological diffusion over the long term 87Conclusion 92Technical Annex: Construction of the summary indexes 92Notes 99References 101

v

Contents

Chapter 3 Determinants of Technological Progress: Recent Trends and Prospects 105Drivers of technological progress: A framework 107External transmission channels 109Nurturing technological adaptive capacity 127Conclusion 150Notes 153References 156

Appendix: Regional Economic Prospects 165East Asia and the Pacific 165Europe and Central Asia 170Latin America and the Caribbean 176Middle East and North Africa 184South Asia 189Sub-Saharan Africa 193

Figures1 Robust growth among developing countries should cushion the developed country

slowdown 22 Scientific innovation and invention is almost exclusively a high-income activity 33 Technological achievement: Converging, but the gap remains large 44 The penetration of older and more recent technologies depends on more

than income 55 Technological achievement tends to level off at different income levels in different

regions 66 Most technologies fail to penetrate deeply into developing economies 77 The urban–rural gap in telephone access in India is huge 78 Domestic absorptive capacity both conditions and attracts external flows 89 Developing countries’ trade in technology goods has risen 10

10 Macroeconomic stability has improved since the early 1990s 1111 Literacy rates have increased in all regions 1212 Developing regions have much poorer governance than do OECD countries 131.1 The perceived riskiness of high-yield corporate bonds increased more than that

of emerging market bonds 191.2 Emerging market asset sell-off more severe than during earlier periods of market

turbulence 191.3 Global equity markets fall, then recover led by emerging markets 201.4 A step-down in growth in 2008 211.5 Volatile patterns of growth among OECD countries 231.6 Tighter credit and weak housing yield slower U.S. growth 231.7 Robust growth in developing country industrial production 241.8 Developing growth retains strong momentum during the first half of 2007... 261.9 ...with growth moderating through 2009 261.10 East Asia now accounts for one-quarter of China’s imports 271.11 External positions vary widely across Europe and Central Asia 271.12 Growth eases in 2007 for the Latin America and Caribbean region 28

C O N T E N T S

vi

1.13 Continued oil revenue gains support growth among Middle East and North Africa oilexporters 30

1.14 South Asia growth is slowing as the Indian rupee appreciates 321.15 Oil exporters drive 2007 growth results for Sub-Saharan Africa 321.16 Weak U.S. growth reduces demand for developing country exports 351.17 Export opportunities for high-income countries 351.18 U.S. current account narrows over 2007 and is likely to continue doing so 361.19 Inflationary pressures are rising in the Middle East and North Africa and

Sub-Saharan Africa 371.20 Inflation is broadly stable elsewhere, though at high levels 371.21 Commodity prices continued gains through 2007 led by metals 381.22 Copper, zinc, and aluminum prices sharply affected by China 381.23 Growth in the world’s demand for oil slows 391.24 OPEC reduces output to support prices 391.25 Agricultural prices surge over 2006–07 401.26 A rise in food prices, led by a ramp-up of the prices of fats, oils, and grains 401.27 Long-term growth, 1980–2030 441.28 Declining capital-led growth for developed countries, 2002–30 451.29 Sustained high productivity growth for developing countries 452.1 Patent activity is rising in middle-income countries 612.2 Electrical consumption varies markedly even at similar income levels 632.3 Rail and road densities rise with income and population density 652.4 Telephone densities are highly correlated with income, but air

transport is not 662.5 The incidence of Internet use varies widely across countries 732.6 Logistics performance in the world 772.7 Distribution of technological achievement by dimension 802.8 Increase in summary technological achievement subindexes, 1990s–2000s 822.9 Alternative summary indexes of technological achievement 832.10 Technological achievement rises with income levels 842.11 Comparison of levels of technological achievement, early 1990s and

early 2000s 853.1 Domestic absorptive capacity both conditions and attracts external flows 1083.2 Rising share of high-tech imports 1123.3 Exports of low-, medium-, and high-technology goods 1143.4 Share of foreign affiliates in business R&D expenditure 1173.5 Licensing payments have risen sharply 1213.6 The brain drain is a severe problem in a number of small countries 1233.7 Share of Ph.D. students still living in the United States five years

after graduation 1243.8 High-skilled emigrants are disproportionately represented in the diaspora 1243.9 Most developing countries have increased their exposure to external

technology 1283.10 Number of countries in conflict worldwide 1293.11 Efficiency of contract enforcement 1323.12 Developing country governance scores relative to OECD average 1323.13 Regional averages of six governance indicators 1333.14 Per capita incomes have accelerated in recent years 134

C O N T E N T S

vii

3.15 Except in Sub-Saharan Africa, life expectancy is improving 1343.16 Educational expenditures have risen in some regions 1373.17 Many developing country students fail to meet literacy standards 1383.18 Levels of intellectual property protection 1463.19 Level of and recent changes in technological absorptive capacity 149A1 East Asian growth moves up in 2007 165A2 Except for China, inflation is now stabilizing across East Asia 166A3 Performance improves for East Asian countries other than China 169A4 Mixed growth outturns across Europe and Central Asia 171A5 External positions vary widely across Europe and Central Asia 171A6 Growth in Europe and Central Asia eases into 2009 173A7 Growth outturns were mixed across Latin America in 2007 176A8 Latin American inflation eases over the last 15 years 177A9 Latin America and the Caribbean sovereign bond spreads decline, then

increase again 177A10 Growth in Latin America and the Caribbean eases into 2009 179A11 Financial test: Credit 183A12 Exchange rate policy dilemmas? 183A13 Export product (value) concentration is increasing 183A14 Export market (value) concentration is falling 184A15 Growth in Middle East and North Africa picks up 184A16 Hydrocarbon exports continue to rise on higher prices, modest volume gains 186A17 Tourism and remittances offset widening trade deficits for Maghreb and

Mashreq countries 187A18 Middle East and North Africa equities rebound from the mid-2007 slump 189A19 South Asian economies ease into 2007 190A20 Monetary policy is tightened in response to a buildup in inflation 190A21 Growth in Sub-Saharan Africa has accelerated markedly. . . 194A22 . . . reaching a 35-year high in oil-exporting countries . . . 198A23 . . . and a 10-year high in oil-importing countries 199A24 Contributions of investment and consumption have increased 199

Tables1.1 Gross capital flows to developing countries, 2005–07 201.2 The global outlook in summary, 2005–09 221.3 Recent economic indicators, developing regions, 2005–07 251.4 Developments and prospects for world trade and payments 341.5 Poverty in developing countries by region, selected years 46 2.1 Disparity among TFP levels remains wide 542.2 Scientific and innovative outputs 612.3 Indicators of the diffusion of older technologies 642.4 Affordability of fixed-line phones falls rapidly with lower incomes 672.5 Immunization rates lag significantly in South Asia and Sub-Saharan Africa 682.6 Diffusion of both water and sanitation technology is low in rural areas 692.7 Diffusion of recent technologies 722.8 Share of high-tech products in total exports 732.9 The quality of logistics services in 2005 varies by income 77

C O N T E N T S

viii

2.10 Indicators included in summary indexes of technological achievement 792.11 Technological achievement in developing countries relative to that in high-income

countries 812.12 Increase in technological achievement in developing countries relative to that in

high-income countries 812.13 Overall technological progress in absolute and relative terms 862.14 Successful diffusion has accelerated 882.15 The pace at which technology diffuses has picked up among successful adaptors 892.16 Slow diffusion means that many developing countries never reach the 25 or 50 percent

threshold 90A2.1 Indicators used to calculate the summary indexes and overall index related to

technological achievement 95A2.2 Indicators used to calculate the summary indexes and overall index of technological

absorptive capacity 96A2.3 Share of total variance explained by principal components, technological achievement

index 96A2.4 Share of total variance explained by principal components, technological absorptive

capacity index 96A2.5 Share of total variance explained by principal components for each subgroup of

indicators 97A2.6 Factor loadings and variable weights for technological achievement

subgroups 98A2.7 Factor loadings and variable weights for technological absorptive capacity

subgroups 98A2.8 Share of total variance explained by main principal components of

technological achievement and technological absorptive capacity using the sub-indexes 99

A2.9 Factor loadings and variable weights obtained from second-stage principal components analysis (2000–03) 99

3.1 Trade in technology goods has increased in developing countries 1113.2 Foreign direct investment as a percent of GDP 1163.3 Foreign direct investment as a percent of fixed capital formation 1163.4 Selected purchases of high-tech firms by companies in developing countries,

early 2000s 1213.5 Increases in exposure to external technologies index, 1990s to 2000s 1293.6 Macroeconomic stability has improved in developing countries 1303.7 The regulatory burden is heavier in developing countries than in the OECD 1313.8 Educational attainment indicators 1353.9 Relatively high youth literacy rates 1363.10 Weak financial intermediation hinders technology in developing countries 1393.11 R&D intensities have increased 1413.12 Private-public sector R&D 141A1 East Asia and Pacific forecast summary 166A2 East Asia and Pacific country forecasts 168A3 Europe and Central Asia forecast summary 170A4 Europe and Central Asia country forecasts 174A5 Latin America and the Caribbean forecast summary 179A6 Latin America and the Caribbean country forecasts 180

C O N T E N T S

ix

A7 Middle East and North Africa forecast summary 185A8 Middle East and North Africa country forecasts 188A9 South Asia forecast summary 192A10 South Asia country forecasts 192A11 Sub-Saharan Africa forecast summary 194A12 Sub-Saharan Africa country forecasts 195

Boxes1 Summary of empirical results 141.1 Developing country exports in the wake of the removal of barriers to

Chinese exports 311.2 Biofuels 411.3 Policy responses to rising food prices 422.1 Technology can contribute to welfare without affecting measures of short-term

output 552.2 Technological innovation may spur further innovation in upstream and downstream

activities 562.3 Promoting appropriate technologies in Rwanda 572.4 Shortcomings of available measures of technological achievement 602.5 Deepwater petroleum technology in Brazil 622.6 The green revolution 682.7 Technology and growth in Latin America’s natural resource–based economies 712.8 Innovative use of communications technology is improving financial access for

the poor 752.9 The technological divide within India 913.1 Technology imports: Different paths for different countries 1133.2 European call centers in the Maghreb have inspired local entrepreneurs and prompted

a specialization in high-value-added services 1183.3 South African investment in Zambia’s retail sector has improved the quality of local

produce and farmers’ earnings 1183.4 Wal-Mart’s entry in Mexico boosted the Mexican soaps, detergents, and surfactants

industry 1193.5 Technological transfers through the diaspora and return migrants:

Some examples 1253.6 Principal market failures impeding technological progress in developing

countries 1433.7 Government sponsored innovation: Brazilian biofuels 1443.8 A successful government program of technological development and innovation

financing in the Republic of Korea 1453.9 Technology in 2020 152

C O N T E N T S

x

xi

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

low- and middle-income countries. Past edi-tions have examined the economic implica-tions of international and regional trade liber-alization, and migration and remittances. Lastyear’s report looked at the recent accelerationin growth among developing countries and itssustainability over the longer term.

This year we take a closer look at technol-ogy, a critical determinant of sustainablegrowth and poverty reduction. We do so bydirectly measuring the extent to which coun-tries use technological inputs (including scien-tific technologies embodied in goods andservices and business processes) and producetechnological outputs. The report also exam-ines trends in the major channels throughwhich technology is transmitted internation-ally, and in the country-specific factors that de-termine how well it is absorbed domestically.

Encouragingly, this Global EconomicProspects finds that, since the early 1990s,technological progress in both low- andmiddle-income countries has increased morerapidly than in high-income countries. As a re-sult, the level of technology used in developingcountries is catching up with high-incomecountries. However, the technology gap be-tween them remains wide. Globalization hasunderpinned much of the recent progress byexposing developing countries to foreign tech-nology through imports of high-tech consump-tion, intermediate and capital goods. Countries

have also benefited from rising levels of foreigndirect investment that often brings with itknowledge of important process technologiesand foreign markets. Finally, highly skilled in-ternational diasporas are exposing developingcountries to technology, both through thetrade and marketing contacts that they provideto their countrymen and through the return offormer émigrés.

Unfortunately, progress in improving thecapacity of developing countries to absorband make use of those technologies through-out their economies has been much weaker.Whether technological progress in developingcountries will continue to outpace high-income countries will depend on the improve-ments in this regard. The main impediments tofurther progress is not access to technologies,but the weakness of domestic skills and com-petencies, which prevents many developingcountries from exploiting these technologies,and rigidities in the regulatory environmentthat prevent innovative firms from being cre-ated and expanding. The diffusion of tech-nologies within countries is often slow, whichmeans that although some firms may havetechnologically sophisticated operations, mostdo not. Moreover, most of the population andmost firms operate in a low-tech environment.As a result, despite having technologicallysophisticated cities and world-class firms, theeconomy-wide level of technological achieve-ment in countries like China and India is notvery different from that in other countries atsimilar levels of development.

Foreword

F O R E W O R D

xii

This report suggests a number of policy di-rections to bolster technology diffusion andabsorption within developing countries. First,developing countries should safeguard theprinciple of openness and actively strengthenskills in the domestic population to ensurethat they are able to take advantage of futureopportunities. Second, to assist diffusionthroughout the economy, policy needs to rein-force technological absorptive capacity at thesubnational and regional levels and tostrengthen dissemination channels withincountries, including the outreach, testing,marketing, and dissemination activities of ap-plied R&D agencies. Third, authorities shouldensure that publicly supplied technologicalservices and technology-enabling infrastruc-ture are widely available, whether they are de-livered directly by the state or by private firms.Fourth, in low-income countries and in those

middle-income countries with uneven accessto quality secondary and tertiary schooling,efforts should concentrate on raising the qual-ity and quantity of schooling.

Finally, governments may need to intervenedirectly to encourage the rapid diffusion oftechnology and a domestic culture of “new-to-the-market” innovation. However, caution isrequired. Although direct interventions havesometimes been associated with some impor-tant technology successes, in many instancesthey have not. Policies that have succeededhave tended to make subsidies conditional onperformance and put in place high-quality andindependent-of-industry oversight systems.

Alan GelbActing Senior Vice President and

Chief EconomistThe World Bank

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. The principal authorsof chapter 1 were Hans Timmer and Elliot (Mick) Riordan. Chapter 2 was written by

Andrew Burns and William Shaw, with written contributions from Antonio David, YvanDecreux, and Annette De Kleine. Chapter 3 was written by Andrew Burns and William Shawwith written contributions from Dilek Aykut, Antonio David, Yvan Decreux, Annette De Kleine,Mariem Malouche, Sanket Mohapatra, and Olga Sulla. Both Chapters 2 and 3 benefitted fromthe expert research assistance of Taras Chernetsky, Shuo Tan, and Teng Jiang.

Several people contributed substantively to chapter 1. The Global Trends Team, under theleadership of Hans Timmer, was responsible for the projections, with written contributions fromJohn Baffes, Paul Brenton, Maurizio Bussolo, Betty Dow, Teng Jiang, Annette De Kleine, DonaldMitchell, Denis Medvedev, Gauresh Rajadhyaksha, Elliot (Mick) Riordan, Cristina Savescu,Shane Streifel, and Dominique van der Mensbrugghe. The poverty numbers originated withShaohua Chen from the Development Research Group.

The accompanying online publication, Prospects for the Global Economy (PGE), was pro-duced by a team led by Cristina Savescu and including Sarah Crow, Teng Jiang, Shunalini Sarkar,and Jennifer Vito, with technical support from Gauresh Rajadhyaksha.

Martha Grotton edited the report, Nigar Farhad Aliyeva and Michael Paul managed the pub-lication process, and Merrell Tuck managed the dissemination activities. Roula Yazigi providedinvaluable assistance with the design of some figures. Book production was coordinated by MaryFisk from the World Bank Office of the Publisher.

The report was produced under the guidance of Uri Dadush, François Bourguignon, and AlanGelb. Several reviewers offered extensive advice and comments throughout the conceptualizationand writing stages. These included Jean-François Arvis, Kevin Barnes, Vandana Chandra, Prof.Carl Dahlman, Mark Dutz, Alan Gelb, Mary Hallward-Dreimeier, Daniel Lederman, JeffreyLewis, William Maloney, Claudia Paz Sepulveda, and Alfred Watkins.

xiii

Acknowledgments

xv

Abbreviations

BACI Banque Analytique de Commerce International (International Trade Analytical Database)

CAGR compound annual growth rate

CAT scan computerized axial tomography scan

CEPII Centre d’Etudes Prospectives et d’Informations Internationales (Institutefor Research on the International Economy)

CIS Commonwealth of Independent States

DAX Deutsche Aktien Exchange

DJIA Dow Jones Industrial Average

DPT diphtheria, pertussis, and tetanus

DSL digital subscriber ink

EAF electric arc furnace

EAP East Asia and the Pacific

ECA Europe and Central Asia

EMBIG Emerging Market Bond Index-G

EPO European Patent Office

EU European Union

FDI foreign direct investment

GDP gross domestic product

GNI gross national income

HIV/AIDS human immunodeficiency virus/acquired immune deficiency syndrome

ICB International Crisis Behavior

IEA International Energy Agency

IMF International Monetary Fund

ISO International Organization for Standardization

LAC Latin America and the Caribbean

LME London Mercantile Exchange

MENA Middle East and North Africa

MSCI Morgan-Stanley Composite Index

NASDAQ National Association of Securities Dealers Automated Quotations

OECD Organisation for Economic Co-operation and Development

OHF open hearth furnace

OPEC Organization of the Petroleum Exporting Countries

PC personal computer

PPP purchasing power parity

R&D research and development

SAR South Asia region

SMEs small and medium enterprises

SSA Sub-Saharan Africa

TFP total factor productivity

TOPIX Tokyo Stock Price Index

UN Comtrade United Nations Comtrade database

UNCTAD United Nations Conference on Trade and Development

UNDP United Nations Development Programme

UNESCO United Nations Educational, Scientific, and Cultural Organization

UNIDO United Nations Industrial Development Organization

USPTO U.S. Patent and Trademark Office

WTO World Trade Organization

A B B R E V I A T I O N S

xvi

1

Overview

This edition of Global Economic Prospects isbeing released during a period of increased un-certainty following four years of recordgrowth in developing countries. In addition toexamining economic prospects over the nearand longer term, it takes an in-depth look atthe current level of and recent trends in tech-nological achievement and the main factorsthat determine the extent to which developingcountries succeed in implementing foreigntechnologies.

Notwithstanding the financial turmoilprovoked by a reassessment of risks in theU.S. mortgage market, and despite large lossesin some financial markets, exposure to asset-backed securities appears to be broadly based.Losses so far have been manageable, althoughcredit conditions have tightened. For develop-ing economies, sovereign risk premiums haveincreased but remain low by historical stan-dards. Equity values, exchange rates, andcommodity prices have become more volatile,and the vulnerability of countries with largecurrent account deficits or pegged exchangerates has become more visible.

Against this background, global growthslowed modestly in 2007, coming in at 3.6 per-cent after a strong 3.9 percent in 2006.Most of the slowdown was attributable toweaker growth in high-income countries.Growth in developing economies was a robust7.4 percent, broadly unchanged from 2006(figure 1). This strong performance in thedeveloping countries has offset somewhat theslowdown in U.S. domestic demand thatstarted with the unwinding of the housingbubble early in 2006. During 2007, developing

countries accounted for more than half thegrowth in world imports, contributing—alongwith the depreciation of the dollar—to strongnet exports for the United States and further-ing the reduction in global imbalances.

Global growth in 2008 should moderate to3.3 percent, as the robust expansion in devel-oping countries partly compensates forweaker results in high-income countries.World output should pick up in 2009, ex-panding by 3.6 percent, as the U.S. economyregains momentum.

Several serious downside risks cast ashadow over this soft landing for the globaleconomy. External demand for the products ofdeveloping countries could weaken muchmore sharply and commodity prices could de-cline if the faltering U.S. housing market orfurther financial turmoil were to push theUnited States into a recession. Alternatively,monetary authorities might overreact to thecurrent climate of uncertainty and overstimu-late the economy. This would be particularlydangerous for developing countries if the bulkof the resulting liquidity were to move intorapidly growing developing regions, provok-ing the same kind of overinvestment condi-tions that arose in the U.S. housing market.

Prospects for the U.S. dollar represent anadditional risk factor. A recession in the UnitedStates or an excessive easing of U.S. monetarypolicy could contribute to further sharp declinesin the dollar. A weaker dollar would benefitdeveloping countries with dollar debt, butimpose losses on those that hold dollar-denominated assets. It would hurt the competi-tiveness of firms exporting to the United States

(and those producing close substitutes for U.S.imports), while benefiting countrieswithcurren-cies pegged to the dollar—at least temporarily.However, the main impact of a precipitous de-cline of the dollar would likely derive from theincreased uncertainty and financial-marketvolatility it would provoke, which would in-crease trading costs, and spreads on developing-country debt—resulting in weaker export andinvestment growth throughout the globaleconomy.

Even should such risks not materialize,several developing countries may be quitevulnerable to sudden adjustments in financialmarkets. Most exposed are those countriesthat combine large current account deficitswith pegged exchange rates and with increas-ing domestic inflation. Also at risk are coun-tries whose domestic banking sectors havebalance sheets characterized by large currencymismatches.

Technological achievement anddiffusion in developingcountries

The special topic of this edition of GlobalEconomic Prospects is technology and its

diffusion within the developing world. Much

of the economic and social progress of the pastfew centuries has been due to technology.Technology has been central to both economicgrowth and many elements of social welfarethat are only partly captured by standard mea-sures of gross domestic product (GDP), includ-ing health, education, and gender equality. Asmeasured by total factor productivity, it ex-plains much of the differences in both the leveland rate of growth of incomes across countries(Easterly and Levine 2001; Hall and Jones1999; King and Levine 1994). And, lookingforward, it is expected to play a central role inmeeting the environmental and climate-changechallenges of the remainder of this century.

The private sector and the efficient func-tioning of markets are key to technologicalprogress. At the same time, the efficient deliv-ery of socially relevant technological goods andservices depends on the direct contribution ofnonmarket actors, including governments,nongovernmental organizations, and interna-tional organizations. Of course, policy alsosupports technological progress by facilitatingthe smooth operation of markets, by ensuringthe acquisition of technological competenciesby the general population, and by providing thephysical infrastructure that is often a necessarycomplement to technologically sophisticatedactivities. Active measures to promote technol-ogy diffusion and strengthen the linkages be-tween firms and research and development(R&D) agencies are also vital.

In exploring technological achievement anddiffusion, this report adopts a broad definitionof technology and technological progress, onethat encompasses the techniques (includingthe way the production process is organized)by which goods and services are produced,marketed, and made available to the public.Understood in this way, technologicalprogress at the national level can occurthrough scientific innovation and invention;through the adoption and adaptation of pre-existing, but new-to-the-market, technologies;and through the spread of technologies acrossfirms, individuals, and the public sector withinthe country.

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 8

2

Figure 1 Robust growth amongdeveloping countries should cushion thedeveloped country slowdown

Source: World Bank.

2007

2008

2009

2004

2005

2006

2003

2002

2001

2000

8

0

2

4

6Developingcountries

ForecastReal GDP, annual percent change

High-incomecountries

The following discussion traces the struc-ture of the overall report, which in chapter 2explores the level of—and recent trends in—technological achievement, as well as theprocess by which technology diffuses betweenand within countries. Chapter 3 concentrateson the process by which countries absorb for-eign technology, both the mechanisms throughwhich they are exposed to foreign technolo-gies and the domestic factors that dictate howsuccessfully they absorb those technologies.Although the chapter identifies a number ofimportant, policy-relevant trends, and it ex-plores their policy implications, it leaves tofuture work a more normative analysis of thepolicies that developing countries shouldfollow to maximize the development benefitsof technological progress.

Policy needs to actively promotetechnological adoption and adaptation aswell as nurturing domestic innovativecapacityA central finding of the report is that most de-veloping countries lack the ability to generateinnovations at the technological frontier.Although the number of patents and scientificjournal articles is strongly correlated withGDP per capita for high-income countries,

almost none of this activity is being performedin developing countries (figure 2). The lack ofadvanced technological competencies in thesecountries means that technological progress indeveloping countries occurs through theadoption and adaptation of pre-existing butnew-to-the-market or new-to-the-firm tech-nologies. Moreover, relatively thin domestictechnology sectors and much better economicand scientific opportunities abroad mean thatmany nationals of developing countries per-form cutting-edge research in high-incomecountries. For example, 2.5 million of the21.6 million scientists and engineers workingin the United States were born in developingcountries (Kannankutty and Burelli 2007).

The level of technological achievement indeveloping countries has converged withthat of high-income countries over thepast 15 yearsA sustained policy of increased openness to for-eign trade and foreign direct investment (FDI),plus increased investments in human capital,have contributed to substantial improvementsin technological achievement in developingcountries over the past 15 years. And despiterapid progress at the technological frontier,technological achievement in both low- and

O V E R V I E W

3

Figure 2 Scientific innovation and invention is almost exclusively a high-income activity

Index

Source: World Bank.

0High-income

countriesUpper-middle-income

countriesLower-middle-income

countriesLow-income

countries

0.05

0.20

0.25

0.15

0.10

middle-income countries has increased muchmore rapidly than in high-income countries. Asa result, developing countries have closedthe relative gap with high-income countries.However, the gap remains large (figure 3).Moreover, the strong aggregate performance oflow-income countries reflects large improve-ments in technological achievement by some,but much more modest advances by themajority. As a consequence, many are onlymaintaining pace with, or even losing groundto, high-income countries.

In general, the level of technologicalachievement observed in a country is posi-tively correlated with income levels. However,considerable variation is apparent within in-come groups. Among other things, this varia-tion reflects the nature of the technology beingobserved, the impact of the overall policyframework on the ability of technologicallysophisticated firms to grow, and the extent towhich governments have given priority to andhad success in delivering services with a strongtechnology component.

The penetration of older technologies, suchas fixed-line telephones, electrical power, trans-portation, and health care services—many ofwhich were originally provided by govern-ments—is only weakly correlated with income.The low-income countries with the highestutilization rates of these older technologiestend to have rates as high as those of the aver-age lower-middle-income country (figure 4).Similarly, the lower-middle-income and upper-middle-income countries with the highest uti-lization rates tend to have rates that match theaverage rate of the next highest income group.

In part, this reflects the nature of thetechnologies in question, such as electricalnetworks, road infrastructure, fixed-linetelephony, and sanitation networks. Manyof these technologies require an infrastructurethat is relatively expensive to create andmaintain, and which relies on large numbersof individuals with scarce technical skills. Inaddition, the observed diffusion of older tech-nologies today depends on the intensity and ef-ficiency with which government services have


4

Figure 3 Technological achievement:Converging, but the gap remains large

Percent change in technological achievement, 2000s versus1990s

Rapid progress in developing countries…

…fueled relative convergence…

…but the gap remains large

Source: World Bank.

0High-income

countries

30

60

150

180

120

90

Upper-middle-income

countries

Lower-middle-income

countries

Low-incomecountries

Technological achievement index

0High-income

countries

0.02

0.10

0.08

0.06

0.04

0.16

0.18

0.14

0.12

Upper-middle-income

countries

Lower-middle-income

countries

Low-incomecountries

Index, high-income countries � 100

0High-income

countries

20

40

100

120

80

60

Upper-middle-income

countries

Lower-middle-income

countries

Low-incomecountries

been delivered in the past. Part of the strongtechnological showing of the countries in theformer Soviet bloc is explained by the heavyemphasis that past governments placed on pro-viding basic infrastructure and education to awide range of the population. Similarly, pastgovernance problems and civil strife help ex-plain the relatively weak penetration of thesetechnologies in many Sub-Saharan Africancountries, whereas macroeconomic turmoiland a relatively unequal distribution of in-comes and skills in Latin America may havecontributed to weak outcomes in that region.

The penetration rates of newer technolo-gies have risen relatively rapidly and are moredirectly correlated with income than is the casefor older technologies. The infrastructure fornewer technologies such as mobile phones,computers, and the Internet is generally lessexpensive to create and requires fewer (thoughmore skilled) workers to maintain. Moreover,in many countries, regulatory reform hasmeant that the private sector now offers theseservices in a competitive environment as com-pared with the state-owned, monopolistic en-vironments of the past. As a result, supply of

O V E R V I E W

5

Figure 4 The penetration of older and more recent technologies depends on morethan income

Index

Source: World Bank.

0High-income countries Upper-middle-income


countriesLow-income countries

0.10

0.35

0.05

0.30

0.25

0.20

0.15

Index

0High-income countries Upper-middle-income


countriesLow-income countries

0.1

0.7

0.6

0.5

0.4

0.3

0.2

0.8

Penetration of older innovations (2000–03)

Penetration of newer innovations (2000–03)

these new technologies has been more respon-sive to market demand and less restrainedby the budget constraints of governments orstate-owned-enterprises. Furthermore, de-mand for these products has been boosted bylow end-user costs as a result of competitivepricing strategies and because some of thesenewer technologies lend themselves more eas-ily to sharing than do some older technologies.

Overall, although technological achieve-ment tends to rise with incomes, this relation-ship is nonlinear and shows a tendency to leveloff. Moreover, it is not uniform across regions.Thus countries in Europe and Central Asiatend to have somewhat higher levels ofachievement than would be expected on thebasis of income alone, but the overall relation-ship between technological achievement andincome in the region tracks relatively well thatof all countries (figure 5). In contrast, techno-logical achievement in Latin America tends tobe lower than what would be expected givenincomes, and the overall relationship suggests

that other factors appear to be restrainingachievement even as incomes progress. Theseresults are consistent with the view that policychoices over the long term (such as thosethat generated the uneven distribution of in-come and educational opportunities in LatinAmerica and the region’s history of weak linksbetween R&D communities and the businessworld) are important determinants of absorp-tive capacity and technological progress.

The level of technology in developingcountries reflects the pace at whichtechnology diffuses within countriesAlthough it can take time for a technology togain a foothold in developing countries, themore serious impediment to technologicalachievement is the speed with which tech-nologies spread within these countries. Onaverage, the time it takes before official statis-tics in a developing country record significantexploitation of a new technology has declinedfrom almost 100 years for innovations


6

Figure 5 Technological achievement tends to level off at different income levels indifferent regions

�0.10

�0.05

0.25

0.30

0.15

0.05

0.10

0

0.20

Source: World Bank.

0 5,000 10,000 15,000

Income per capita (PPP)

20,000 25,000 30,000 35,000 40,000

Europe and Central Asia

Latin America

All countries

East Asia and the Pacific Europe and Central Asia

Latin America and the Caribbean Middle East and North Africa

Middle East and North Africa South Asia Sub-Saharan Africa

High-income OECD countries

High-income other countries

Technological achievement versus per capita income by region

Index

discovered in the 1800s to about 20 yearstoday.1 However, technological progress alsodepends on how rapidly the technologyspreads within the country. Here the story isless encouraging. For technologies discoveredduring 1950–75, only a quarter of the devel-oping countries that have achieved at least a5 percent penetration level have gone on toreach the 25 percent threshold, and all of theseare upper-middle-income countries (figure 6).

The story is somewhat better for newertechnologies. Not only have these technologiesspread more quickly between countries, butalso the share of countries that have achievedthe 25 percent threshold is higher, at 33 per-cent. Indeed, developing countries have nowreached the same average level of penetrationof mobile phones as was observed in high-income countries in 1995.

The unevenness of technological diffusionacross countries is often mirrored withincountries, especially large countries. Althoughtechnology spreads relatively rapidly amongelites living in major cities, it takes muchlonger for it to find its way to the rest of thepopulation or from top-performing companiesto the average firm. Specific sectors in ad-vanced urban centers in China and India, for

example, use world-class levels of technology,but the incidence of these technologies else-where in the country, and in rural areas in par-ticular, remains low (figure 7). Even withinsectors, technology may diffuse only slowly. InBrazil and India, for example, the most so-phisticated firms use technologies and achievelevels of productivity that rival world leaders,but the vast majority of firms operate at levelsof productivity that are less than one-fifththose of the top performers.

A framework for understanding thediffusion of technology within developingcountries The bulk of technological progress in develop-ing countries has been achieved through theabsorption and adaptation of preexisting andnew-to-the-market or new-to-the-firm tech-nologies, rather than the invention of entirelynew technologies. Given the still wide technol-ogy gap, this is likely to remain the case forthe vast majority of developing countries.

A developing country’s ability to absorband adopt foreign technologies depends ontwo main factors: the extent to which it isexposed to foreign technologies (the pace at

O V E R V I E W

7

Figure 6 Most technologies fail topenetrate deeply into developingeconomies

Source: World Bank calculations using data from Comin andHobijn 2004.

1800s 1900–50 1950–75 1975–2000

120

0

20

40

80

60

100

5%

25%

50%

Percent

Figure 7 The urban–rural gap in telephoneaccess in India is huge

Percent of subscribers with phone service

0

1998

1999

2000

2001

2002

2003

2004

2005

2006

2007

a

10

20

30

60

50

40

Source: Telecommunications Regulatory Authority of India.

a � estimated.

Cities

Ruralareas

depends on the technological absorptivecapacity of the economy (represented by themultiple-ringed drum). Absorptive capacitydepends on the overall macroeconomic andgovernance environment, which influencesthe willingness of entrepreneurs to take riskson new and new-to-the-market technologies;and the level of basic technological literacyand advanced skills in the population, whichdetermines a country’s capacity to undertakethe research necessary to understand, imple-ment, and adapt them. In addition, becausefirms are the basic mechanism by which tech-nology spreads within an economy’s privatesector, the extent to which financing for inno-vative firms is available—through the bank-ing system, remittances, or governmentsupport schemes—also influence the extentto and speed with which technologies areabsorbed.


8

Pro-active policies

Finance of innovative firms

Basic technological literacy

Governance and the business climate

Source: World Bank.

Figure 8 Domestic absorptive capacity both conditions and attracts external flows

Technological frontier

Tech

nolo

gica

lab

sorp

tion

Diasporaand othernetworks

FDITrade

Tran

smis

sion

chan

nels

Tech

nolo

gica

l abs

orpt

ive

capa

city

Policies to - create competencies - build infrastructure - foster an innovation- friendly business climate

Returns toscale

Spillovereffects

Domestic technological achievement

Dynamiceffects

magnifytechnology

transfer

which technologies diffuse across countries)and its ability to absorb and adapt those tech-nologies to which it is exposed (the pace atwhich technology diffuses within the coun-try). Figure 8 presents a stylized description ofhow a developing country absorbs technology.As a first step, an economy is exposed tohigher-tech business processes, products, andservices through foreign trade; foreign directinvestment; and contacts with its diaspora andother communications channels, includingacademia and international organizations (thelarge arrows at the top of the figure). Thelarger these flows, the greater the exposure ofthe economy to the global technologicalfrontier.

However, exposure to new ideas and tech-niques is not sufficient to ensure that thetechnology diffuses throughout the economy.Successful absorption of foreign technology

Government policy also has a crucial roleto play. Governments are often the primarychannel through which certain technologies,such as electricity, fixed-line telephones,transportation infrastructure, and medical andeducational services, are delivered. Moreover,government policy is largely responsible forcreating a business environment that facili-tates easy firm entry and exit and that is nothostile to the profits to be made from exploit-ing new technologies. Too often, rules and/orspecific features of the domestic market pre-vent firms from making money by exploiting anew technology, and, as a result, the technol-ogy does not spread within the country. Policyshould also ensure that R&D and dissemina-tion efforts give priority to creating and intro-ducing products for which a market (domesticor foreign) exists and to helping firms exploitthose opportunities.

The overall process is, of course, muchmore complicated and much less mechanisticthan is depicted in Figure 8. Technologicalflows and technological absorptive capacityinfluence each other. How well technology dif-fuses depends on various market imperfec-tions, including increasing returns to scale andtechnological spillovers (the smaller light bluerings toward the bottom of the figure). Herethe existence of a financial sector that inter-mediates between savers and innovators maybe necessary to overcome the initial cost ofsome new technologies. In particular, accessto finance may be essential if innovative firmsare to achieve the necessary scale to unleash apotential virtuous circle, so that the additionalincome garnered by the successful exploita-tion of one new technology permits the acqui-sition of another, thus resulting in furthergains.

Increased openness to trade, FDI, anddiaspora contacts have boostedtechnological diffusionThe dismantling of trade barriers in many de-veloping countries over the past two decadeshas dramatically increased developing coun-tries’ exposure to foreign technologies. Their

imports of capital and intermediate goods(which permit the production of technologi-cally sophisticated goods and services) nowrepresent between 6 and 14 percent of theirGDP, an increase of more than 80 percentsince 1994. The ratio of high-tech importsto GDP more than doubled during the sameperiod. Partly as a result, developing-countryexports of high-tech goods have also in-creased, rising from 11 percent of total ex-ports in the mid-1990s to 19 percent in2002–04 (figure 9). In the case of lower-middle-income countries, high-tech goods rep-resent broadly the same 23 percent share intotal exports as in high-income countries (15percent if China is excluded).

The easing of restrictions on FDI also hascontributed to technology diffusion within de-veloping countries. FDI is a major source ofprocess technology and learning by doing op-portunities for individuals in developing coun-tries. Over the past 15 years, FDI inflows todeveloping countries have almost doubled as apercentage of GDP. In addition, foreign firmsare making important contributions to thetechnological capacity of host countries, per-forming more than 40 percent of the totalR&D in some countries. At the same time, thecompetition, standards and knowledge offoreign markets that foreign firms bring tothe domestic market can have importantspillover effects. Finally, many firms in devel-oping countries have increased their access tocutting-edge technology by purchasing techno-logically sophisticated firms domiciled in high-income countries.

In addition to dismantling barriers to for-eign investment, some middle-income coun-tries have encouraged greater FDI flows byimplementing stronger regimes governing in-tellectual property rights (evidence suggeststhat stronger intellectual property rights areassociated with a rise in knowledge flows toaffiliates and in inward FDI flows towardmiddle-income and large developing coun-tries, but not in poor countries). A few coun-tries have encouraged joint ventures ratherthan FDI to maximize technology transfers to

O V E R V I E W

9

local firms. However, this strategy seems towork only for countries with substantial mar-ket power. In particular, fear of losing controlover cutting-edge technologies sometimescauses multinational firms forced into jointventures to reserve their best technologies forthe domestic market and transfer only olderless efficient ones.

Substantial technology transfers are also as-sociated with international migration and thediasporas of developing countries. Not all ofthese are positive. Even though 93 percent ofuniversity-educated individuals from develop-ing countries return to or remain in their coun-try of origin (Docquier and Marfouk 2004), thebrain drain is a serious problem for a number ofmostly small countries. However, the existenceof a well-educated diaspora (more highly-skilled individuals migrate than lower-skilledindividuals) constitutes an important techno-logical resource for the home country—a brainbank, as it were. This is especially the case whenweak employment prospects in the home coun-try reduce the economic benefits initiallyforgone by the individual’s departure.

For most countries, high-skilled out-migration remains at managable levels and thesetechnologically savvy diasporas contribute totechnological transfers by strengthening trade

and investment linkages with more advancedeconomies through networks that provideaccess to technology and capital and through re-mittances. Remittances not only contribute todomestic entrepreneurship and investment, butalso, along with the introduction of mobilephone services, have greatly expanded the pro-vision of banking and other arm’s-length fi-nancial services within developing countries—themselves a critical enabling processtechnology. Finally, returning migrants canprovide important resources, such as entrepre-neurship, technology, marketing knowledge,and investment capital. The effect of a single re-turning émigré armed with skills acquired in adeveloped economy can have (and has had)large economic and technological effects on thecountry of origin.

Better macroeconomic and educationalpolicies have improved absorptivecapacity in developing countries...Although increases in the flows of the princi-pal international transmitters of technologyhave been marked, improvement in the factorsthat determine the capacity of developingcountries to absorb and effectively use thattechnology has been much more gradual. Onthe positive side, most developing countries


10

Figure 9 Developing countries’ trade in technology goods has risen

Percent of GDP Percent of merchandise exports

Source: World Bank.

0

2

8

4

6


Upper-middle-income

countries

Lower-middle-income

countries

Low-incomecountries

0

5

15

10

20

25


Upper-middle-income

countries

Lower-middle-income

countries

Low-incomecountries

Imports of high-tech goods Exports of high-tech goods

2002–04

1994–962002–04

1994–96

has reached 70 years and continues to rise. Inlow-income countries outside of Sub-SaharanAfrica, life expectancy is up from 59 yearsin 1990 to 64 years in 2005 (in Sub-SaharanAfrica, extremely low incomes and theHIV/AIDS epidemic have led to a drop inlife expectancy since 1990). The labor forcein most developing countries has also becomebetter educated. Adult literacy rates have in-creased in every developing region over thepast 15 years (figure 11). The share of chil-dren graduating from primary school has alsoincreased in all regions except East Asia andthe Pacific (where it stood at 98 percent in2005). Meanwhile, secondary school and col-lege enrollment rates are up across the board.Increased school enrollment has raised youthliteracy rates to close to 100 percent in allthe predominantly middle-income regions. Ac-cording to official statistics, almost 75 percentof 15- to 24-year-olds in Sub-Saharan Africacan read and write. That rate compares favor-ably with an adult literacy rate of 60 percentand suggests that over time, the technologicalliteracy of the population will rise. Althoughpolicies to promote literacy and extend school

O V E R V I E W

11

Figure 10 Macroeconomic stability has improved since the early 1990s

Percent

Source: World Bank.

a � average absolute value of the monthly change in the real-effective exchange rate.

0High-income

countries

5

20

25

15

10

Upper-middle-income

countries

Lower-middle-income

countries

Low-incomecountries

Percent

0High-income

countries

6

8

4

2

Upper-middle-income

countries

Lower-middle-income

countries

Low-incomecountries

2002–06

1990–94

2002–06

1990–94

Median inflation rate Real-effective exchange rate volatilitya

have improved their investment climates.Their macroeconomic and political environ-ments have become more stable over the past15 years. The number of international con-flicts has fallen by more than 50 percent sincethe 1990s, median inflation has dropped fromabout 20 percent in the early 1990s to lessthan 5 percent, and exchange rate volatilityhas fallen by more than 50 percent in every de-veloping region (figure 10). All these changesreduce risk and increase the likelihood that en-trepreneurs will take a chance and introduce anew technology within a country. These samechanges have contributed to improved percapita GDP and a significant decline in thenumber of people living in absolute poverty,which has eased the constraints on the abilityof poor countries to generate resources forinvestment, and has increased the willingnessof firms and individuals to take risks.

Improvements in the quality of humancapital in most developing countries have in-creased the countries’ capacity to adopt andadapt technologies. Poor health is receding asa factor that impedes technological progress.Life expectancy in middle-income countries

attendance are critical, in too many cases, thequality of the education delivered in many de-veloping countries remains low. Large propor-tions of students officially classified as literatefail to pass international standardized tests ofliteracy and numeracy.

Technological progress requires additionalimprovements in the quality of the labor forcebeyond strengthening educational systems.Training can make an important contributionto both the productivity of private firms and theefficiency of public services. For example, thedissemination of the simple skills required tobuild rainwater collection systems can improveaccess to clean drinking water and to reduce theincidence of disease. And investing in the do-mestic skills required to support high-skill andhigh-value-added industries can help maximizethe technology spillovers from FDI.

...but improvements in the businessclimate and governance lagsIn contrast to improvements in the quality ofhuman capital, business climate and gov-ernance indicators have shown little improve-ment, on average, over the past decade.Governance in several countries hasimproved, notably in Central Europe and theBaltic countries, proving that motivated

political leadership can make a difference.But, in many other countries, the quality ofgovernance has declined or remained stable.

Progress in the dismantling of regulatorybarriers that impede technology diffusion hasalso been slow. Restrictions on labor mobilitythat constrain firms’ ability to reallocateworkers within the firm can be important bar-riers to the adoption of new technologies, andrestrictions on firm entry and exit tend to propup inefficient firms and limit the expansionand creation of innovative ones. Overall, thetime and cost involved in starting a business,the efficiency of contract enforcement, thetime required to resolve insolvencies, the aver-age amount recovered, and the degree of cor-ruption in developing countries generates anoverall investment climate that is much lessconducive to innovation than that observed inthe industrial countries (figure 12).

Along with eliminating unnecessaryrequirements, technological progress oftenrequires the strengthening of regulatory initia-tives. For example, improvements in theeffectiveness of public-sector institutions havecontributed to more efficient logistics services,a key determinant of trade competitiveness.Strengthening contract enforcement, the effi-ciency of court operations, the security ofproperty rights (including the reliability andtimely update of property registries), and theeffective regulation of financial markets canbe critical to ensuring an adequate return toinvestments in technology. Governments alsocan play a key role in boosting technologicalprogress through the definition and promo-tion of product standards, and in helpingfirms comply with them.

Despite the limited amount of at-the-frontier scientific innovation performed in de-veloping countries, technological progressdepends on R&D and especially technologydissemination activities. In most developingcountries and sectors, R&D should focus onthe adoption and adaptation of preexistingtechnologies, not on efforts to expand theglobal technological frontier. For low-incomecountries, policy should focus on strengthening


12

Figure 11 Literacy rates have increasedin all regions

0

20

40

60

80

100

East A

sia

and

the

Pacific

Europ

e an

d

Centra

l Asia

Latin

Am

erica

and

the

Caribb

ean

Midd

le Eas

t and

North

Afri

ca

South

Asia

Sub-S

ahar

an

Africa

Percent of adult population

Source: World Bank.

2005

1990

Figure 12 Developing regions have much poorer governance than do OECD countries

Regional averages of six governance indicators: percent share of average OECD percentile rank

0 20 30 40 50 60 70 80 9010 100

Source: Kaufman, Kraay, and Mastruzzi 2007.

Note: OECD � Organisation for Economic Co-operation and Development.

OECD

Caribbean

Europe and the Baltics

East Asia

Developing countries average

Latin America

Middle East and North Africa

South Asia

Sub-Saharan Africa

Former Soviet Union

the infrastructure necessary for the successfuldiffusion and implementation of technologies,on facilitating the diffusion of already existingtechnologies, and on developing domesticcompetencies. More technologically advancedmiddle-income countries should emphasizethe same points but should strengthen theirR&D and technical competencies in order toincreasingly compete at the global technologi-cal frontier. In both low- and middle-incomecountries, policy should place special empha-sis on incentives and on maintaining strongties to private-sector firms.

Some policy directions

This review of the level of and trends intechnological achievement in developing

countries, of the major transmitters of techno-logical knowledge, and of the determinantsof countries’ ability to absorb them suggests anumber of empirical conclusions (box 1). Thisreport does not offer a comprehensive expla-nation of why technological progress occurs,nor does it include an in-depth analysis ofthe policies that governments can adopt toincrease the rate of technological progress.Nevertheless, the preceding analysis makesclear that some combination of opennessto foreign technology, strong domestic

technological competencies, a motivated publicsector, and a well-financed private sector arekey ingredients for success. In addition, severalgeneral policy directions suggest themselves.

First, much of the technological progress indeveloping countries over the past 15 yearshas been associated with the increase in open-ness that occurred during the same period.This openness has increased developing coun-tries’ exposure to foreign technologies, buttheir capacity to absorb them has improvedmuch less. To the extent that technologicalabsorptive capacity limits the level of techno-logical achievement that an economy canreach (as suggested by the tendency for tech-nological achievement in Latin America tolevel off), the relatively weak improvement inabsorptive capacity may result in a futureslowing of the rate of technological progress insome countries unless they take significantsteps to raise the quality of domestic humancapital, improve the regulatory environment,and increase the efficiency with which theydeliver government services. This risk may bemost marked for those countries such as In-donesia and Mexico that have taken advan-tage of globalization in a relatively passivemanner, exploiting their low-wage compara-tive advantage without taking strong steps toimprove domestic competencies.

O V E R V I E W

13

Second, because of the complementarity oftechnologies and infrastructure, countrieswhere older technologies have yet to penetrateparticularly deeply may also face limits to theextent to which other technologies are able todiffuse. Therefore, the authorities should focuson ensuring that publicly supplied technologi-cal services are available as widely, reliably, andeconomically as possible, whether they are de-livered directly by the state or by private firms.

Third, a main remaining challenge is toensure that technologies diffuse throughoutthe country, not just to major centers or top-performing firms. This does not mean tryingto create research centers everywhere, but itdoes require reinforcing absorptive capacity atthe subnational level. Moreover, it means pay-ing attention to dissemination channels withincountries, including domestic transportationinfrastructure, and the essential role to beplayed by the outreach, testing, marketing,and dissemination activities of applied R&Dagencies.

Fourth, notwithstanding the relativelystrong improvement in technological achieve-ment by some low-income countries, manyothers have improved only marginally or notat all. In particular, improvements in techno-logical absorptive capacity have been limited.

Efforts to concentrate on increasing thequality of human capital must continue, notonly by ensuring that more students stay inschool longer, but also by raising standards,which in too many cases are too low.

Fifth, given the importance of market fail-ures (for example, increasing returns to scale,the potential for coordination failures, the dif-ficulties in appropriating the full returns to in-novation owing to imitators, and capital-mar-ket imperfections), governments may need tointervene directly to encourage the rapiddiffusion of technology and the growth of a vi-brant domestic culture of technology adapta-tion and new-to-the-market innovation. Poli-cies that have been tried include, among others,support for industry-specific research, subsidiesfor specific products, barriers to trade thatfavor technology-intensive activities, and di-rected credit programs. Such policies have beenassociated with economic miracles, particularlyin several East Asian countries. However, theyhave also been associated with significant fail-ures, notably in some Latin American and Sub-Saharan African countries. In those caseswhere direct interventions have been success-ful, they have tended to make support condi-tional on performance and have maintainedhigh-quality government monitoring programs


14

First, on most fronts, developing countries haveprogressed markedly over the past 15 years. As a

result, technological achievement in all incomegroups and in every region has advanced morequickly than in high-income countries.

Second, the technological frontier has advanced ashigh-income countries (and some developing coun-tries) continue to innovate at a rapid rate. Thus thetechnology gap between developed and developingcountries remains large, particularly for low-incomecountries.

Third, to a large extent the convergence in tech-nological achievement reflects a substantial increasein the openness of developing countries to foreign

Box 1 Summary of empirical resultstrade, foreign direct investment, and internationalmigration, which has dramatically increased both theexposure of developing countries to new technologiesand the opportunities to use foreign markets toexploit increasing returns to scale.

Fourth, progress has also been made in increas-ing countries’ absorptive capacity through im-proved literacy, enhanced educational attainment,and better macroeconomic stability. However,progress in improving the business climate andgovernance indicators has been much more mixed.As a result, technological absorptive capacity hasadvanced much less quickly than technologicalachievement.

that have avoided being “captured” by indus-trial interests.

Note1. Significant is defined here to be a penetration

rate that is at least 5 percent of the average level incountries with the highest rate of exploitation.

ReferencesComin, Diego, and Bart Hobijn. 2004. “Cross-Coun-

try Technology Adoption: Making the TheoriesFace the Facts.” Journal of Monetary Economics51(1): 39–83.

Docquier, Frederic, and Abdeslam Marfouk. 2004.“Measuring the International Mobility of SkilledWorkers (1990–2000).” Policy Research WorkingPaper Series 3381. World Bank, Washington, DC.

Easterly, William, and Ross Levine. 2001. “It’s NotFactor Accumulation: Stylized Facts and Growth

Models.” The World Bank Economic Review15(2):177–219.

Hall, Robert E., and Charles I. Jones. 1999. “Why DoSome Countries Produce So Much More Outputper Worker Than Others?” The Quarterly Jour-nal of Economics 114(1):83–116.

Kannankutty, Nirmala, and Joan Burrelli. 2007. “WhyDid They Come to the United States? A Profile ofImmigrant Scientists and Engineers.” Info Brief.National Science Foundation: Directorate forSocial Behavioural and Economic Sciences.Arlington, VA. June.

Kaufmann, Daniel, Aart Kray, and Massimo Mas-truzzi. 2007. “Governance Matters: GovernanceIndicators for 1996–2006.” World Bank PolicyResearch Working Paper No. 4280, World Bank,Washington, DC.

King, Robert G., and Ross Levine. 1994. “CapitalFundamentalism, Economic Development, andEconomic Growth.” Carnegie-Rochester Confer-ence Series on Public Policy 40:259–92.

O V E R V I E W

15

17

Prospects for Developing Countries1

Following the sudden and sharp drop inmarket valuations of U.S. mortgage-backedsecurities in mid-2007, global markets haveentered a phase of heightened uncertainty.This has been reflected in increased volatilityin equity markets, commodity prices, andexchange rates.

Notwithstanding the increased volatility,the impact on developing countries has beenrelatively minor to date. Risk premiums haveescalated, but remain relatively low in a his-toric context, and capital inflows remainplentiful, although bank lending has droppedoff. Aggregate growth in developing coun-tries continues to be strong, reflectingimproved fundamentals in many countries,sizable revenues from commodity exports,and continued access to international financeat moderately higher cost. Their strong grossdomestic product (GDP) growth is partiallyoffsetting weaker U.S. domestic demand,which is now expected to remain subduedwell into 2008.

Despite the resilience demonstrated by theglobal economy, risks exist and increasedvolatility has made several developing coun-tries more vulnerable to financial disturbance,especially those with large current accountdeficits, pegged exchange rates, or domesticbanking sectors that have borrowed heavilyin international markets.

Growth outlook

On average, developing countries havebeen affected only modestly by the slow-

down in the United States during 2007, whichis now anticipated to continue into 2008

before picking up in 2009. GDP growthamong low- and middle-income economieseased just 0.1 percentage point in 2007 fromthe strong 7.5 percent recorded in 2006.Despite weaker U.S. import growth, continuedrobust spending by oil-exporting countries andvibrant expansions in China and India are pro-jected to keep developing-country growthstrong at 7 percent or more in 2008 and 2009.

Over the longer term, the resilience of de-veloping countries’ improved fundamentalswill be tested. More prudent macroeconomicmanagement and technological progress (seechapters 2 and 3) have contributed to an in-crease in total factor productivity (TFP) andreal income growth over the past 15 years.Over the next 10 years, these same factors areexpected to enable developing countries toachieve annual per capita income gains of3.9 percent, and perhaps as much as 3.4 per-cent in the following decade. These projec-tions imply per capita income growth that ismore than twice as fast as that in high-incomecountries. Growth of such magnitude wouldreduce the number of people living on lessthan a dollar a day from 1.2 billion in 1990and 970 million in 2004 to 624 million by2015. Such aggregate outcomes are not guar-anteed, however, and performance across in-dividual countries is likely to be diverse.

Inflation has remained remarkably mutedworldwide despite four years of stronggrowth. Many developing countries have con-tained domestic inflation following a tighten-ing of monetary and fiscal policies. The sharpincreases in commodity prices mainly hadone-time direct impacts on inflation, with only

limited second-round effects. Moreover, the in-creasing integration of developing countriesinto global markets and their rising shares inworld trade have helped dampen inflationglobally through heightened internationalcompetition. In some countries, however, infla-tion may become an increasing challenge. Inseveral oil-exporting countries, spending ofvast export revenues is heating up domesticmarkets. In China, efforts to slow growth maynot succeed in quickly reversing a recent accel-eration of inflation, and demand pressuresremain pronounced in several countries inEurope and Central Asia and Latin Americaand the Caribbean. In Sub-Saharan Africa, thecombination of strong domestic demand andrising international grain prices could pushalready mounting inflation still higher, particu-larly in import-dependent coastal states.

Continued high and increasing oil priceshave stimulated the use of food crops for bio-fuels and raised fertilizer costs. Prices of maizeand vegetable oils increased by 33 and 50 per-cent, respectively, during 2007. Wheat produc-tion fell short of consumption partly becauseit has been displaced by maize and partlybecause of adverse weather conditions. As aresult, stocks have reached historic lows, andwheat prices have jumped 30 percent. From amacroeconomic perspective, these price in-creases have hit low-income countries thehardest, resulting in a terms-of-trade loss equalto 0.5 percent of their GDP, with the pooresturban and nonfarming rural segments of thepopulation bearing the greatest burden. Whileexperience shows that direct and targeted in-come support, rather than price controls, is themost effective way to help these vulnerableconsumers, the institutional requirements forsocial safety nets can be daunting.

Risks

The financial turbulence that emerged in mid-2007 has demonstrated how sudden and

pervasive adjustments in financial markets canbe. Because the dynamics of financial behaviorare inherently difficult to control, and because

new securitized instruments have made identify-ing the location or magnitude of underlying riskdifficult, the possibility of a breakdown in a keyfinancial institution or system cannot be fullydiscounted. Moreover, the likelihood of finan-cial problems would increase rapidly if homeprices in the United States were to fall precipi-tously, an event that could push the U.S. econ-omy into recession. Such circumstances, and thelikely U.S. monetary policy reaction, wouldreinforce the dollar’s slide, with a consequentdestabilizing effect on global markets.

To date, strong fundamentals in developingcountries have helped mitigate the slowdown inthe United States, but in the case of a major dis-ruption, adverse effects in emerging markets areunlikely to be avoided, which at some pointwould exacerbate the U.S. slowdown. Substan-tially tighter financial conditions could generatea credit crunch that would have consequencesfor investment and growth in middle-incomecountries. Low-income countries would alsosuffer substantial repercussions resulting fromweaker global demand for commodities, pricedeclines, and terms-of-trade losses. Even with-out further turmoil in international financialmarkets, several developing countries havebecome more vulnerable to financial pressureas a result of heightened anxiety and increasedvolatility in foreign exchange markets.

Another important risk is that the looseningof monetary policy in response to the U.S. sub-prime mortgage crisis could cause growth toovershoot. Commodity markets could tightenfurther, inflationary pressures would mount,and financial imbalances would increaserather than recede. Such a scenario could sowthe seeds of a much sharper slowdown in themedium term and illustrates the current chal-lenge facing monetary authorities in both high-income and developing countries.

Financial markets: Neededcorrection or major disruption?

The financial market turmoil of the secondhalf of 2007 resulted from the interaction

of several factors. An extended period of


18

abundant liquidity and low interest rates world-wide sparked a search for yield that inducedmany investors to take on additional risk. Thiswas supported by robust global growth andfavorable financial conditions, fueling a four-year expansion in the global credit cycle. Rapidgrowth in the market for asset-backed securitiesand structured financial products (collateral-ized debt obligations in particular) throughoutmajor financial centers facilitated both lending(by making the calibration and offloading ofrisk easier) and borrowing (by effectively in-creasing liquidity and the availability of credit).Emerging market bond spreads declined torecord lows, and equity prices increased rapidlyin many developing countries during the firsthalf of 2007. However, the degree of risk wasespecially underestimated in the lower creditsegments of the U.S. mortgage market (sub-prime and “alt-A” loans), and hence the valueof many asset-backed securities was grosslyoverestimated.

Corrections to this overvaluation begansuddenly in late July, and rising default ratesin the U.S. subprime mortgage market spilledover into equity, currency, and bond marketsworldwide. Credit conditions for corporateborrowers tightened significantly, while gov-ernment bond yields declined sharply in whatis known as a “flight to quality.” Spreads onnoninvestment grade U.S. corporate securitieswidened by 200 basis points in July and thefirst half of August, indicating that investors’appetite for risk had diminished considerably(figure 1.1). In mid-August, the U.S. FederalReserve and the European Central Bank pro-vided ample liquidity to the banking system tohelp stabilize financial conditions.

The sell-off in risky assets served to widenemerging market bond spreads by about 100basis points by mid-August, raising the cost ofcapital for corporate borrowers in both ma-ture and emerging markets. As financial con-ditions tightened once more near the end ofthe year, U.S. high-yield spreads jumped to600 basis points by the end of Novemberand emerging market spreads retreated, thenincreased to 270 basis points, with the overall

widening attributable to the current episodemoving to 170 points.

Even though the turmoil has affectedemerging markets, so far the financial fallouthas been limited, though nevertheless more se-rious than other, fairly short-lived episodes ofmarket turbulence and volatility that haveoccurred since 2005 (figure 1.2). Flight to

P R O S P E C T S F O R D E V E L O P I N G C O U N T R I E S

19

Figure 1.1 The perceived riskiness of high-yield corporate bonds increased more thanthat of emerging market bonds

Sources: Bloomberg, JPMorgan-Chase, and World Bank.

Note: EMBIG � Emerging Market Bond Index-G.

Jan.

2, 2

007

Feb. 1

1, 2

007

Mar

. 23,

200

7

May

2, 2

007

Jun.

11, 2

007

Jul. 2

1, 2

007

Aug. 3

0, 2

007

Oct. 9

, 200

7

Nov. 1

8, 2

007

600

100

550

500

450

400

350

300

250

200

150

High-yield spreads indeveloped markets

Basis points

EMBIG spreads

Figure 1.2 Emerging market asset sell-offmore severe than during earlier periods of market turbulence

Sources: Bloomberg and World Bank.

Mar.–Apr.2005

Oct.–Nov.2005

May–June2006

July–Oct.2007

0

�10

�20

�60

�50

�40

�30

Size of drawdown over peak to trough (percent)

Emergingmarketequities

Emerging marketbonds

quality and the need to cover losses in thesubprime market provoked a sell-off acrossthe entire spectrum of high-yield assets in ma-ture and emerging markets. Equity price de-clines in emerging market economies initiallyexceeded those in mature markets, but emerg-ing markets rebounded sharply, outpacinggains in mature markets (figure 1.3). TheMorgan-Stanley composite index of emerg-ing-market stocks picked up close to 50 per-cent from the beginning of the year, wellabove the developed markets, before bothretreated in tandem by late November. Therebound in emerging market equities was

underscored by a resumption of inflows toequity funds, which had experienced outflowsof some $5 billion during late July and earlyAugust. Until recently, corrections were globalin nature, and stock exchanges in East Asiaand the Pacific and Latin America and theCaribbean were continuing to drive solidrecovery in emerging market equities.

Gross capital flows to developing countriesshowed strong gains in 2007 before financialuncertainties arose. Bond issuance, bank loancommitments, and equity placements togetheraveraged $53 billion a month from Januarythrough July, up from $41 billion during2006, but a decline in August dropped flowsto $42 billion (table 1.1). The surge in flowsbefore August was concentrated in bond is-suance and equity placements, and these cate-gories initially experienced the steepest falloffafter the turmoil. By October, bond and eq-uity flows had recovered fully or almost fully,but a sharp falloff in bank lending emerged,with commitments dropping $25 billion dur-ing the month. Viewed on the basis of onlymoderate increases in sovereign spreads, thelack of bond issuance in August and Septem-ber may have reflected decisions by govern-ments in developing countries to postponenew issuance because of limited financingneeds rather than an inability to access themarket. However, for corporate borrowers inemerging economies, which accounted for80 percent of bond issuance during 2007,financial conditions have deteriorated. Thedecline in banking flows is a concern, possiblyreflecting a partial near-term withdrawal from


20

Table 1.1 Gross capital flows to developing countries, 2005–07(monthly averages, $ billions)

2007

2005 2006 January–July August October

Total 30 41 53 42 50Bond issues 11 11 22 3 15Bank loan commitments 15 20 18 35 10Equity placements 5 9 13 4 25

Source: Dealogic Loanware and Bondware.

Figure 1.3 Global equity markets fall, thenrecover led by emerging markets

Source: Thomson/Datastream.

Note: DJIA � Dow Jones Industrial Average; MSCI �Morgan-Stanley Composite Index; TOPIX � Tokyo StockPrice Index; DAX � Deutsche Aktien Exchange.

Equity market index (January 1, 2007 � 100)

Jan.

1, 0

7

Feb.

15, 0

7

Apr. 1

, 07

May

16,

07

Jun.

30, 0

7

Aug. 1

4, 0

7

Sep. 2

8, 0

7

Nov. 1

2, 0

7

150

80

140

130

120

110

100

90

MSCI emergingmarkets

DAX (Germany)

DJIA (U.S.) TOPIX (Japan)

emerging markets, as banks tighten creditcriteria and assume a more risk-averse postureas they replenish reserves after sharp losses insubprime securities.

Global growth

After four years of robust GDP and tradegrowth, steadily increasing commodity

prices, low bond market spreads, graduallychanging interest rates, and relatively stableexchange rates, volatility in international mar-kets has increased. Conditions in global finan-cial markets have turned from exceptionallyfavorable to less stable and less predictable.

More than in recent years, reserves andother buffers will be needed to absorb unex-pected shocks. Policy makers must prepareboth for the possibility that their economiesmay slow sharply and for the possibility thatgrowth may continue to exceed potential.Similarly, they must prepare for the possibilityof an abrupt depreciation of their currenciesas well as the possibility that continued capi-tal inflows could push them up. Commodityprices may spike, or they could give up part ofthe gains realized this decade.

Despite such a volatile climate, aggregategrowth is likely to remain robust for the de-veloping countries, mainly because of strongdomestic momentum in most of them. Indeed,economic performance for many developingeconomies was exceptionally robust duringthe first half of 2007, much stronger than an-ticipated in Global Development Finance inearly 2007 (World Bank 2007a).

Table 1.2 and figure 1.4 summarize recentdevelopments and the base case outlook.World growth eased from 3.9 percent in 2006to 3.6 percent in 2007, with the slowdown ledby members of the Organisation for EconomicCo-operation and Development (OECD).Their GDP dipped by 0.3 percentage points to2.5 percent in the year. The downturn wasmore marked in the United States, withgrowth slowing from 2.9 percent in 2006 to2.2 percent in 2007. Much of the decline re-

flected the direct fallout of the weakeninghousing market, with residential investmentfalling rapidly, and credit conditions for bothfirms and consumers tightening.

Among developing countries, growth re-mained firm at 7.4 percent in 2007, after anequally strong 7.5 percent in 2006, under-pinned by continued strength in East andSouth Asia. If China and India are excluded,activity in low- and middle-income countriesslipped by 0.2 percentage points to 5.7 percentin the year.

In 2008, global growth is expected to mod-erate further, as the effective cost of capital re-mains elevated for financial institutions, firms,and households. Weak domestic demand is ex-pected to keep U.S. GDP growth below 2 per-cent in 2008, while growth in Europe andJapan should continue to ease under the addi-tional weight of appreciating currencies.OECD import demand is projected to movefrom a solid 6.8 percent gain in 2007 to 5.4 per-cent during 2008, slowing export growth in de-veloping countries by a point to 11 percent anddampening their output growth to 7.1 percent.

The OECD countries are anticipated to re-cover during the course of 2009, as returningstability in financial markets helps revive con-sumer and business confidence and residential


21

Figure 1.4 A step-down in growth in 2008

Source: World Bank.

1995

1997

1999

2001

2003

2005

2007

2009

World

Developing countries


8

6

4

2

0

Real GDP (percent change)Forecast


22

Table 1.2 The global outlook in summary, 2005–09(percent change per annum, except where otherwise indicated)

Estimate Forecast

2005 2006 2007 2008 2009

Global conditionsWorld trade volume 7.8 10.1 9.2 7.6 9.2Consumer prices

Group of seven countries a,b 2.0 2.0 1.7 1.7 1.9United States 3.4 3.2 2.8 2.6 2.4

Commodity prices (US$ terms)Non-oil commodities 13.4 24.5 15.3 �0.7 �4.6

Oil price US$ per barrel c 53.4 64.3 71.2 84.1 78.4Percentage change 41.5 20.4 10.8 18.1 �6.8

Manufactures unit export value d 0.0 1.6 2.3 0.8 0.8Interest rates

$, 6-month (percent) 3.7 5.2 5.3 4.8 5.0€, 6-month (percent) 2.2 3.2 4.3 4.0 4.3

Real GDP growth e

World 3.4 3.9 3.6 3.3 3.6World (PPP weights) f 4.8 5.3 5.2 4.9 5.1High-income countries 2.6 2.9 2.6 2.2 2.6

OECD countries 2.4 2.8 2.5 2.1 2.4Euro Area 1.5 2.8 2.7 2.1 2.4Japan 1.9 2.2 2.0 1.8 2.1United States 3.1 2.9 2.2 1.9 2.3Non-OECD countries 5.8 5.6 5.1 4.8 5.0

Developing countries 6.8 7.5 7.4 7.1 7.0East Asia and the Pacific 9.1 9.7 10.0 9.7 9.6

China 10.4 11.1 11.3 10.8 10.5Indonesia 5.7 5.5 6.3 6.3 6.5Thailand 4.5 5.0 4.3 4.6 5.2

Europe and Central Asia 6.1 6.9 6.7 6.1 5.7Poland 3.6 6.1 6.5 5.7 5.1Russian Federation 6.4 6.7 7.5 6.5 6.0Turkey 7.4 6.1 5.1 5.4 5.7

Latin America and the Caribbean 4.6 5.6 5.1 4.5 4.3Argentina 9.2 8.5 7.8 5.7 4.7Brazil 2.9 3.7 4.8 4.5 4.5Mexico 2.8 4.8 2.9 3.2 3.6

Middle East and North Africa 4.3 5.0 4.9 5.4 5.3Algeria 5.1 1.8 3.4 4.0 3.8Egypt, Arab Rep. of 4.4 6.8 7.1 7.0 6.8Iran, Islamic Rep. of 4.3 4.6 5.0 5.0 4.7

South Asia 8.7 8.8 8.4 7.9 8.1Bangladesh 6.0 6.6 6.5 5.5 6.5India 9.2 9.4 9.0 8.4 8.5Pakistan 7.7 6.9 6.4 6.5 6.7

Sub-Saharan Africa 5.8 5.7 6.1 6.4 5.8Kenya 5.8 6.1 6.3 5.3 5.1Nigeria 6.6 5.6 5.9 7.4 6.1South Africa 5.0 5.4 5.0 5.1 5.3

Memo itemsDeveloping countries

Excluding transition countries 6.9 7.5 7.5 7.2 7.1Excluding China and India 5.2 5.9 5.7 5.3 5.2

Source: World Bank.Note: OECD � Organisation for Economic Co-operation and Development; PPP � purchasing power parity.The CPI figures in this table that were released on January 9, 2008, reflected typographic errors. Current figures were released onJanuary 11, 2008.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 dollars; 2000 prices and market exchange rates.f. GDP measured at 2000 PPP weights.

investment bottoms out. On aggregate,growth in developing countries is expected tobe robust in both 2008 and 2009, remainingat or above 7 percent.

The high-income countries Among OECD countries, the first quarters of2007 appeared to be a prelude to morevolatile growth (figure 1.5). U.S. GDP weak-ened sharply in the first quarter before re-bounding to 3.8 and 4.9 percent in the secondand third quarters on the strength of businessinvestment in the second quarter, surprisinglystrong consumer demand and stock-buildingin the third, and strong net exports in both.But high-frequency data point to weaker con-sumption growth in the fourth quarter, andfor the year as a whole, 2.2 percent growth isexpected, 0.7 percentage points below 2006results (figure 1.6).

In contrast, Japan and the Euro Area main-tained a favorable pace of growth in the firstquarter, with business confidence breachingrecord highs, but developments in the secondquarter were disappointing. In Europe, a re-trenchment in business capital outlays morethan halved GDP gains of the previous quar-ter, while in Japan, a slide in fixed investmentturned growth into a decline. Third quarter re-sults for Europe provided an upside surprise,

with growth returning to a favorable 2.9 per-cent. GDP gains were broadly based acrosscountries, while business investment, stocks,and consumer spending in France andGermany revived to spur overall growth.

The Japanese economy rebounded mod-estly in the third quarter as well to registergrowth of 1.5 percent after a 1.8 percent de-cline in the previous quarter based on muchimproved net exports and a moderate boost tohousehold spending. For 2007 as a whole,European growth is expected to register astrong 2.7 percent, eclipsing the United Statesfor the first time in more than a decade, andgrowth in Japan should register 2 percent.

GDP growth in the United States isprojected to weaken further in 2008, falling to1.9 percent. During the year, continuing diffi-culties in the commercial paper market, thesource of working capital for most U.S. busi-ness, implies a boost in the effective cost ofshort-term funds, despite a cumulative reduc-tion of 100 basis points in Federal funds overSeptember through December, which carriedthe rate to 4.25 percent. Recovery is antici-pated for 2009, with growth registering


23

Figure 1.5 Volatile patterns of growthamong OECD countries

Source: National agencies and Eurostat.

Note: Q � quarter.

United States Japan Euro Area Germany

6

�2

0

2

4

Growth at seasonally adjusted annualized rates

Q3, 2007

Q2, 2007

Q1, 2007

Figure 1.6 Tighter credit and weakhousing yield slower U.S. growth

Source: World Bank.

Note: Q � quarter.

Q1, 2

005

Q3, 2

005

Q1, 2

006

Q3, 2

006

Q1, 2

007

Q3, 2

007

Q1, 2

008

5Forecast

4

2

3

1

0

Real GDP growth at annual rates (percent)

2.3 percent, grounded in a stabilization of fi-nancial markets and a revival of business andhousehold spending.

As a result of weaker domestic demand andstronger export performance (in part based ona substantially weaker dollar and continuedsolid demand from emerging markets), theU.S. current account deficit is expected to de-cline from its high of 6.6 percent of GDP in2006 to around 5 percent by 2009. Inflation isanticipated to moderate toward 2.5 percent,in step with slower growth, while householdsavings could move toward positive groundfor the first time in many years. Developmentsin the United States are expected to influenceconditions in Japan, both because of Japan’sreliance on trade as a source of growth andbecause of the sensitivity of business invest-ment to the costs of long-term capital. This isof particular note given Japan’s experiencewith the yen carry trade.1 Because such tradecan potentially unwind rapidly, with local cur-rency proceeds used to redeem yen, substan-tial appreciation of the currency could hamperexports, production, and incomes. Growth inJapan is expected to slow to 1.8 percent in2008 before picking up to 2.1 percent by2009.

Household spending in the Euro Area hasnot yet fully recovered from the increase in theGerman value added tax of January 2007,although unemployment in Europe eased to

7.3 percent in September 2007, the lowestlevel since figures began to be compiled in1993. Business investment continues to movein close step with export performance, whichdeteriorated toward the end of the year, inpart because of the appreciation of the euroand slowing U.S. demand. As exports softenand business investment declines, GDPgrowth is projected to slip to 2.1 percent in2008 before reviving to 2.4 percent in 2009.

The developing countries The first half of 2007 was marked by anacceleration in industrial production acrossdeveloping regions, notably in East Asia(20 percent, year over year) (figure 1.7 andtable 1.3).2 A pickup in China’s production,together with momentum gains in Indonesia(7.3 percent) and Thailand (6.0 percent), re-sulted partly from a bottoming out of thehigh-tech slump in the second quarter of theyear. South Asia’s continued buoyant produc-tion gains are linked to double-digit growthin India, which is tied in turn to strong do-mestic demand. Latin America and theCaribbean achieved a 6.4 percent increase inindustrial production during the second quar-ter, at a seasonally adjusted annual rate, upfrom 2.4 percent during the first. Within theregion, strong performance in Brazil (10 per-cent), Colombia (13 percent), and Mexico(5.5 percent) have offset weakening output


24

Industrial production (percent change, seasonally adjusted annualized rate)

Figure 1.7 Robust growth in developing country industrial production

28

24

20

16

12

8

4

0Jan.2005

May2005

Sep.2005

Jan.2006

May2006

Sep.2006

Jan.2007

May2007

Sep.2007

Source: World Bank.

Developingcountries

OECD

World


25

Table 1.3 Recent economic indicators, developing regions, 2005–07

Growth year- Seasonally adjusted Growth year-on-year annualized growth on-year

Indicator and region 2005 2006 Q1, 2007 Q2, 2007 H1, 07 Latest

GDP growth (percent) a

Developing countries 6.8 7.5 8.5 7.0 7.5 —East Asia and the Pacific 9.1 9.7 10.3 10.5 9.8 10.2Europe and Central Asia 6.1 6.9 3.8 4.3 6.3 —Latin America and the Caribbean 4.6 5.6 7.5 5.0 5.9 5.7Middle East and North Africa 4.3 5.0 4.8 4.5 4.7 —South Asia 8.7 8.8 15.6 7.7 9.2 —Sub-Saharan Africa 5.8 5.7 4.9 4.0 5.0 —

Industrial production growth (percent)Developing countries 9.3 10.1 14.6 18.5 14.0 11.4

East Asia and the Pacific 13.8 14.3 19.6 28.9 19.6 16.4Europe and Central Asia 1.4 4.5 5.6 0.0 3.7 5.0Latin America and the Caribbean 3.9 4.3 2.4 6.4 3.8 4.6Middle East and North Africa 4.4 �0.3 �1.9 1.5 �0.5 �1.2South Asia 9.1 11.0 15.6 11.3 13.0 8.0Sub-Saharan Africa 5.0 5.3 — — — —

Consumer prices (year over year growth rates in percent)Developing countries 6.4 6.2 5.7 6.1 5.9 6.1

East Asia and the Pacific 7.2 4.9 4.3 3.4 3.8 3.1Europe and Central Asia 4.4 5.6 4.5 4.8 4.7 7.1Latin America and the Caribbean 5.4 5.6 5.8 5.0 5.4 4.7Middle East and North Africa 2.8 1.8 4.9 4.2 4.5 5.0South Asia 7.0 7.6 7.1 6.4 6.7 6.9Sub-Saharan Africa 3.4 4.6 6.3 7.1 6.7 7.1

Spreads on debt (basis points)Developing countries 306 198 176 162 169 243

East Asia and the Pacific 265 180 138 129 134 207Europe and Central Asia 185 137 147 134 141 192Latin America and the Caribbean 364 213 183 167 175 259Middle East and North Africa 324 338 418 434 426 548South Asia 199 199 166 171 169 482Sub-Saharan Africa 277 266 283 270 277 321

Source: National statistical agencies through Thomson/Datastream; spreads: JPMorgan.Note: Q � quarter. Quarterly 2007 growth for developing regions based on data available for key economies.EAP: China, Indonesia, Malaysia, Philippines, Thailand. SAR: India. ECA: Hungary, Poland, Russian Federation, Turkey. LAC: Argentina, Brazil, Chile, Colombia, Mexico. MENA: Arab Rep. of Egypt. SSA: Nigeria, South Africa. — � not available; H1 � 1st half.a. Growth rates at annual or annualized rates unless otherwise indicated.

gains in Argentina and the RepúblicaBoliviarana de Venezuela.

Robust production data are also reflectedin GDP results. Led by the large economies,notably China, India, and the Russian Federa-tion, output for the developing countriesaveraged 7.5 percent (year-on-year) in the firsthalf of 2007, matching the record pace of2006 (table 1.3 and figure 1.8). Becausedeveloping economies have been less affectedby the fallout from the subprime crisis than high-income economies, the anticipated slow-

down of growth in 2008 should be less pro-nounced. Gains among developing countriesare projected to slow from 7.4 percent in 2007to 7.1 percent in 2008, with easing across allregions except the Middle East and NorthAfrica and Sub-Saharan Africa, in part becauseof higher oil revenues. For the remaining coun-tries, slower export market growth, only grad-ual improvement in financing conditions, anddeclines in the terms of trade account for mostof the slowdown. Though a further easing ofgrowth to 7 percent is anticipated for 2009,


26

global conditions should favor a smootherstep-down in activity for those countries andregions more dependent on trade, notably EastAsia and Latin America (figure 1.9).

Developing region perspectivesGDP in East Asia and the Pacific is expected togrow about 10 percent in 2007, the strongest

performance since 1994. Growth in China isexpected to exceed 11 percent (table 1.2). Else-where in the region, strong investment demandboosted growth in Indonesia by almost apercentage point, from 5.5 percent in 2006 to6.3 percent in 2007. Tighter monetary policyand the economy’s absorption of the directeffects of the removal of energy subsidiesin 2006 have nearly halved inflation, from13 percent in 2006 to 7 percent in 2007.Growth is expected to remain in the 6 percentrange. Growth in Malaysia and the Philippinesappears to have picked up as well, coming in atnear 6 percent or better, reflecting a relaxationof monetary policies and improved externaldemand for electronics. Growth in Thailandhas been more subdued during 2007, at a littleover 4 percent, tied in part to political unrest,as well as a lagging response to the high-techrevival.

Assuming that China continues its double-digit growth and that the authorities succeedin dampening overheated sectors, GDPgrowth in East Asia should slow gradually to9.7 percent in 2008 and to 9.6 percent by2009. Many countries in the region could,however, experience a stalling of exportgrowth and a loss of business and consumerconfidence, leading to a moderate falloff inGDP gains in 2008.

The effects from the turmoil in the world’sfinancial centers may be minimal for mosteconomies in East Asia. Except for China, di-rect exposures of financial institutions in theregion to mortgage-backed securities (or sub-prime risks) are limited.3 Economies in EastAsia and the Pacific are entering this moreturbulent period from a position of relativestrength: improved macroeconomic fundamen-tals, further movement into current accountsurplus, and rapidly increasing reserves. Yetseveral of these economies have been the objectof international investor interest through carrytrades.3 Should conditions in the mature mar-kets deteriorate so that assets are shifted out ofclasses perceived as risky to cover other losses,the threat of an unwinding of the yen carrytrade, with attendant down-spikes in local

Figure 1.8 Developing growth retainsstrong momentum during the first halfof 2007...

Sources: World Bank and National agencies.

Note: H1 � 1st half.

East A

sia

South

Asia

Europ

e an

d

Centra

l Asia

All dev

elopin

g

coun

tries

Latin

Am

erica

and

the

Caribb

ean

Sub-S

ahar

an A

frica

Midd

le Eas

t and

North

Afri

ca

10

2

4

6

8

GDP growth (percent)

H1, 2007

2006

2005

Figure 1.9 ...with growth moderatingthrough 2009

Source: World Bank.

East A

sia

South

Asia

Europ

e an

d

Centra

l Asia

All dev

elopin

g

coun

tries

Latin

Am

erica

and

the

Caribb

ean

Sub-S

ahar

an A

frica

Midd

le Eas

t and

North

Afri

ca

10

2

4

6

8

Real GDP (percent change)

2009

2008

2007

equity markets and rapid currency deprecia-tion, is a concern for policy makers.

Possible side effects related to financialmarket turbulence—notably a distinct soften-ing of U.S. and European import demand—could affect all East Asian countries. The EuroArea and the United States together accountfor 43 percent of China’s export market, whileJapan accounts for 7.5 percent. With the de-velopment of China as a hub for the final as-sembly and shipment of goods to destinationmarkets, with parts and materials supplied byother East Asian economies (figure 1.10),slower import demand from developed coun-tries could adversely affect the entire region.Some East Asian economies could experiencea double hit, sustaining a loss of both directand indirect import demand.

In the Europe and Central Asia region,investment and external demand are bothslowing, leading to a moderate easing ofgrowth from 6.9 percent in 2006 to 6.7 per-cent in 2007. Private consumption, fed byrobust credit creation, accounted for 4.6 per-centage points of the advance in 2007. Invest-ment, which accounted for 3.4 percentagepoints of growth in 2007, was driven by pol-icy reforms, improved business confidence

tied to European Union (EU) accession by sev-eral Central and Eastern European countries,and continuing high oil prices that have fueleda construction boom in several countries ofthe Commonwealth of Independent States(CIS).

The falloff in external demand took hold inthe second half of 2007, slowing growth by2 percentage points, as net exports continuedto deteriorate. Inflation has risen in severalcountries, tied to sustained strong growth indomestic demand, rising fuel and food prices(the latter aggravated by drought in Bulgariaand Romania), and higher world grain prices.Rapid credit expansion in most of the regionhas contributed to a worsening of externalbalances while exerting upward pressure onasset, goods, and labor market prices (fig-ure 1.11). Signs of overheating are evident inBulgaria and the Baltic states, for example,where external positions have deterioratedfrom already high levels. Effects on the regionstemming from the crisis in the U.S. subprimemortgage market have thus far been limited.Initial downward adjustments to asset prices


27

Figure 1.10 East Asia now accounts forone-quarter of China’s imports

Source: National sources through Haver Analytics.

Hong

Kong,

China

Indo

nesia

Philipp

ines

Korea

, Rep

. of

Singap

ore

Thaila

nd

12

0

4

8

10

2

6

(percentage of total imports)

Taiw

an, C

hina

Europ

e an

d

Centra

l Asia

Centra

l and

Easte

rn E

urop

e

Uzbek

istan

Comm

onwea

lth o

f

Inde

pend

ent S

tate

s

Russia

n Fed

erat

ion

Ukrain

e

Poland

Hunga

ry

Roman

ia

Latvi

a

20072006

2005

Figure 1.11 External positions vary widelyacross Europe and Central Asia

Source: World Bank.

�25

0

15

20

10

5

�5

�10

�15

�20

Current account balance as a percent of GDP

and currencies have since been recouped andbond spreads increased, but to a lesser extentthan in other markets.

Regional GDP growth is expected to slowto 6.1 percent in 2008 and 5.7 percent in2009. Given tighter international credit condi-tions and weakening external demand, thegrowth falloff in 2008 is likely to affect almostthe entire region. Domestic demand is alsoprojected to soften slightly from its recentstrong growth, with contributions to GDP ofboth private consumption and investmentdropping by 0.2 percentage points during2008.

Three exceptions to this growth forecastfor 2008 are Albania, Hungary, and Turkey. InAlbania, growth is likely to firm on continuedstrong domestic demand, including an in-crease in public investment. That investment iscrucial to overcome substantial problems ininfrastructure, as power shortages now form asignificant bottleneck to growth. In Hungaryand Turkey, further easing in monetary policyis expected to strengthen domestic demand,leading to a pickup in growth.

External demand is projected to revive by2009 as GDP growth among the OECD coun-tries picks up. After slowing GDP by 2.2 per-centage points in 2008, net exports areprojected to boost growth by 1.7 percentagepoints in 2009. This gain is expected to be off-set by further slowing in domestic demand,particularly investment in the CIS countries,yielding GDP growth of 5.7 percent in 2009.The expected falloff in investment in the CIS isdriven by the completion of hydrocarbon in-frastructure projects that have supported pro-duction and export capacity in recent years.

In Latin America and the Caribbean, GDPadvanced by 5.1 percent in 2007, the fourthconsecutive year of sustained growth. Theaverage rate of output gains over 2005–07was 5.3 percent, twice the 2.7 percent regis-tered during the previous 15 years. Recentgrowth has also been broadly based, with pos-itive results shared across subregions: SouthAmerica, Central America, and the Caribbean(figure 1.12). A favorable external environ-

ment, together with better macroeconomicmanagement, helped improve fundamentals,reduce the volatility of economic indicators,and sustain growth. GDP in the region pickedup to a 5.9 percent pace in the first half of2007 on the back of continued strong activityin Argentina, Brazil, Chile, and the RepúblicaBoliviarana de Venezuela. In addition, themid-2007 credit market turmoil that hit theUnited States seems to have had limited effectson Latin America and the Caribbean to date.To a degree, the recent upturn in growth willserve as a buffer. Even though any stagnationin the United States will eventually affect re-gional prospects, countries seem better pre-pared for exogenous shocks than during priorepisodes of financial market disturbance.

In contrast with previous expansionphases—and indeed, with previous episodes ofcrisis—Latin America and the Caribbean isnow recording a healthy current account sur-plus and accumulating large stocks of interna-tional reserves. The improvements of recentyears might indeed be sufficient to ward off


28

Figure 1.12 Growth eases in 2007 for theLatin America and Caribbean region

GDP growth 2005–07 (percent)

0 3 6 9 12

Sources: World Bank and national agencies.

Latin Americaand the Caribbean

Venezuela, R.B. de

Argentina

DominicanRepublic

Colombia

Honduras

Chile

Brazil

Mexico

Jamaica

200720062005

some of the adverse effects of developments inthe United States. For example, at the time ofthe 1998 Russian crisis, the current accountdeficit for the region was close to $89 billionor 4.4 percent of GDP. In 2003, the deficit re-versed to a surplus, and in 2006, the regionhad a surplus of more than $46 billion, equiv-alent to 1.6 percent of GDP. Trade liberaliza-tion and flexible exchange rates are among thefrequently cited policies that facilitated theseimprovements in external balances.

On the heels of strong growth in 2007,regional GDP growth is expected to ease to4.5 percent in 2008 and further to 4.3 percentby 2009. Such measured regional slowdownis underpinned by continued strong growthin Brazil and a rebound from a weak 2007 inMexico, while growth in other countries—notably Argentina and the RepúblicaBoliviarana de Venezuela—is likely to slow.Excluding the latter two countries, GDP mod-erated slightly from 4.7 percent in 2006 to4.4 percent in 2007 and, following an antici-pated dip to 4.2 percent in 2008 because ofweak import demand in the United States, isexpected to recoup to 4.3 percent by 2009.Should these outturns be realized, they wouldrepresent the longest positive growth spell forLatin America and the Caribbean since the1960s. Despite a gradual worsening of currentaccount positions because of stabilizing com-modity prices and slower growth in globaldemand, strong growth is likely to persist,supported by continued expansion in con-sumption and investment, buoyed by an envi-ronment of low inflation (excluding Argentinaand the República Boliviarana de Venezuela),improved fiscal policy (particularly in Mexico),and continued strong capital inflows (especiallyto Brazil).

High oil prices have continued to supportgrowth for the oil-exporting countries in theMiddle East and North Africa, coming on theheels of a 5 percent regional GDP advance in2006, the fastest in a decade for the region’sdeveloping countries.4 Oil prices, whichspiked to almost $100 per barrel late in theyear and averaged $71 per barrel for 2007, are

further buttressing government revenues,some of which are dedicated to infrastructurespending in the developing oil exporters (prin-cipally Algeria and the Islamic Republic ofIran). At the same time, a solid pickup inEuropean growth over the course of 2007 hasfavored economic activity among the diversi-fied exporters in the Maghreb and Mashreq,which have especially close trade and servicesties with the Euro Area. GDP in the MiddleEast and North Africa appears to have faredwell over the first half of 2007, though easingfrom 2006 to a 4.7 percent pace during the pe-riod, with the slowing tied in part to a returnto drought conditions that have affectedseveral countries in the Maghreb, notablyMorocco.

The run-up of surplus funds among oil ex-porters and the availability of new investmentopportunities across the region boosted for-eign direct investment (FDI) flows to newhighs of more than $24 billion in 2006, or3.4 percent of regional GDP. At the same time,recovery of local equity markets offerspromise that after bouts of overheating during2004–05 and stabilization following the re-cent tensions in global financial markets, a lessvolatile source of funds may play a larger rolein the region’s growth.

Developing country oil exporters in theMiddle East and North Africa registeredgrowth of 4.5 percent in 2007, half a percent-age point up from 2006, as GDP in the IslamicRepublic of Iran advanced to 5 percent, whileoutput in Algeria almost doubled from thepoor 1.8 percent showing of 2006. Hydrocar-bon revenues (oil and natural gas) flowing todeveloping country exporters amounted to$167 billion in the year, representing an in-crease of 5.7 percent, or $9 billion, over 2006results. In contrast, export revenues for thehigh-income country exporters grew by 6 per-cent to $360 billion, an increment of $20 bil-lion, over 2006 levels (figure 1.13).5

For diversified exporters, trade respondedwell to increased European demand, and to adegree by a depreciation of local currencies inrelation to the euro. Exports of goods and


29

services for Jordan, Morocco, and Tunisia as agroup jumped to 10.2 percent growth in theyear, in step with a 2.5 percent increment inEuro Area demand, setting the stage for im-proved output growth. At the same time, a riskfacing the region’s textiles and clothing ex-porters, as well as many other low- and mid-dle-income country exporters, is the comingremoval of remaining barriers to Chinese ex-ports of specific textiles and clothing products(box 1.1).

A number of factors are likely to shape thegrowth profile for the Middle East and NorthAfrica region. A softening of OECD demandis anticipated for 2008, although it may be ac-companied by additional firming of global oilprices tied to continued robust demand inemerging markets. This should benefit the oilexporters for a time, and should supportregional growth of 5.4 percent. As the globalenvironment improves by 2009, the MiddleEast and North Africa region should be ableto maintain the broader pace of growth estab-lished in 2008 for several more years. Domes-tic conditions will vary decidedly across the

disparate economies of the region, as willefforts at reform, which in most cases areaimed at spurring private sector or non-oil ac-tivity. Tensions related to continuing conflictin Iraq, unsettled conditions in the Levant,and international disputes with the IslamicRepublic of Iran will tend to affect global andregional investors’ confidence concerning theregion, and any projections exercise about theregion should take these into account as a risk.

In South Asia, regional GDP growth re-mained vibrant at 8.4 percent in 2007, thougheasing somewhat from the stellar 8.8 percentgains of 2006. Industrial production and GDPgrowth are being driven by strong domesticdemand, with private consumption and in-vestment each contributing about 4 points toGDP growth in the year. An expansion ofcredit, rising incomes, and strong worker re-mittance receipts are underpinning privateconsumption, while improvements in businesssentiment—both foreign and domestic—alongwith rising corporate profits, are providing afurther boost to investment. Despite more re-strictive monetary policy and progress towardgreater fiscal consolidation in a number ofcountries, domestic demand growth haspicked up, building on the momentum of re-forms undertaken in recent years. Moreover,because of tighter policy conditions, inflationpressures moderated during the first half of2007 in the larger regional economies of Indiaand Pakistan. The risks of revived inflation re-main, in part because of the still incompletepass-through of high energy prices and up-ward pressures on food prices.

Current account balances deteriorated in anumber of countries in 2007, with deficitsreaching close to 5 percent of GDP in Pakistanand Sri Lanka and about 2 percent in India.Pakistan’s current account deficit is a concern,having deteriorated by the equivalent of morethan 5 percentage points of GDP during thelast four years. In India, monetary tighteningand large capital inflows led to substantial ap-preciation of the rupee over the year, with thecurrency reaching a near decade high of Rs39.30 against the dollar by late November


30

Figure 1.13 Continued oil revenue gainssupport growth among Middle East andNorth Africa oil exporters

502002 2003 2004 2005 2006 2007

100

150

200

250

300

350

400

Oil price(right axis)

450

500

550

25

30

35

40

45

50

55

60

65

70

75

$/bn

Crude oil and product revenues ($ billions) oil price,World Bank ($/barrel)

$/barrel

Sources: UN Comtrade database; International MonetaryFund; International Energy Agency; World Bank.

Low- and middle-income country exporters

High-income exporters


31

The system of quantitative restrictions that man-aged rich countries’ imports of textiles and cloth-

ing from developing countries for 30 years, especiallythose produced in China and India (the Multi-FiberArrangement), was finally dismantled at the end of2004, although restrictions for a number of cate-gories of Chinese textile and clothing exports to theEU and the United States remained because ofmeasures that are due to expire in 2008.

China’s exports of clothing soared 22 percent in2005 and 32 percent in 2006, increasing its marketshare in those two years to 24 percent and 28 per-cent, respectively, but the impact on competitors hasbeen less drastic than some had feared. The increasein the size of the world market for clothing hasallowed exports from many other countries togrow, including the Arab Republic of Egypt, India,Peru, Sri Lanka, and Turkey. In Bangladesh, where 1 million jobs were predicted to be lost, exports tothe EU and the United States gained continuouslybetween 2004 and the first four months of 2007.

Nevertheless, some countries have seen declinesin clothing exports that may entail substantialadjustment. For example, exports to the U.S. andEU markets from Brazil, the Dominican Republic,Swaziland, and Taiwan (China) declined substan-tially in 2005 and 2006. With the exception ofSwaziland, clothing exports from these countriescontinued to decline into 2007. In addition, for Sub-Saharan Africa as a whole, where the end of theclothing sector had been foreseen, exports to the EUand the United States fell by 7 percent in 2004 and17 percent in 2005 (on a trade weighted average). In2006, Sub-Saharan African textile exports to the EUgrew 3 percent, whereas exports to the United Statesdeclined by 6 percent. In 2007, to the extent thatdata are available, Sub-Saharan African textile ex-ports to both the EU and the United States grewby 7 and 2 percent, respectively. A number of coun-tries, including Madagascar, Mauritius, and Swazi-land, managed to reverse an initial decline inclothing exports and return to growth in 2006 orearly 2007.

How vulnerable are other countries when thefinal restrictions on Chinese textile and clothing

Box 1.1 Developing country exports in the wake of the removal of barriers to Chinese exports

exports to the EU and the United States expire?In 2006, 19 percent of Chinese exports to the EUand 20 percent of exports to the United States weresubject to quota restrictions, and exports of theseproducts will likely grow significantly after removalof the quotas. In the EU market, Colombia, theDominican Republic, Mauritius, Peru, and Sri Lankaappear to be most at risk with more than 40 percentof their 2006 exports in product categories for whichChina is currently still subject to quotas. For othercountries, the ratio is between 20 and 40 percent,and for Sub-Saharan Africa as a whole stands at51 percent, mainly driven by the high exposure ofMauritius (74 percent). In the U.S. market, exposureis generally lower: only the Dominican Republic,India, and Sri Lanka export more than 20 percentof their textiles and clothing in categories whereChinese exports are currently subject to quotas. Formost other countries, the ratio is between 5 and20 percent. However, looking at the impact of theelimination of Multi-Fiber Arrangement quotas in2004, many competitors managed to defend theirmarket shares in recent years, and they might be ableto do so in 2008 as well.

The clothing sector still provides an opportunityfor export diversification and the expansion of man-ufactured exports for low-wage countries, even inthe face of unfettered competition from China. Thecountries best able to expand their exports of cloth-ing will be those that have a supportive business en-vironment, low trade costs (efficient customs, ports,and transport infrastructure), and competitive firmsthat are flexible enough to meet the changing de-mands of the global buyers that now dominate theindustry.

At the same time, significant adjustment pressuresmay arise as more efficient firms expand, while thoseunable to compete in the global market decline. Inthe absence of other employment opportunities, espe-cially for women, workers made redundant fromthe textile and clothing sector may fall back intopoverty. Minimizing the costs incurred by releasedworkers and their families and facilitating their ad-justment into alternative employment will be a majorchallenge for a number of developing countries.

(figure 1.14). The initial effects of increasedvolatility in international credit marketsduring the latter part of mid-2007 led to a fall-off in equity prices for both India and Pak-istan, though both countries have recoupedlosses in the interim and returns remain wellin the positive territory for the year. In localcurrency, equity markets are up 36 percentin India (50 percent in dollar terms) and20 percent in Pakistan (19 percent in dollarterms) as of the end of November 2007.

Tighter credit conditions, greater marketvolatility, increased risk of recession in theUnited States, and easing growth in the EUwill place downward pressures on regionalexports before improvement sets in by 2009.At the same time, increased competition fromChina will be a factor in shaping the path ofexport growth over the next few years, and anadditional 18 percent increase in crude oilprices in 2008 will contribute to a marked de-terioration in external balances. On balance,these factors should lead to an easing ofregional growth from 8.4 percent in 2007 to7.9 percent in 2008. Growth is projectedto pick up to 8.1 percent in 2009 as externaldemand revives and given easing pressures on

the import bill as oil price gains diminish.Domestic demand is anticipated to revive into2009, assuming inflation conditions permitsome easing of monetary policy in the secondhalf of 2008.

Growth in Sub-Saharan Africa looks poisedto remain buoyant by historic standards,maintaining a growth pace near 6 percentfrom 2007 through 2009, notwithstandingslowing demand in the United States and theEuro Area. GDP continued to grow stronglyin 2007, with output expanding 5 percent inthe first half of the year, and is expected toamount to 6.1 percent for 2007 as a whole.This follows a solid 5.7 percent advance in2006, grounded in sharp gains for regional oilexporters and South Africa (figure 1.15).Global commodity demand and prices werepushed higher in recent years, notably by con-tinued brisk economic expansion in China.Sub-Saharan Africa is one of few regions thathas witnessed a strong supply response tohigher oil prices, with crude oil production up14.3 percent in 2004, an additional 7.6 per-cent in 2005, and 8.1 percent in 2006 (whenNigeria, which suffered from a number ofshutdowns of facilities, is excluded).


32

Figure 1.14 South Asia growth is slowingas the Indian rupee appreciates

Jan.

2005

May

200

5

Sep. 2

005

Jan.

2006

May

200

6

Sep. 2

006

Jan.

2007

May

200

7

Sep. 2

007

18 US$-Rs

Industrial production15

12

9

6

3

Source: World Bank.

0

Industrial output (annual percentagechange; year-on-year)

90

94

92

96

98

110

100

102

104

108

106

US$-Rs index(January 2005 � 100)

Figure 1.15 Oil exporters drive 2007growth results for Sub-Saharan Africa


Oilexporters

Oil importers,excluding theRepublic ofSouth Africa

South Africa

9

0

3

6

GDP growth (percent change)

2007

2006

2005

Sub-SaharanAfrica


33

Improved macroeconomic stability is alsoplaying an important role in sustaininggrowth, as is a pickup in both domestic andforeign investment. Debt relief in recent yearshas freed budgetary resources for spending oninfrastructure and social programs. A com-mon trait across economies is a notablepickup in capital spending focused on thetransport, telecommunications, and construc-tion sectors. In addition, recovery fromdrought in many areas of the region is trans-lating into improved performance in agricul-ture, adding impetus to growth, while theincome effects stemming from several years ofhigh non-oil commodity prices are stimulatingprivate consumption.

Recent turbulence in international financialmarkets resulted in a moderate depreciation ofthe South African rand against the dollar, butthis followed a period of appreciation of therand because of anticipated capital inflows re-lated to merger and acquisition activity. Dueto weakness in the U.S. dollar, the nominal ef-fective exchange rate of the rand has returnedto levels prevailing in July, declining 11.2 per-cent during the first 10 months of the year.There are as yet no signs of a sharp sell-off ofSouth African assets, and there appears to belittle evidence of marked adverse effects on thedomestic growth outlook.

Much of the impetus for regional growthover 2008–09 will come from strong domesticdemand, notwithstanding softer privatedemand in South Africa, the region’s power-house, where higher interest rates and anerosion of real incomes are curbing real out-lays. A sharp decline in farmers’ income incountries affected by recent floods will consti-tute a near-term drag on growth, and onprivate consumption in particular, althoughgovernment and donor transfers and assistancemay mitigate some of the effects. Investment isexpected to remain strong, notwithstandingthe tightening of international credit condi-tions and lower commodity prices, in part be-cause of large strategic investments by rapidlygrowing developing economies such as Chinaand India. A notable activity is the Moatize

coal project in Mozambique, investment inwhich amounted to $1.44 billion in the firsthalf of 2007. Madagascar is also experiencinghuge investments in its economy. Against thisbackground, Sub-Saharan African GDP isanticipated to be sustained at a pace above6 percent through 2008, before slipping to5.8 percent growth in 2009 as oil exportersrespond to international conditions and re-strain output moderately.

World trade

The globalization of markets for goods andservices is continuing at an unabated pace.

Over the past seven years, world trade volumeshave increased at an average rate of 6.7 per-cent, virtually the same as during the 1990s(table 1.4). Trade volumes are expanding morethan twice as fast as industrial production(global GDP has grown 3 percent a year since2000, up from 2.8 percent a year duringthe 1990s). In current dollar terms, worldtrade doubled during the 1990s and has dou-bled again since 2000.

While world trade has grown steadily for thelast 15 years, developing country trade hasaccelerated in recent years. During the 1990s,developing country export volumes increasedat an annual pace of 6.8 percent, roughly thesame as the 6.9 percent export gains of the high-income countries. Since 2000, however, develop-ing countries’ exports have been growing twiceas fast as exports from high-income countries:10.8 percent a year versus 5.1 percent a year.

On the import side, the acceleration is evenmore impressive. During the 1990s, developingcountries’ import growth of 5.7 percent a yearlagged behind that in high-income countries(7.0 percent), but over the last seven years,those positions have reversed. In 2006, importgrowth in developing countries registered14.3 percent, compared with 7.9 percent in thehigh-income countries. Imports across devel-oping regions were growing at double-digitrates during 2006, as export revenues, whichhad been boosted by high-volume growth andsharp increases in commodity prices, were

being expended. The pickup of trade in devel-oping countries also shows in their shares ofworld markets: in current dollars, developingcountries’ market share increased gradually inthe 1990s from 20 to 25 percent, but since2000, their share has jumped to 35 percent,supported in part by higher commodity prices.

The rapid growth of trade and productionin developing countries, seemingly indepen-dent of growth in the high-income countries,is sometimes referred to as “delinking.” At thesame time, the rapid integration of developingcountries into global markets, a key factorunderlying high growth rates, could also be


34

Table 1.4 Developments and prospects for world trade and payments

Estimate Forecast

Growth in percent 1991–2000 2000–05 2006 2007 2008 2009

World trade volume a (growth in percent) 6.8 5.7 9.8 8.7 7.4 9.3

World exports (growth in percent) 6.9 5.7 10.1 9.2 7.6 9.2High-income countries 6.9 4.3 9.2 8.2 6.3 7.6

OECD high-income 6.8 3.6 8.8 8.7 6.5 7.7United States 7.1 1.9 8.4 7.8 8.5 9.0Japan 4.4 5.9 9.6 6.5 4.0 4.3Euro Area 6.9 3.6 9.0 11.9 7.2 8.6

Developing countries 6.8 10.4 12.7 12.0 11.0 13.1East Asia and the Pacific 11.7 15.4 17.7 17.8 15.2 18.5Europe and Central Asia 0.9 9.1 10.3 9.2 8.5 8.7Latin America and the Caribbean 8.1 5.3 7.8 4.8 5.5 5.8Middle East and North Africa 4.4 6.1 9.5 4.3 3.8 5.2South Asia 9.0 11.2 9.0 8.5 8.4 10.5Sub-Saharan Africa 4.7 5.4 4.4 5.7 7.3 6.6

World imports (growth in percent) 6.7 5.8 9.5 8.3 7.2 9.5High-income countries 7.0 4.5 7.9 6.8 5.4 7.9

OECD high-income 6.8 4.1 7.4 6.8 5.0 7.8United States 9.3 4.3 5.9 2.0 1.3 6.8Japan 3.6 3.9 4.6 3.6 3.0 5.2Euro Area 6.3 3.4 7.5 10.1 7.4 9.4

Developing countries 5.7 10.1 14.3 12.5 11.9 13.3East Asia and the Pacific 11.3 13.7 14.8 14.9 14.3 19.2Europe and Central Asia �0.9 10.5 14.2 12.8 11.6 10.4Latin America and the Caribbean 10.7 4.2 13.3 9.2 9.3 8.1Middle East and North Africa 1.6 9.7 19.1 12.0 10.0 7.4South Asia 7.9 12.9 12.1 12.1 10.5 11.9Sub-Saharan Africa 4.4 8.5 12.4 8.2 8.7 8.8

Current account (percent of GDP)High-income countries �0.1 �0.8 �1.1 �0.7 �0.6 �0.5

OECD high-income �0.2 �1.1 �1.9 �1.5 �1.4 �1.2United States �1.8 �4.9 �6.6 �6.0 �5.4 �5.1Japan 2.4 3.1 3.9 4.1 3.7 3.8Euro Area 0.0 0.2 �0.2 0.4 �0.1 0.1

Developing countries �1.5 1.4 3.6 3.1 2.5 1.8East Asia and the Pacific 0.1 3.3 8.4 10.1 8.6 7.6Europe and Central Asia �1.0 1.4 0.6 �1.3 �1.9 �2.6Latin America and the Caribbean �2.6 �0.6 1.6 0.5 0.1 �0.2Middle East and North Africa �0.3 5.5 9.6 8.2 9.5 6.2South Asia �1.6 0.3 �1.3 �2.4 �3.0 �2.7Sub-Saharan Africa �1.9 �0.3 0.9 0.4 0.9 �0.2

Source: World Bank.Note: a. Exports and imports of goods and non-factor services in 2000 US$.

considered as an intensification of the linkageof developing countries to high-incomecountries, as they have become an integralpart of the global business cycle. The combi-nation of delinking in terms of growth ratesand increased linkage in terms of cyclicalchanges in growth gives developing countriesa prominent role in the current economic en-vironment: they have become a driving forcefor global trade, and their strong trade linkscan help mitigate the slowdown in high-income countries. In 2007 and the yearsfollowing, developing countries are expectedto be the source of more than half of growthin global imports.

The shift of import demand away from theUnited States (and high-income countries gen-erally) and toward developing countries isclearly visible in recent high-frequency data.The slowing of high-income countries’ importsand of developing countries’ exports reflectsmuch weaker investment and consumptiongrowth in the United States in line with thefallout in the housing market (figure 1.16).

At the global level, the slowing of U.S.imports has been offset in part by a strength-ening of import demand across developingcountries driven, among other factors, by ro-

bust domestic demand, especially in the largeemerging market economies (figure 1.17).Looking forward, vibrant growth in develop-ing countries should offer a cushion fromrecessionary conditions in the key OECDeconomies over the critical 2008 period.

Indeed, among recent regional develop-ments, dollar-based import growth has beenstrong: Sub-Saharan African imports increased23 percent over the year to May 2007; SouthAsian imports jumped 32 percent through Au-gust, driven by strong investment and importdemand in India; and Latin America and EastAsia were importing at rates of 15 and 14 per-cent, respectively, through September. How-ever, the key contribution to the pickup in im-ports stems from Europe and Central Asia (anincrease of 28 percent through September, upfrom 20 percent a year earlier), where in Cen-tral and Eastern Europe, EU accession coun-tries are increasing their import potential indollars with currencies effectively pegged to anappreciating euro, while in the CIS, exportersof oil and other hydrocarbons continue tospend portions of their accumulated revenues.

Such rotation of trade growth is contribut-ing to an unwinding of global imbalances that

Figure 1.16 Weak U.S. growth reducesdemand for developing country exports

Source: World Bank.

1999

15

�15

10

5

0

�5

�10

U.S. imports

Growth of investment and imports, four quarter movingaverage, (percent)

U.S. investment

2000

2001

2002

2003

2004

2005

2006

2007


35

Figure 1.17 Export opportunities forhigh-income countries

Source: World Bank.

Note: M3 � March.

1994

M3

30

�10

20

10

0

Developing-country imports

(Percent) nominal growth in US$, 12-month movingaverage

High-income-country exports

1996

M3

1998

M3

2000

M3

2002

M3

2004

M3

2006

M3

started in 2006. The U.S. current accountdeficit declined from a peak of 6.8 percent ofGDP in the fourth quarter of 2005 to 5.5 per-cent by the second quarter of 2007. Theshrinking deficit reflected weaker U.S. imports,caused in part by the housing slump, andstronger exports, supported by demand in de-veloping countries and the falling dollar (fig-ure 1.18). The mirror image of the narrowingU.S. current account may be found in develop-ing countries, as well as selected OECD coun-tries, where surpluses are broadly in decline.Based on the strong correlations shown in fig-ure 1.16, export revenues in developing coun-tries are likely beginning to slow in line withthe softening of import demand in the high-income countries. At the same time, spendingin developing countries remains strong. Suchgradual unwinding of global imbalances isexpected to continue, at least through 2009, asoil price levels decline, driven by the sameelements that emerged in 2007.

Inflation and commoditymarkets

In high-income countries, inflationary pres-sures were sufficiently under control during

2007 to allow central banks to end a period ofmonetary tightening. In the United States, theFederal Reserve had become more concernedabout a possible recession than about acceler-ating inflation and lowered Federal funds by50 basis points to 4.75 percent on September18—the first reduction in four years. A further25 basis point cut was enacted on October 31and a like reduction on December 11, carryingFederal funds to 4.25 percent. In Japan, infla-tion is not yet permanently in positive territory,making it unlikely that the Bank of Japan willfollow its initial tightening steps with furtherincreases in interest rates. In Europe, inflationis fluctuating around the European CentralBank’s upper target, recently with more up-ward than downward momentum. Given ex-pectations for softening growth in 2007, aswell as an appreciating currency, the EuropeanCentral Bank will probably hold its policy rateat 4 percent.

The inflation picture is more diverse acrossdeveloping countries. In several rapidly grow-ing economies, where inflation had picked upover the last two years, policy interest rateshave been increased gradually. This occurredacross South Asia (India, Pakistan, and SriLanka), where signs of overheating becameevident, and also in several Latin Americanand Caribbean countries (Argentina, Chile,Colombia, and the República Bolivariana deVenezuela). China and the Czech Republic areexamples of economies where monetary tight-ening continues. But in some countries(Belarus, Brazil, the Lao People’s DemocraticRepublic, and the Philippines) inflation easedand monetary policy, which was tight, is nowloosening moderately. In several countriesmonetary policy has reversed. In Hungary andTurkey, for example, policy rates were cut in2007 after being on a long uptrend. This turn-about was a reaction to the recent slowdownin growth after a long period of strong gains


36

Figure 1.18 U.S. current account narrowsover 2007 and is likely to continuedoing so

U.S. current account components ($ billions, three quartermoving average)

Source: U.S. Department of Commerce.

�250

�200

�150

�100

�50

0

Q12000

50

Income Services Non-oil balance

TransfersOil balance

Q12001

Q12002

Q12003

Q12004

Q12005

Current account

Q12006

Q12007

in domestic demand that led to large currentaccount deficits and inflation pressures.

Inflation has remained remarkably mutedafter four years of global growth that has beenstronger than that experienced in the last30 years. Average consumer price inflation inhigh-income countries registered 1.8 percentin 2007, slightly below that experienced dur-ing the downturn of 2001. Median inflation indeveloping countries registered a moderate (ina historic context) 5.9 percent, only 0.6 per-centage points higher than at the beginning ofthe decade. Hyperinflation, once a trademarkof several developing countries, has been elim-inated, with Zimbabwe now being the notableexception. The number of developing coun-tries with double-digit inflation has halvedsince the beginning of the decade.6 Inflationwas lower in the first half of 2007 than it wasin the previous year in more than 60 percentof all countries, and among the 56 developingcountries for which inflation data are avail-able for the first half of 2007, more countriesexperienced an easing than a pickup of pres-sures. Moreover, average inflation is still onthe rise in just two of six developing regions(figures 1.19 and 1.20). In Sub-Saharan

Africa, inflation is volatile, but is trendinghigher, reflecting rising food prices, and theMiddle East and North Africa region shows aclear uptick in inflation, likely triggered by theexpenditure of burgeoning oil revenues in theregion, notably in the Islamic Republic ofIran. In other regions, inflation is volatile, butbroadly trending flat or slightly downward.

Developing countries have played animportant part in restraining inflationworldwide. Improved macroeconomic policieshave tended to reduce domestic pressures oninflation across a wide range of countries.Independent monetary policy, more cautiousfiscal stances, strong currencies, and rapidTFP growth have increasingly kept domesticinflation under control. In addition, greaterengagement in international trade and compe-tition has helped spread price restraint acrosstrading partners. In many countries, passingcost pressures through to output and to con-sumer prices has become progressively moredifficult.

The lack of pass-through, in particular, hasenabled monetary policy in the high-incomecountries to remain fairly passive in the face ofrising non-core prices. Partly as a conse-quence, global growth has been maintained at


37

Figure 1.19 Inflationary pressures arerising in the Middle East and NorthAfrica and Sub-Saharan Africa

Source: World Bank.

Jan.

200

2

Jul. 2

002

Jan.

200

3

Jul. 2

003

Jan.

200

4

Jul. 2

004

Jan.

200

5

Jul. 2

005

Jan.

200

6

Jul. 2

006

Jan.

200

7

6.0

4.5

3.0

1.5

0

7.5


Sub-Saharan Africa

9.0

Consumer prices (percent change, year-on-year)

Figure 1.20 Inflation is broadly stableelsewhere, though at high levels

Source: World Bank.

Jan.

200

2

Oct. 2

002

Jul. 2

003

Apr. 2

004

Jan.

200

5

Oct. 2

005

Jul. 2

006

Apr. 2

007

9.0

1.5

7.5

6.0

4.5

3.0

South Asia

Europe andCentral Asia

Latin America

Consumer prices (percent change, year-on-year)

East Asia

a higher rate much longer than in the past,generating increasing capacity constraints incommodity markets. As a result, commodityprices continued to increase and remained ele-vated into the fourth quarter of 2007. Com-modities have become more volatile, however,and the increase in several commodity prices isnow moderating (figure 1.21).

The increased use of food crops for pro-duction of biofuels is an important factor thatled to large increases in the prices of vegetableoils and grains in 2007, which in turn con-tributed to an overall 15 percent increase inthe index of agricultural prices and a 20 per-cent rise in food prices. The latter is of specialconcern for poor consumers in developingcountries.

The prices of metals have increased morethan other commodity prices over the last fouryears, largely because of especially strongdemand in China. Underinvestment duringearlier periods of low prices and numeroussupply problems and delays in bringing on newcapacity have also played a part. Shortages ofequipment and skilled workers have signifi-cantly increased development costs, and oregrades are deteriorating. The price of metalsfor which China is a net importer, especially

copper, have experienced sharp gains, whilethose for which China is a net exporter—mainly aluminum and to some extent zinc—have increased much less (figure 1.22). Follow-ing the global credit squeeze in August andSeptember, the prices of metals dropped morethan 10 percent (metals tend to be particularlysensitive to slowing economic activity andmay have been exposed to speculative pres-sures when investors closed their positions tofinance other losses in their portfolios). Theprices of metals are generally expected to peakin 2007, to decline by 5 percent in 2008, andto continue lower into 2009 as rising capacitytips markets into surplus.

Nominal oil prices, measured in dollars,broke historic records in November, reachingnearly $100 a barrel. Measured in euros, oilprices stood 4.5 percent above their 2006peak, while in real terms (corrected for overallinflation) oil prices remain 4.2 percent belowthe peaks reached in November 1979.

Higher oil prices have reduced growth inglobal oil demand, particularly in high-incomecountries. Oil demand in the OECD declinedfor six consecutive quarters beginning in thefourth quarter of 2005, with an average drop ofmore than0.4millionbarrels aday (figure1.23).


38

Figure 1.21 Commodity prices continuedgains through 2007 led by metals

Source: World Bank.

Jan.

1, 2

000

Jan.

1, 2

001

Jan.

1, 2

002

Jan.

1, 2

003

Jan.

1, 2

004

Jan.

1, 2

005

Jan.

1, 2

006

Jan.

1, 2

007

350

50

300

250

200

150

100

Metals

Crude oil

Agriculture

Commodity price indexes (January 2003 � 100)

Figure 1.22 Copper, zinc, and aluminumprices sharply affected by China

Sources: London Mercantile Exchange and World Bank.

Jan.

3, 2

000

Jan.

3, 2

001

Jan.

3, 2

002

Jan.

3, 2

003

Jan.

3, 2

004

Jan.

3, 2

005

Jan.

3, 2

006

Jan.

3, 2

007

9,000

0

7,000

8,000

5,000

6,000

3,000

4,000

2,000

1,000

Copper

Aluminum

Zinc

$/metric ton

In non-OECD economies, the demand for oilhas grown by just over 1 million barrels of oila day since 2005, down sharply from the surgeof 2004. Supply in several producers that arenot members of the Organization of PetroleumExporting Countries (OPEC), especiallyRussia and countries in western Africa, has in-creased in recent months, and among OECDcountries, Canadian production continues togrow, predominantly from oil sands, whilesignificant new output in the deep water U.S.part of the Gulf of Mexico is starting up. Theseincreases have been partially offset by moder-ately falling production in the North Sea. Asdemand eased and non-OPEC supply in-creased, OPEC countries reduced their outputto prevent further increases in stocks and a fallin prices (figure 1.24).

Because OPEC has limited spare capacityand is holding down production, oil priceswill likely remain quite elevated and volatile;however, high prices and increasing environ-mental concerns should continue to moderategrowth in demand. At the same time, rising

upstream investment in oil-producing coun-tries (both in OPEC and in non-OPEC coun-tries) should result in new capacity that ex-ceeds the growth in oil demand. Nevertheless,oil markets are expected to remain finely bal-anced over 2007–09, in part because of pro-duction discipline by exporters, and prices areexpected to remain above $75 a barrel for thecoming two years. In the longer term, the oilmarket balance is expected to loosen andprices are projected to fall toward $50 perbarrel.

The rise in agricultural prices during 2007was underpinned by strong demand for foodimports, especially by oil-exporting countries,which contributed to a 20 percent increase inglobal food prices for the year. Higher cocoaand robusta coffee prices raised beverageprices by 13 percent, while raw materialsprices were moderately higher (figure 1.25).The increase in food prices was led by fats andoils, up 50 percent for the year, and grains, up22 percent (figure 1.26).7 Among other com-modities, sugar prices declined 32 percent, as


39

Figure 1.23 Growth in the world’s demandfor oil slows

Sources: International Energy Agency and World Bank.

Q1, 2

001

Q1, 2

000

Q1, 2

002

Q1, 2

003

Q1, 2

004

Q1, 2

005

Q1, 2

006

Q1, 2

007

4.5

2.5

0

2.0

�1.0

3.0

3.5

4.0

1.5

0.5

1.0

Millions of barrels of oil/day

�0.5

World Demand

Other Other Asia China OECD

Figure 1.24 OPEC reduces output tosupport prices

Sources: International Energy Agency and World Bank.

5

3

0

�3

4

2

1

Millions of barrels of oil/day

OPEC Other Former Soviet Union

�1

�2

World supply

Q1, 2

001

Q1, 2

000

Q1, 2

002

Q1, 2

003

Q1, 2

004

Q1, 2

005

Q1, 2

006

Q1, 2

007

growing conditions in India, Pakistan, andThailand improved, and new plantings and fa-vorable weather boosted Brazilian production.

Food prices have risen nearly 75 percentsince their lows of 2000. The increases stempartly from the stepped-up use of food crops forbiofuels and partly from other more fundamen-tal factors, such as rapid income growth in de-veloping countries, high fertilizer prices, low

stocks, and droughts. Biofuels are playing anincreasingly important role in agriculturalcommodity markets as their share of globalproduction and trade increases. In 2006, bio-fuels accounted for 5–10 percent of the globalproduction of the primary biofuel feedstocksand up to 77 percent of the volume of trade.Among the largest biofuel producers, theUnited States used 20 percent of its maize pro-duction for biofuels; Brazil used 50 percent ofits sugarcane for biofuels; and the EU used68 percent of its vegetable oil production,primarily rapeseed, and also imported addi-tional vegetable oils. Such large usage reducessupplies of these crops for food and feed andhas contributed to substantial price gains(box 1.2).

The anticipated spike in grains prices, iden-tified as a concern in the World Bank’s GlobalDevelopment Finance report published inMay 2007 (World Bank 2007a), has largelymaterialized. Monthly wheat prices have in-creased 90 percent since mid-2007. Wheatstocks are expected to fall to record lows rela-tive to consumption, and prices may increasefurther in 2008 before production recoupsenough to rebuild stocks. In the meantime, alarge number of food-importing countriesmay suffer substantial terms-of-trade lossesover the course of 2007 and into 2008 (box1.3). Price increases for vegetable oils andgrains primarily affect low-income countries,with the rise in prices since the end of 2004leading to a terms-of-trade loss equivalent to0.5 percent of GDP. This represents 1 percentof GDP in 29 countries, and nearly 5 percentof GDP for the most affected country, Eritrea.The impact on middle- and high-income coun-tries is considerably less because imports ofthese commodities represent a smaller share oftrade, and higher prices on other commodityexports tends to offset terms-of-trade lossesresulting from higher food prices. Agriculturalprices are expected to remain nearly flat athigh levels in 2008, as biofuels productioncontinues to ramp up in response to consump-tion mandates and production subsidies,drawing resources from other crops.


40

Figure 1.26 A rise in food prices, led by aramp-up of the prices of fats, oils, andgrains

Source: World Bank.

Jan.

1, 2

000

Jan.

1, 2

001

Jan.

1, 2

002

Jan.

1, 2

003

Jan.

1, 2

004

Jan.

1, 2

005

Jan.

1, 2

006

Jan.

1, 2

007

200

50

175

125

150

75

100

Fats and oils

Other foods

Commodity price indexes (1990 = 100)

Grains

Figure 1.25 Agricultural prices surge over2006–07

Source: World Bank.

Jan.

1, 2

000

Jan.

1, 2

001

Jan.

1, 2

002

Jan.

1, 2

003

Jan.

1, 2

004

Jan.

1, 2

005

Jan.

1, 2

006

Jan.

1, 2

007

175

50

150

100

125

75

Raw materials

Food

Beverages

Commodity price indexes (2003 = 100)


41

Biofuels are not new: the first Ford automobilewas originally intended to run on ethanol. How-

ever, the interest in biofuels has increased in recentyears along with increases in energy prices andheightened concern about the environment. Coun-tries want not only to achieve greater energy security,but also to reduce the use of fossil fuels to lowergreenhouse gas emissions and improve air quality.The demand for biofuels also got a boost whenmethyl tertiary-butyl ether was banned as a fueladditive to meet clean air regulations in many U.S.states and localities and for fear it was contaminat-ing groundwater. The political gain attached to sup-porting new demand for agricultural products, andthus enhanced crop prices, has added to interest onthe part of politicians and encouraged generous sup-port policies in many countries, including EU coun-tries and the United States.

Biofuels can be produced from a variety offeedstocks. Current technology, often called first-generation technology, relies primarily on food cropssuch as sugarcane and maize to produce ethanol andon vegetable oils from rapeseed, soybeans, palms,and other crops to produce biodiesel fuel. So-calledsecond-generation technology may be able to pro-duce biofuels from an even wider range of feed-stocks, such as switch grass, timber waste, andmunicipal garbage, but such technology is not yetcommercially viable and many experts do not expectit to become so for at least a decade.

Global production of biofuels totaled about45 billion liters in 2006, representing slightly morethan 1 percent of global road transport fuels on anenergy equivalent basis. Biofuels can be used to re-place their fossil fuel counterparts or can be blendedwith fossil fuels to achieve certain benefits, such asreduced tailpipe emissions and increased octane

Box 1.2 Biofuelslevels to improve engine performance. One of theirgreatest advantages is that they can be used in con-ventional gasoline and diesel engines without modifi-cation of the engines and can be dispensed throughexisting distribution channels.

Nevertheless, their use has some limitations.Ethanol can be used in conventional gasoline enginesonly up to about a 10 percent blend with gasolinewithout engine or fuel system modifications. It alsorequires special handling in transport to prevent con-tamination. Specially designed flex-fuel engines canuse a wider range of blends of ethanol and gasolineand are available in Brazil and the United States.Biodiesel fuel can be used in any blend with fossilfuel diesel in standard diesel engines, but its use islimited in colder climates. The energy content ofethanol is lower than that of gasoline, providingabout 20–30 percent fewer miles per gallon thangasoline, while biodiesel provides 5–10 percentlower mileage than diesel.

While biofuels have thus far had little impact oncrude oil prices, they have already had large effectson prices of commodities used as feedstocks for bio-fuels, as well as for competing crops. For example,maize prices rose by about 60 percent from mid-2005 to mid-2006, largely because of the increaseduse of maize for ethanol production in the UnitedStates. This prompted a huge shift of land fromwheat into maize in the following season, whichcontributed to a sharp increase in wheat prices.Vegetable oil prices have also increased because oftheir stepped-up use for biodiesel production inEurope and the United States, with palm oil pricesup 48 percent in the last year and soybean oil pricesup 25 percent. These price shifts have set off a foodversus fuel debate that is causing some to questionthe contribution of biofuels.

downward spiral in housing is mitigated bystrong export growth. On the financial side ofthe economy, the projections assume thatlosses on holdings of asset-backed securitiesare widely distributed and that interventionsby the Federal Reserve, the European CentralBank, and other institutions restore calm tofinancial markets. However, the effective cost

Risks and uncertainties: Danger of a banking crisis anda U.S. recession

The baseline projections assume, on the realside of the economy, that the U.S. housing

recession does not spill over in a large-scaleway to the rest of domestic demand, as the

of capital is likely to increase further, reflectedin tightened credit criteria for firms as well ashouseholds in the United States. Elsewhere,however, tightening is expected to be moremoderate. Under such conditions, weakness inU.S. housing would continue, but the contrac-tion in residential investment will have bot-tomed out by mid-2008.

Should unexpected and large-scale newlosses occur in financial markets—concentrated among commercial and invest-ment banks or among major investors—creditconditions globally could tighten much more.Such a scenario would tend to increase losseson asset-backed securities, potentially carryingfinancial markets and the real economy into a


42

Higher prices of imported staples, strong growthof domestic and regional demand that pushes

up prices of domestically produced goods, and in-creased prices of production inputs such as fertilizersand energy are causing rapid rises in food prices inmany countries. The price increases will hurt thepoor, who spend a large share of their income onfood, but will also help the rural poor who producea marketable surplus. Governments are under pres-sure to take action to blunt the impact of higherfood prices, but many countries have liberalized orpartially liberalized their domestic food markets andimports (Bangladesh, Brazil, Egypt, India, Mali,Morocco, Russia, Ukraine, and the Republic ofYemen, to name a few) and no longer have policyinstruments to control food prices. Those countriesthat have targeted safety net programs can rely onthose channels to provide assistance to the poorestpeople, but countries without safety net programswill feel pressure to impose price controls or to rein-troduce government controls. This would undo suc-cessful policy reforms and send a negative message tothe private sector.

Bangladesh offers an example of the success ofopen market food policies. It has transformed itsagricultural sector into one of the most productive inSouth Asia. The country is largely self-sufficient inrice, a basic staple, and is an emerging exporter ofhigh-value agricultural products. One of the keys tothis success was the government’s decision to liberal-ize food imports in the early 1990s. Private tradersimported food grains during times of domestic short-fall, providing needed supplies and price stabiliza-tion, as well as removing a financial burden from thegovernment. By 2000, the private sector was import-ing 100 percent of imported food, and the govern-

Box 1.3 Policy responses to rising food pricesment reoriented its large public food distributionsystem away from mass distribution in favor of atargeted safety net program for the poor. Such aresponse would be effective in the current situationof high food prices, but a complicating factor is thatpart of the current price increases might be morepersistent than in the past.

Past periods of food price increases were tempo-rary and lasted only two or three years, such as theincreases during the 1970s or the more recent in-creases in 1995–96; however, the current increaseshave a structural component that may persist be-cause it is closely tied to the rise in global energyprices. If energy prices remain high, food crop pricesare unlikely to decline significantly. Over the longerterm, supplies of food are expected to increase andprices to fall, but the current price increases are ex-pected to continue for several years, and thus mostcountries will not be able to shelter their consumersfrom them. Current food price increases also have atemporary component caused by low stocks and pro-duction shortfalls stemming from drought. These canbe expected to dissipate as supplies respond to highprices. Countries should aim to protect consumersfrom temporary price increases caused by shortages,but few countries can afford to protect consumersagainst structural changes in food prices.

Consumers in food-exporting countries are also see-ing their food prices increase as supplies are exportedat high international prices, and several countries haveimposed export bans to contain domestic food priceinflation. Such bans unfairly penalize the producers ofthese crops and may encourage smuggling and corrup-tion. A more appropriate policy response is to providea targeted safety net program for the poor while allow-ing exports to continue unfettered.

downward spiral and requiring an aggressiveloosening of monetary policy.

Under such circumstances, contraction inquarterly GDP in the United States would be-come more likely, pushing growth in 2008 to1 percent, or almost half of baseline growth.Equity markets in high-income countrieswould likely decline substantially, and the ef-fective cost of capital could increase by some200 basis points in 2008, compared with thebaseline. Such a pronounced credit crunchwould be reflected in a sharp decline in U.S.business investment, declining employment,weaker consumer outlays, and a prolongedperiod of depressed consumer prices.

Adverse developments in the United Stateswould spill over to the rest of the worldthrough weaker U.S. imports and a substan-tial further decline in the dollar, spurred byan aggressive loosening of monetary policy.The spillover could be exacerbated by a re-versal of the yen carry trade and a reducedappetite for U.S. assets among internationalinvestors as growth slows and assets of fi-nancial institutions become more risky.Largely because of reduced exports and in-vestment growth, GDP growth in Europe andJapan could fall to 1.5 and 1.3 percent,respectively, about half a percentage pointbelow the baseline.

For a number of middle-income countries,the most important transmission channel ofeffects stemming from the OECD countrieswould be a tightening in international creditconditions, which would cut into investmentand reduce growth. Middle-income economieswith large current account deficits and coun-tries whose currencies are pegged to the eurowould likely feel the greatest effects. Growthin middle-income countries would fall a per-centage point below the baseline; however, theimpact would be quite diversified, in partdepending on how economies are linkedfinancially to the U.S. dollar. Central Euro-pean economies that accumulated euro-denominated debt will be more vulnerablethan countries that hold dollar debt.


43

The repercussions for low-income coun-tries could also be sizable, as weaker globaldemand for commodities (metals, agriculturalproducts, and fuels) could worsen their termsof trade by 2 percent of GDP, and the pace ofexpansion could decline by 0.6 points in 2008.Overall, given such a scenario, global growthmight decline by three-quarters of a percent-age point in 2008 compared with the baseline.

The loosening of monetary policy in re-sponse to the subprime crisis might alsocause growth to overshoot. As a result, com-modity markets could tighten further; infla-tionary pressures would mount, especially indeveloping countries; and financial imbal-ances would increase rather than recede.Such a scenario could sow the seeds of amuch sharper growth slowdown in themedium term and illustrates the challengefacing monetary authorities in both high-income and developing countries.

Long-term prospects andpoverty forecastsA potential for catching upBeginning in the mid-1990s, per capita incomegrowth in developing countries has acceler-ated, with growth being particularly vibrantsince 2000. Several factors suggest that thishigh growth will be sustained over the longerterm. First, economic policies are on a moresolid footing than in earlier periods, with bothinflation and fiscal deficits broadly under con-trol, and structural policies more conducive totaking advantage of more open global mar-kets, and a business climate more favorablefor investment. Second, many countries haveentered, or are entering, a period of demo-graphic transition that combines rapid laborforce growth with declining dependency ratios.This shift provides a window of opportunityfor rapid economic gains. Third, the incomedisparity between developing and developedcountries is still large, but with broader accessto information and technology-laden capital

and imports, developing countries have theability—and the incentive—to narrow this gap.

The long-term forecast for 2010–30 pro-jects sustained growth across developingcountries, albeit with a moderating trend (fig-ure 1.27). For developing countries as a whole,per capita growth is expected to ease from anaverage of 3.9 percent in the first decade ofthe 2000s to 4.5 and 3.4 percent in the secondand third decades, mainly reflecting slowinggrowth in the East Asia and the Pacific region.Sub-Saharan Africa could see a slight accelera-tion, with average per capita income growth of3.0 percent in the current decade, increasing toa more sustainable 3.2 to 3.4 percent in subse-quent decades. This is a substantial improve-ment over the 1980s and 1990s, when percapita incomes in Sub-Saharan Africa declined.

Years of sustained increases in per capitaincomes should see average real incomes(stated in 2001 prices) more than double by2020, rising from $1,300 in 2001 to $2,800by 2020. By 2030, they are projected to reachnearly $4,000. However, significant variationwill occur around these numbers, and despite

a relatively optimistic growth scenario, thegap between rich and poor countries will re-main extremely wide. The average citizen in adeveloping country is projected to earn only7.8 percent as much as a citizen of a high-income country, a ratio that should rise toabout 10.0 percent by 2030.

This year’s long-term scenario represents asignificant upward revision from last year’slong-term forecast. The revision reflects a con-fluence of factors, including the simple recog-nition that developing country growth hasaccelerated over the last decade and has beenbroadly based and sustained. Among thefactors that likely contributed to sustainingthis high level of growth is the relatively rapidconvergence in technological achievementbetween high-income countries and develop-ing countries (see chapter 2). This progress isprojected to continue over the next 20 years,and the returns to knowledge and capital willremain more or less constant. These sustainedreturns combined with rising incomes willtranslate into increased investment in educa-tion and research and development, helping to


44

High-in

com

e

coun

tries

Low- a

nd m

iddle-

incom

e co

untri

es

East A

sia a

nd

the

Pacific In

dia

China

Europ

e an

d

Centra

l Asia

Latin

Am

erica

and

the

Caribb

ean

Midd

le Eas

t and

North

Afri

ca

South

Asia

Sub-S

ahar

an A

frica

Figure 1.27 Long-term growth, 1980–2030

Source: World Bank data and simulations with the Linkage model.

�2

0

8

10

6

1981–90

1991–2000 2011–20

2021–30

2001–10

4

2

Annual per capita GDP growth (percent)

establish a virtuous cycle of further technolog-ical progress and increased incomes. Thosecountries where technological progress hasstagnated are expected to benefit from a sup-portive global environment characterized byaccelerating exports and continued strongincome gains from commodities, partly ex-plained by the growing importance of devel-oping countries in global growth.

Part of the strong projected performance fordeveloping countries derives from strongerlabor force growth, but much can be attrib-uted to technological progress, measured in fig-ures 1.28 and 1.29 by TFP growth. The strongTFP growth in developing countries of the lastseveral years is consistent with the findings inchapter 2 of this report, which suggest that(based on a different measure of technology)the technology gap between middle- and high-income countries has narrowed over the last10 years. Over the forecast horizon, the rela-tive strength of TFP growth in developingcountries is expected to persist, which is alsoconsistent with the finding in chapter 3 that the


45

Figure 1.28 Declining capital-led growthfor developed countries, 2002–30

Source: Simulation results from the Linkage model.

Note: TFP � total factor productivity.

2007

2002

2012

2017

2022

2027

3.5

2.5

1.0

�1.0

3.0

2.0

1.5

Growth decomposition (percentage contribution per year)

0.5

0

�0.5

Capital Labor TFP

Figure 1.29 Sustained high productivitygrowth for developing countries

Source: Simulation results from the Linkage model.

Note: TFP � total factor productivity.

8.0

7.0

4.0

0

6.0

5.0

Growth decomposition (percentage contribution per year)

3.0

2.0

1.0

Capital Labor TFP

2007

2002

2012

2017

2022

2027

drivers of technological diffusion and absorp-tion in developing countries have strengthened.However, as stressed there, the realization ofthis potential during the coming decades willdepend on the extent to which countries, espe-cially low-income countries, can strengthentheir technological absorptive capacity andremain open to new technology flows.

Poverty declines significantly in thebaseline, though not uniformlyThe upward revisions to the long-term fore-cast generate a more positive poverty forecastfor 2015, albeit a modest improvement com-pared to the forecast presented in the GlobalMonitoring Report for 2007 (World Bank2007b). The percentage of the population indeveloping countries living on less than $1 aday in 2015 under the current long-term fore-cast is 10.2 percent, down from 11.8 percentin the Global Monitoring Report (table 1.5).The rapid improvements in reducing povertyin Asia since 1990 imply that the target ofhalving the percentage of the poor living on


46

Table 1.5 Poverty in developing countries by region, selected years

Region or country 1990 2004 2015

Number of people living on less than $1/day (millions)East Asia and the Pacific 476 169 40

China 374 128 29Rest of East Asia and the Pacific 102 41 11

South Asia 479 446 256India 376 371 217Rest of South Asia 103 76 39

Europe and Central Asia 2 4 2Middle East and North Africa 5 4 2Sub-Saharan Africa 240 298 290Latin America and the Caribbean 45 47 34

Total 1,247 970 624Excluding China 873 841 595

Number of people living on less than $2/day (millions)East Asia and the Pacific 1,113 684 296

China 819 452 186Rest of East Asia and the Pacific 294 232 110

South Asia 954 1,116 997India 734 868 772Rest of South Asia 220 248 226

Europe and Central Asia 20 46 16Middle East and North Africa 49 59 38Sub-Saharan Africa 396 522 567Latin America and the Caribbean 115 121 102

Total 2,647 2,548 2,017Excluding China 1,828 2,096 1,831

Percentage of the population living on less than $1/day

East Asia and the Pacific 29.8 9.1 2.0China 33.0 9.9 2.1Rest of East Asia and the Pacific 22.1 7.1 1.6

South Asia 43.0 30.8 15.1India 44.3 34.3 17.6Rest of South Asia 38.9 20.6 8.5

Europe and Central Asia 0.5 0.9 0.3Middle East and North Africa 2.3 1.5 0.7Sub-Saharan Africa 46.7 41.1 31.4Latin America and the Caribbean 10.2 8.6 5.5

Total 28.7 18.1 10.2Excluding China 27.1 20.7 12.6

Percentage of the population living on less than $2/day

East Asia and the Pacific 69.7 36.6 14.5China 72.2 34.9 13.4Rest of East Asia and the Pacific 63.7 40.4 16.9

South Asia 85.7 77.1 59.0India 86.4 80.4 62.7Rest of South Asia 83.4 67.6 49.2

Europe and Central Asia 4.3 9.8 3.4Middle East and North Africa 21.7 19.7 10.3Sub-Saharan Africa 77.1 72.0 61.5Latin America and the Caribbean 26.3 22.2 16.3

Total 60.8 47.6 32.9Excluding China 56.8 51.6 38.7

Source: World Bank.


47

$1 a day or less between 1990 and 2015 willbe achieved at the global level, though not nec-essarily at the country or regional level. Thenew forecast leads to some improvement inthe outlook for Sub-Saharan Africa, but theregion is still significantly off target.8,9

Agricultural productivity has importantpoverty implications for low-incomecountriesAgricultural technology is a particularly im-portant determinant of overall technology andpoverty reduction in low-income countries(World Bank 2007c), mainly because in mosteconomies the majority of workers remain inagriculture, and the poor are concentratedin rural areas. Moreover, productivity growthin agriculture is one of the main drivers ofrising incomes in agricultural economies.Increased agricultural productivity frees upworkers to take on more lucrative manufac-turing jobs and reduces food costs relative towages. Moreover, by increasing yields, highagricultural productivity generates a mar-ketable surplus that can be used to purchasehigher-quality inputs for production andconsumer goods, which in turns leads to a vir-tuous cycle between the agricultural andnonagricultural sectors of the economy. Themarketable surplus can also be used to in-crease exports or to reduce food imports.

In the baseline scenario, agricultural pro-ductivity is projected to increase by a uniform2.5 percent a year (Martin and Mitra 1999).Historically, however, not all countries haveachieved this average increase. For example,the green revolution that lifted agriculturalproductivity in South Asia over the last40 years largely bypassed Sub-Saharan Africa,where yields have largely stagnated. Lookingahead, agriculture in Sub-Saharan Africa isfaced with some of the same challenges as inthe past, for example, lack of access to creditand poorly integrated markets. Both SouthAsia and Sub-Saharan Africa could be facingadditional challenges that will threaten futuregains in agricultural productivity, for instance,

environmental degradation caused by globalclimate change or insufficient investment innew and locale-specific varieties of crops.

To illustrate the sensitivity of future out-comes to the possibility of weaker agriculturalproductivity performance, an alternative sce-nario looks at the impact on global growth,incomes, and poverty from assuming zeroagricultural productivity growth over the pe-riod 2008–15 in the two largely low-incomeregions of South Asia and Sub-SaharanAfrica.10 One key result would be higherprices for agricultural produce compared withthe baseline. Agricultural producer prices inSub-Saharan Africa would increase around6 percent, with a more modest increase of2 percent for processed food. The increase inprices is more acute in South Asia: around11 percent for primary agriculture, with a4 percent increase in processed foods. Theseincreases lead to a loss of competitiveness inboth domestic and export markets.

In the case of Sub-Saharan Africa, importsof crops and livestock products rise by some40 percent in 2015 relative to the baseline andexports drop by 30 percent. Overall outputdeclines by some 12 percent. Agriculturallabor demand barely changes, as the decline inoutput is offset by a decline in productivity,leaving overall labor demand more or less un-changed, albeit with depressed wages. InSouth Asia, labor demand actually increases.The output impact is a more modest drop of9 percent, and the loss of productivity is re-flected in an overall increase in labor demandof 1 percent, and thus less rural-to-urbanmigration occurs. The difference between thetwo regional impacts is linked to the degree ofautonomy of their respective agricultural andfood markets. A greater share of Sub-SaharanAfrica’s agricultural output is exported and agreater share of its demand is imported. Theincrease in domestic producer prices willtherefore be dampened by external markets inSub-Saharan Africa, compared with themore self-sufficient markets in South Asia.11

South Asia therefore witnesses more price

adjustment and less volume adjustment rela-tive to Sub-Saharan Africa.

Lower agricultural productivity, higherprices, and lower wages for unskilled workersincrease poverty in both South Asia and Sub-Saharan Africa. In Sub-Saharan Africa, overallconsumer prices rise by 1.4 percent and foodprices by 4 percent. In South Asia, the in-creases are even more marked: 10 and 3 per-cent, respectively. Combined with a significantfall in wages for the unskilled, which creates asharp drop in food wages (the quantity offood that can be purchased with the averagewage) that affects the poorest the most, thepoverty headcount index in Sub-SaharanAfrica would increase by some 5 percentagepoints in 2015 relative to the baseline, eventhough the average income loss would be amore moderate 3 percent, suggesting that theloss in agricultural productivity harms thepoor, on average, more than others.12

Notes1. Carry trade is an approach undertaken to lever-

age investments in higher-yielding securities intermedi-ated through a low-interest cost center. For example, apurchase of emerging market equities or fixed incomesecurities through borrowing in yen at low Japanese in-terest rates and converting yen to local currencies tocomplete the transaction would be classified as a carrytrade. Even though estimates of funds intermediatedthrough the yen carry trade are highly uncertain, an in-dicator of their potential size can be discerned from thesubstantial weakness of the yen-dollar exchange rateduring 2006–7, when it fell some 5.8 percent.

2. A full analysis of recent developments and theoutlook for each region is available on the WorldBank’s Web site at http://www.worldbank.org/prospects.

3. China is the largest overseas holder of U.S.mortgage-backed securities, around $260 billion,largely through China’s official reserve holdings andholdings of Chinese commercial banks. However,almost all of these instruments enjoy the backing ofU.S. government-sponsored enterprises Fannie Maeand Freddie Mac, and the risks associated with theseholdings appear to be minimal.

4. Developing countries of the Middle East andNorth Africa region, which account for regional

aggregate figures in this chapter, are Algeria, the ArabRepublic of Egypt, the Islamic Republic of Iran, Jor-dan, Lebanon, Morocco, Oman, the Syrian Arab Re-public, Tunisia, and the Republic of Yemen. Low- andmiddle-income economies with insufficient data forcoverage are Iraq, Libya, and the West Bank and Gaza.High-income countries of the region, which are ex-cluded from aggregates in the chapter, are Bahrain,Kuwait, Qatar, Saudi Arabia, and the United ArabEmirates.

5. Changes in the volume of oil and gas productionhave been modest in recent years, ranging from 0.5 to1.0 percent annual gains in Algeria to a decline in theIslamic Republic of Iran and the Republic of Yemen.Hence the buildup in export revenues is largely due tothe large-scale increases in global oil prices during2005 through 2007.

6. Inflation now exceeds 10 percent in Angola,Argentina, Botswana, Costa Rica, Ghana, Haiti,Indonesia, the Islamic Republic of Iran, Kenya,Madagascar, Malawi, Mozambique, Sri Lanka,República Boliviarana de Venezuela, and Zimbabwe.

7. Among individual vegetable oils, palm oil wasup 48 percent, coconut oil was up 41 percent, and soy-bean oil was up 25 percent. Among individual grains,maize prices rose 33 percent and wheat prices in-creased 30 percent.

8. Even though the revisions to the long-term fore-cast imply roughly a doubling of per capita growth forSub-Saharan Africa, that growth will have more im-pact after 2015 than before, as recent poverty forecastshave already incorporated the strong upward trend inper capita growth rates since the end of the 1990s.

9. The poverty numbers presented here do not yettake into account the results of the recent InternationalComparison Project, which will provide an updated setof price levels across countries. New purchasing powerparity exchange rates could—although not necessarilywill—lead to a new set of poverty estimates.

10. These projections involved six modeled coun-tries or regions—Bangladesh, India, Pakistan, the restof South Asia, Nigeria, and the rest of Sub-SaharanAfrica excluding South Africa.

11. One plausible explanation for South Asia’sgreater self-sufficiency has been its higher agriculturalproductivity over time because of the green revolution.

12. The poverty impact is generated by the WorldBank’s Development Economics Prospects Group’sglobal income distribution dynamics poverty tool, andwas performed only for Sub-Saharan Africa because ofits high projected level of poverty in 2015. The globalincome distribution dynamics tool probably underesti-mates the true poverty impact, because information onconsumption in most surveys is insufficient to factor in


48

price changes fully. To the extent that the poor spenda larger share of their meager budget on food, onewould anticipate an even greater negative impact forthe poor.

ReferencesMartin, Will, and Devashish Mitra. 1999. “Produc-

tivity Growth and Convergence in Agriculture

and manufacturing.” Policy Research WorkingPaper Series 2171. World Bank, Washington, DC.

World Bank. 2007a. Global Development Finance2007. Washington, DC: World Bank.

_____. 2007b. Global Monitoring Report 2007. Wash-ington, DC: World Bank.

_____. 2007c. World Development Report 2008:Agriculture for Development. Washington, DC:World Bank.


49

51

Technology and TechnologicalDiffusion in Developing Countries

2

Technological progress—improvements in thetechniques (including firm organization) bywhich goods and services are produced, mar-keted, and brought to market—is at the heartof human progress and development. At thenational level, technological progress canoccur through invention and innovation;through the adoption and adaptation of pre-existing but new-to-the-market technologies;and through the spread of technologies acrossfirms, individuals, and the public sectorwithin a country.

For developing countries, the bulk of tech-nological progress occurs through the lattertwo channels. Much of this chapter is con-cerned with measuring the extent to whichthis process has occurred in countries in dif-ferent regions and at different income levels.Although the current state of technologicalachievement is itself illuminating, the pace atwhich it is changing is equally important, andthis is estimated by comparing the level oftechnological achievement in the early 1990swith its current level and inferring the pace ofchange for different countries. The chapterconcludes by looking at the speed with whichspecific technologies spread both across coun-tries and within them. Armed with this broadview of technological progress within devel-oping countries, chapter 3 explores in moredetail the main factors that influence techno-logical progress in individual developingcountries. That chapter places equal emphasis

on the international connections and net-works that expose firms and individuals indeveloping countries to cutting-edge tech-nologies and on the domestic factors that de-termine how successfully countries are able toabsorb and apply those technologies. To es-tablish a baseline for future work, both chap-ters adopt a positive (empirical) rather than anormative or prescriptive approach. Never-theless, some clear conclusions with policyimplications do emerge from the analysis.

This chapter begins its empirical examina-tion by reviewing existing estimates of thecontribution of technological progress to eco-nomic growth as measured by gross domesticproduct (GDP) and of income levels in devel-oping countries. It reviews the mechanisms bywhich technology contributes to GDP and in-comes, but also stresses the contributions oftechnology to other important developmentgoals that are not well captured by GDPalone, such as health, education, and the envi-ronment. In addition, it discusses some of theprincipal limitations of this kind of empiricalanalysis and accompanying caveats. The chap-ter then discusses a wide range of previouslypublished indicators of the extent to whichvarious technologies have penetrated theeconomies of developing countries. For ease ofexposition, these indicators are arranged inthree groups: those showing the extent of sci-entific innovation and invention; those mea-suring the penetration of older technologies,

such as railroads and telephones; and thosemeasuring the penetration of newer technolo-gies, such as personal computers and mobilephones. The chapter then develops an aggre-gate measure of technological achievementusing a statistical technique (principal compo-nents analysis) that combines some 20 sepa-rate indicators of technological achievementalong these three dimensions, plus anadditional dimension, the extent to whichcountries are exposed to external technologies(explored in more detail in chapter 3). Thedistribution of overall technological achieve-ment across countries and changes over thepast decade are examined to evaluate both thespeed with which technological achievementin countries is advancing and the dimensionsalong which change is occurring most quickly.The chapter concludes by examining a newlongitudinal data set (Comin and Hobijn2004) that tracks both the speed with whichindividual technologies are transmitted acrosscountries and the pace with which they diffusewithin countries.

Eight main results emerge from thischapter:

1. While technological achievement is re-lated to income levels across countries, thenature of this relationship differs depend-ing upon the dimension of technology beingexamined.

• While a strong correlation exists be-tween scientific innovation and inven-tion and income in high-income coun-tries, almost none of this kind of activityis being performed in developing coun-tries. As a result, virtually all technolog-ical progress in developing countriescomes from the adoption and adapta-tion of preexisting technologies.

• At an aggregate level, the use of oldertechnologies is positively related to in-come in developing countries. However,the extent to which they are employedvaries substantially within incomegroups, suggesting that developing

countries’ history, geography, and pastgovernment success in deliveringinfrastructural technologies are equallyimportant determinants of the extent towhich older technologies are used.

• Penetration rates of more recent tech-nologies vary more regularly with in-come. In part this reflects their relativelylower start-up and infrastructure coststhan those of older technologies andtheir more flexible delivery structure.

2. While technological achievement tends torise with income, it tends to level off. The levelof technological achievement at which this lev-eling off occurs differs across countries accord-ing to their geography, history, and level oftechnological absorptive capacity (discussed inmore detail in chapter 3). Thus, reflecting anemphasis on equal access to education andstate-provided technological services, countrieswithin the Europe and Central Asia regionhave significantly higher levels of technologicalachievement than other countries at similar in-come levels. By the same token, countries inLatin America and the Caribbean are some-what less advanced than might be expected, asearlier inward-looking policies and weak basictechnological literacy in the overall populationhave limited the extent to which technologieshave permeated economic activity.

3. Technological achievement within coun-tries can vary widely. Despite a level of tech-nological achievement in major cities that canrival that in high-income countries, low levelsof technological advancement in rural areasmean that, viewed as a whole, countries suchas China and India are not particularly tech-nologically advanced. Moreover, because tech-nology spreads slowly across firms, there arewide differences in the technological sophisti-cation of production, even within the samesector in the same country.

4. Overall, the technology gap betweenmiddle-income and high-income countries hasnarrowed over the past 10 years. Evidence ofcatch-up is particularly strong in Chile,Hungary, and Poland, where the overall level of


52

technological achievement increased by morethan 125 percent during the 1990s.

5. On average, technology is advancing morerapidly among low-income countries. Amongthose low-income countries for which sufficientdata are available, the penetration of technol-ogy is progressing more rapidly than in eithermiddle- or high-income countries. However,this reflects very strong catch-up in some coun-tries and more modest improvements, or evenrelative declines, in the majority. Moreover,technology in high-income countries is also ad-vancing, and the absolute increase in thesecountries is larger than in developing countries.

6. The pace at which technology spreadsbetween countries is accelerating. Whereas anew technology in the 1800s could take aslong as 100 years to reach 80 percent of theworld’s countries, for a new technology toreach 80 percent of the world’s countries nowtakes less than 20 years.

7. Ultimately, however, what matters mostfor technological achievement is the speedwith which technology spreads within a coun-try. Here too the evidence suggests a pickup inthe pace of internal diffusion, but there is alsowidespread divergence across countries, evenacross those at similar income levels.

8. Changes in the regulatory environmentand in the nature of technologies partly ex-plain the acceleration in the rate at which theypenetrate into developing countries. Many oldinfrastructure technologies, such as roads,railroads, sanitation, and fixed-line telephonesystems, are often provided by the governmentand are thus subject to public sector budgetconstraints and the risk of government failure.By contrast, the most common new technolo-gies, such as the Internet, mobile phones, andcomputers, are being delivered in a regulatoryenvironment that encourages competition andthat harnesses private capital (domestic andforeign) to provide basic infrastructure. More-over, the past 10 years have been more stablepolitically than the 1980s and 1990s, whichhas likely given a boost to the diffusion ofnewer technologies.

The role of technology in development

Technological progress is at the heart ofhuman progress and development. As the

1998 World Development Report on theknowledge economy (World Bank 1998) em-phasized, the understanding of how things arecreated and the communication of that know-ledge are critical drivers of economic progress.Central to understanding the role of technol-ogy is the recognition that technology andtechnological progress are relevant to a widerange of economic activities, not just manufac-turing and computers. For example, some esti-mates suggest that technological progress hasboosted productivity in agriculture four timesas quickly as in manufacturing (Martin andMitra 2001). Indeed, seemingly low-tech prod-ucts such as corn or flowers can be the resultof relatively high-tech production processes,while in some countries the production ofostensibly high-tech products such as comput-ers is an outcome of relatively low-tech assem-bly activities. Finally, in many cases technologyis embodied in production and managementsystems rather than in physical goods or soft-ware algorithms. A computer loaded with thelatest software that sits unused on a desk formost of the day is a very different manifesta-tion of technology than the same computerthat is running a production process or man-aging an accounts payable system.

This report defines technology and techno-logical progress in this wider sense, althoughdata limitations may give some of the measuresdeveloped the flavor of a more narrow, physi-cal, and manufacturing-oriented definition.

Technology is both a critical determinantand an outcome of rising incomesTraditionally, economists view the process bywhich goods and services are produced as onethat combines capital, labor, and other factorsof production (land and natural resources)using a particular technology. The relative ef-ficiency with which a given economy produces

T E C H N O L O G Y A N D T E C H N O L O G I C A L D I F F U S I O N I N D E V E L O P I N G C O U N T R I E S

53

goods and services given a certain quantity oflabor and capital is called total factor produc-tivity (TFP). TFP is commonly interpreted as ameasure of the technology of production andits rate of growth as a measure of technicalprogress.1

International comparisons of TFP suggestthat enormous gaps exist between high-income and low- and middle-income countriesin the efficiency with which they producegoods and services (table 2.1). In 2005, the av-erage level of TFP in low-income countrieswas only slightly more than 5 percent ofU.S. levels. The technology lower-middle-income countries employed was roughly twiceas efficient and that of upper-middle-incomecountries was approximately four times as effi-cient. While these gaps have been narrowingfor low-income and lower-middle-incomecountries, upper-middle-income countrieshave only managed to maintain their relativeposition in relation to high-income countries.At the regional level, these gaps have widened

or remained stagnant in three of six develop-ing regions, with TFP growing faster in high-income countries than in Latin America andthe Caribbean, the Middle East, and Sub-Saharan Africa.

The relationships between income growth,technological progress, capital accumulation,and welfare are, of course, much more complexthan can be summarized in a simple measureof TFP, partly because each factor of produc-tion and the technology with which factorsare combined are dependent on one another.As discussed in chapter 3, capital goods oftenembody significant technological progress andthere is no simple way to distinguish betweenthe contribution that each makes to growth.Similarly, technology in the form of knowl-edge of business processes and of science andgeneral experience is embodied in labor.Moreover, the contribution of technology towelfare is only imperfectly measured by its im-pact on GDP (box 2.1).

Technological progress can lower costs,improve quality, create new products, and help reach new markets Even though measures of TFP and its progressgive us a sense of the relative dispersion oftechnological progress, they tell us little aboutthe mechanisms by which technology influ-ences development. Technological progress in-volves much more than doing the same thingsbetter or with fewer resources. It is more dy-namic, involving both the creation of new andnew-to-the-market products and productiontechniques, but also the spread of these tech-niques across firms and throughout the eco-nomy. While the mechanisms by which tech-nological progress contributes to developmentare in some sense obvious, the following de-serve special mention:

• Technological progress can spurdevelopment by lowering the costs ofproduction and enabling the exploitationof increasing returns to scale. By improv-ing the efficiency with which existing


54

Table 2.1 Disparity among TFP levelsremains wide

TFP relative to that Annual TFP of the United States, growth,

2005 1990–2005

(annual (index, percentage

Regions U.S. � 100) change)East Asia and the Pacific 8.4 5.1Europe and Central Asia 21.7 2.2Latin America and the Caribbean 19.3 0.2Middle East and North Africa 13.3 0.5South Asia 5.8 2.3Sub-Saharan Africa 5.6 0.2

Income groupsHigh-income OECD

countries 77.1 1.3High-income non-OECD

countries 53.1 0.7Upper-middle-income

countries 23.7 1.2Lower-middle-income countries 9.6 3.2Low-income countries 5.2 1.7

Source: Poncet 2006.Note: OECD � Organisation for Economic Co-operation andDevelopment; TFP � total factor productivity.

products are produced, new technologiescan open up the possibility of increasingoutput and, assuming that markets areavailable, taking advantage of previouslyunexploited increasing returns to scale.

• Technological progress in one sector cancreate new economic opportunities inother sectors. Lower production costscan create whole new products, or evensectors. A new-to-the-market innova-tion in one sector can result in a flower-ing of activity in other sectors bycreating a demand for and supply ofgoods and services that did not existpreviously (box 2.2).

• The benefits of a new technology canextend well beyond the immediatesector or good in which the technologyexists. This is the case if the initialproduct is an important intermediategood in the production of other goods,

for example, telecommunications orreliable electrical service.

• Technology can yield quality improve-ments. Such improvements can enable adeveloping country to penetrate moredemanding consumer and intermediatemarkets. This can be as simple as em-ploying machinery and equipment thatproduce goods and services that corre-spond to the more exacting expectationsand standards of consumers and busi-ness clients in high-income countries.Technology in this sense extends beyondengineering technology to include man-agement techniques. For example, one ofthe big challenges facing Ugandan fish-eries was creating systems of quality as-surance that allowed them to meet phy-tosanitary standards in the EuropeanUnion on a sustained basis (Chandra andKolavalli 2006).


55

While the relative level of TFP provides a senseof the efficiency with which factors are com-

bined, it ignores the welfare contributions of techno-logy that do not have an immediate impact on GDP.For instance, in national accounts, the purchase ofmachinery that reduces air and water pollution, suchas scrubbers for smokestacks, may not increaseGDP. While the purchase of machinery will berecorded as income accruing to the producing firm,this may be offset by reduced profits and other fac-tor payments of the purchasing firm. Thus, eventhough over the longer term the machinery may con-tribute to a reduction in days of work lost because ofrespiratory illnesses, and therefore to an increase innational income, over the medium term the machin-ery would have little measurable effect on GDP orTFP despite the improvement in air quality, whichwould provide a general, if not monetized, benefit.Similarly, technological advances that reduce the costof public services may have little impact on recorded

Box 2.1 Technology can contribute to welfarewithout affecting measures of short-term output

income, but may have important implications for thequality of life.

In developing countries, the diffusion of suchtechnology as water and sanitation systems, oral re-hydration techniques to treat diarrhea, immuniza-tion, malaria prevention, and contraceptives havebeen tremendously important for improving house-hold well-being, but such innovations will affect output only over time as improved child health even-tually pays off in terms of greater adult productivity(Alderman, Hoddinott, and Kinsey 2006; Behrmanand Rosenzweig 2004; Glewwe, Jacoby, and King2001). These technologies may also have importantnoneconomic societal benefits, such as improved gen-der equality, which are not recorded in GDP becausewomen are more likely to engage in nonmarket pro-duction, or may appear only with a lag as improvedhealth technologies facilitate women’s entry into thelabor force over time (Bailey 2006; Miller 2005;Schultz 2007).

Even relatively simple technologies canhave far-reaching development impactsTechnological advances do not need to be ex-traordinarily complex or reliant on the mostsophisticated technology to have importantdevelopment impacts. In many low-incomecountries, fairly commonplace technologiesare often in short supply because of weak ca-pacities to implement them (box 2.3), and rel-atively simple innovations can have profoundeffects. The green revolution is a dramatic ex-ample of the effectiveness that even modesttechnological advances can have in boostingincomes among the poor. In addition, greateraccess to the technologies required to storeand process food can increase food security,particularly in communities without access toreliable electricity or means of refrigeration.The use of sawmill waste (sawdust, planershavings, and chipper dust) to produce car-bonized briquettes for use in householdcooking can increase access by the poor to fuelfor cooking while reducing deforestation

pressures.2 Dissemination of the simple skillsrequired to build rainwater collection systemscan greatly improve access to clean drinkingwater and reduce the incidence of diarrhea, amajor cause of infant mortality. Insecticide-treated mosquito nets are a well-known, cost-effective strategy for preventing the spread ofmalaria, but the main challenge in many coun-tries remains developing and implementing amechanism for distributing them to thosemost in need and ensuring that they are used.

Despite these advantages, technologicalchange can also be disruptiveWhile technological progress generates sub-stantial benefits, it can also be disruptive, be-cause its benefits are not necessarily evenlydistributed. In particular, while the introduc-tion of an advanced technology may meannew opportunities for the innovator andreduced costs for consumers, it can result insignificant short-term losses in incomes forcompetitors using older technologies. For


56

In Chile, the creation of a viable internationalsalmon farming industry involved the simultaneous

development of a number of related new-to-the-market products, including the domestic productionof fish tanks, fish eggs, salmon food, and vaccines,and eventually the introduction of additional vari-eties of farmed fish. New process technology wasalso introduced, including systems for feeding, pro-cessing, and stocking fish that met global qualityand phytosanitary standards.

The introduction of a cut flower industry inKenya to serve the European market represents theindirect effect of the successful introduction of theindustry in Colombia to serve the U.S. market. Thenew activity generated a wide range of additionalnew-to-the-market innovations in the form ofgreenhouses and postharvest care facilities topreserve the freshness of blossoms. Process

Box 2.2 Technological innovation may spur furtherinnovation in upstream and downstream activities

technology involved learning how to use chemicalsand mastering the logistical challenge of deliveringthis fragile product to the local airport on time andwith sufficient regularity to meet customers’ just-in-time requirements.

Success in one activity may well lead to furtherinnovation and technological deepening. The movefrom producing carnations to more fragile and ex-pensive roses is an example. Another example is theshift to higher-quality products such as chilled ratherthan frozen fish fillets. Yet another example of deep-ening is palm oil production in Indonesia, where newprocesses include the production of new varieties ofpalms; the introduction of new crude and processedpalm oil refining technologies; and, notably, the in-troduction of oleo chemical technologies.

Source: Chandra 2006.


57

Arecent study of Rwanda identified simple tech-nologies whose greater use could have a substan-

tial impact on development. For example, the studyidentified a lack of qualified plumbers and watersanitation technicians as a major factor holding backthe implementation of simple rainwater collectionstrategies that have helped improve the quality ofdrinking water supplies in neighboring countries.Similarly, a lack of basic skills, including those neces-sary to manufacture stainless steel products, preventsthe implementation of simple food processing

techniques, such as passion fruit pasteurization andpulping, that could reduce the share of crops lost tospoilage, which sometimes results in the loss of asmuch as 30 percent of a crop. Public sector dissemi-nation of best practices is hindered by poor skills andinappropriate incentives, which result in researchcenters producing local products that take insuffi-cient account of users' needs and requirements. Thetable provides a snapshot of the status of efforts topromote the diffusion of simple technologies inRwanda.

Box 2.3 Promoting appropriate technologies in Rwanda

Diffusion of selected “appropriate” technologies in Rwanda

Rural energy • Biogas for institutions: installations ongoing and spreading• Biogas for households: pilot program of 163 units to start 2007• Micro hydropower: 6 projects in preparation, more in future?• Biofuel: no national program or policy as yet• Wind: no program or policy as yet• Peat: large stocks but limited exploitation• Efficient stoves for urban areas: national program ongoing• Efficient stoves for rural areas: some programs ongoing• Rice and coffee husks for briquette production: limited programs• PV systems: technology available but slow market• Solar water heating: technology available but slow market

Water and sanitation • Roof water harvesting: only on limited scale for households• Boreholes: few and expensive• Hand pumps: imported from region or India• VIP and Ecosan latrines technology: available, limited uptake

Agricultural technologies • Irrigation through treadle and motorized pumps: limited uptakeand transport • Drip irrigation: starting

• Animal traction for tillage and transport: promoted in certain areas• Small tractors for rice puddling and transport: few units imported• Rice threshing and winnowing: few machines available and locally

produced• Rice hulling: opportunities for small-scale processing• Maize milling: machines imported and locally made• Oil presses for sunflower, soya, essential oils: starting• Livestock spraying: locally made machine now available

Low-cost building • Rice and coffee husks and peat for brick burning: some use• Hand brick press machines: locally made and imported• Engine brick press machines: imported

Source: Watkins and Verma 2007.

example, improved production and processingof sugarcane in Brazil has allowed production,incomes, and employment in that country toincrease significantly, but it has done so at theexpense of sugar producers in other countries,who have been unable to compete. While theassociated income losses may be painful,the global impact tends to be positive, becausethe income losses promote the reallocation ofresources and activity to more effective uses.

Technological progress may also benefitcertain classes of workers over others. Tech-nological change that uses high-level skillsmore intensively may hurt less skilled workersin high-income countries by increasing the de-mand for skilled workers and simplifyingtasks or allowing the outsourcing of tasks thatpreviously were accomplished by relativelywell-paid semiskilled workers. Many econo-mists cite the recent tendency for technologi-cal progress to benefit more skilled workers asa major source of the rise in earnings inequal-ity in most advanced countries.3 Note, how-ever, that technical change does not alwaysraise the demand for skilled workers relativeto unskilled workers, nor does the disruptionnecessarily occur to the detriment of low-skillworkers. Thus the weaving and spinning ma-chines that benefited lower-skilled workers byenabling them to produce textiles formerlyproduced by skilled artisans were destroyed inthe Luddite and Captain Swing riots of the19th century (Acemoglu 2002).

Moreover, if changes in earnings indeveloping countries are taken into account, itis no longer clear that technical change hasbeen biased toward skilled workers. By somemeasures, global inequality has not increasedover the past two decades.4 Global incomedistribution has benefited from the rapidgrowth in China and India, which has enabledhundreds of millions of people to escapepoverty. Technical change interacting with in-creased globalization may have increased in-equality within some countries by increasingthe demand for skilled workers. By openingup opportunities for technical progressthrough the production of export goods that

require relatively high-level skills, greater par-ticipation in international trade has led to in-creasing demand for skilled workers, and thusto greater income inequality in some countries(Arbache, Dickerson, and Green 2004; Zhuand Trefler 2005). At the same time, technicalprogress can be strongly pro-poor, for exam-ple, the discovery of simple technologies tostore and process food in areas with insuffi-cient access to electricity or to enable low-costapproaches to combating disease.

The disruptive nature of technologicalprogress can generate important benefits tosociety by spurring competition. For example,the introduction of mobile phone technologyin several developing countries has introducedan important element of competition not onlyin the telecommunications sector, but also inbanking and other information-sensitive sec-tors. Partly as a result, many of the informa-tional asymmetries generated by a lack of effec-tive communications that various middlemenused to exploit have been eliminated, raisingproducer prices and lowering consumerprices.5 These benefits are often accompaniedby shifts in the distribution of income wherebysome groups can lose either relative to othersor in absolute terms. These losses can be diffi-cult for the poor to absorb, underlining theimportance of safety nets to minimize socialconflict and to ensure that overall progressdoes not come at too high a cost for someindividuals.

Measuring technologyin developing countries

The remainder of this chapter is concernedwith measuring the level of technological

achievement in developing countries and re-cent progress in this regard. This first sectiongoes beyond indirect measures of technologylike TFP, and seeks a more direct measure oftechnological achievement by exploring the ex-tent to which specific technologies have per-meated economic activity in developing coun-tries and the intensity of scientific innovationand invention.


58

Measuring technology directly is difficult,mainly because, unlike pencils or automobiles,technology has no easily counted physicalpresence. Nor does it have a well-definedprice that would allow it to be measured andaggregated in the same way that services are.Rather, technology is embodied in products,intermediate inputs, and processes. As aresult, most efforts to measure it have beenforced to use indirect techniques (seeArchibugi and Coco 2005 for a review). Someindexes emphasize inputs into technologicaladvancement, such as education levels, num-bers of scientists and engineers, and expendi-tures on research and development (R&D) orR&D personnel, for example, the index ofinnovation capability put out by the UnitedNations Conference on Trade and Develop-ment (UNCTAD 2005). Other indexes alsoincorporate information on the diffusion oftechnologies and on indicators of innovation,such as the number of patents granted. Thetechnology achievement index, published bythe United Nations Development Programmeis an example. Still other indexes focus on out-puts, such as the share of high-tech activitiesin manufacturing value added and exports, forinstance, the index of competitive industrialperformance published by the United NationsIndustrial Development Organization (UNIDO2002). Some indexes focus more on the mech-anisms by which technological progress isachieved (Sagasti 2003) or by which techno-logical learning occurs (Soubattina 2006). Forexample, the national innovative capacityindex reflects government and firm-levelpolicies associated with successful innovation(Porter and Stern 2003).

Each of these approaches has its strengths,but none of them is entirely satisfactory, bothbecause the indicators used fail to do justice tothe broad definition of technology adoptedhere (box 2.4), and because the methods bywhich these indexes are constructed are some-times arbitrary (Archibugi and Coco 2005). Toovercome these deficiencies, the indexes devel-oped in the remainder of this chapter include anumber of indicators not previously included

in technology indexes. Summary indexes arederived from these along three dimensions oftechnological achievement: the extent of scien-tific innovation and invention, the diffusion ofolder technologies, and the diffusion of newertechnologies. We begin by reviewing currentlevels of technology and their dispersion andrecent trends in a number of indicators that theliterature pertaining to these three dimensionsof achievement has used.

In a subsequent section, summary indica-tors of achievement along each of these di-mensions are derived using principal compo-nents analysis. Their current levels and recenttrends are discussed, an overall index of tech-nological achievement is generated from thesesummary indicators, and a fourth indicator(developed in chapter 3) summarizes the ex-tent to which external technology is used inthe production process.

Scientific innovation and inventionMost technological improvements in develop-ing countries are at least partially dependenton the diffusion of technology from moreadvanced countries. Nevertheless, scientificinnovation is important in some developingcountries, and advanced technologies oftenneed to be adapted to local conditions, whichmay require further innovation.

The intensity of innovation is closelyrelated to per capita income . . .The degree of scientific innovation in develop-ing countries, as measured by the number ofjournal articles and patents granted (scaled bypopulation), varies sharply with per capita in-come (table 2.2).6 Authors from high-incomecountries report 7 times as many published ar-ticles than those from upper-middle-incomecountries and 88 times as many as authors fromlow-income countries. Variations for measuresof patents granted and license fees earned areeven larger. This result is generally reflected inregional data, with countries in regions withhigher incomes such as Latin America and theCaribbean reporting higher levels of patentsand journal articles than regions such as South


59

Asia or Sub-Saharan Africa, which are domi-nated by lower-income countries. The ratio ofpatents granted to residents to the total numberof patents a country grants (an indicator of theextent to which innovations are generateddomestically) is only weakly correlated withincome. Equally important influences includethe domestic economic structure, the country’sopenness to foreign direct investment (FDI) (seechapter3), thedomestic costsofmakingapatentapplication, local intellectual property rights,and the legal environment—all factors that dic-tate the potential benefits from holding a patent.

Patent activity in middle-income countrieshas increased over the past 20 years (figure 2.1), primarily because of a sharp jumpin patenting (relative to population) among

upper-middle-income countries in the early1990s following the integration of thetransition economies of the former SovietUnion into the world economy. The continu-ous increase in patent activity among lower-middle-income countries mainly reflects activ-ity in China, whose share in world patentapplications rose from about 1.5 percent inthe late 1980s to a peak of nearly 10 percentin 2004. Excluding China, additional patent-ing activity in lower-middle-income coun-tries has been relatively modest. Whilepatent activity has also risen in low-incomecountries, it remains far below that inmiddle-income countries both in the ab-solute numbers of patents issued and relativeto the population.


60

Available indicators provide only a partial view ofthe level of technological achievement in devel-

oping countries and of the gap with high-incomecountries. Most available indicators reflect the quan-tity of technology used, whereas the quality of deliv-ery is often what is critical. For example, the value ofelectricity in production is a function of both theamount consumed and its reliability. In general,global indicators of technology levels do not takedifferences in quality sufficiently into account. To theextent that quality of delivery varies systematicallywith income levels, the indicators likely understatethe differences between rich and poor countries. Forexample, Kaufmann, Leautier, and Mastruzzi (2005)find that access to infrastructure services (similar towhat we measure here) and the quality of infrastruc-ture services in urban areas are both closely relatedto the strength of governance, which is itself highlycorrelated with income levels.

Nor do the available indicators reflect the dispar-ity of achievement within countries. National indica-tors of technological achievement are based on coun-try averages, but large gaps exist in the extent towhich technologies are used within regions, income

Box 2.4 Shortcomings of available measures of technological achievement

groups, and countries. For example, the relativelylow performance of South Asia reflects the slownessof technology diffusion from the relatively advancedmajor cities to rural areas, as well as from the rich tothe poor within urban areas. Indeed, the degree oftechnological diversity across Chinese regions orIndian states mirrors the extent of diversity acrossdeveloping countries, with regions containing largetechnologically sophisticated cities, such as Mumbaior Delhi, being well ahead of areas that lag behind ineconomic development.

Finally, most indicators tend to be biased towardgoods (as opposed to services), and among these, toward electronic and other high-tech goods. Mostmeasures also focus on product technology (goodsand services that themselves are highly technical)rather than final (or intermediate) goods and servicesthat may be technologically unremarkable, but whichare the result of a technologically sophisticated pro-duction process, for example, maize that is producedusing sophisticated crop rotation methods, enhancedirrigation and fertilization strategies based on satel-lite imaging, and bioengineered seeds.

. . . although the Europe and Central Asiaregion is an outlierReflecting a history of advanced scientificand engineering work in a number of formerSoviet bloc countries, the Europe and CentralAsia region has relatively high levels of scien-tific innovation and invention (table 2.2).

Publication rates there are equal to those inmany high-income countries, and patent ac-tivity is more than twice the level in any otherdeveloping region. The region is also themost self-reliant of developing regions interms of patent activity, with only 28 percentof patents being filed by nonresidents, a fig-ure that is even lower than the high-incomecountry average of 38 percent. The East Asiaand Pacific region also scores high in terms ofpatents, although its publication record ismore in keeping with that of other develop-ing regions. In some countries, such as Chinaand India, conscious efforts to raise R&Dspending have led to higher levels of scientificinnovation than might be expected based onincome (Lederman and Saenz 2005), whilelow levels of innovation in Latin Americaand the Caribbean reflect an academic re-search tradition with few links to industry(Maloney 2006).

Penetration of older technologiesThe clear dominance of high-income countriesin the number of scientific and technical jour-nal articles published, the number of patents


61

Figure 2.1 Patent activity is rising in middle-income countries

Total patent applications per 10,000 people

Source: World Development Indicators.

0

1985

1987

1989

1991

1993

1995

1997

1999

2001

2003

0.2

0.4

0.6

1.8

1.2

1.0

0.8

1.6

1.4

Upper-middle-income countries

Low-incomecountries

LMY excl. China

Lower-middle-income countries (LMY)

Table 2.2 Scientific and innovative outputs

Scientific and Number of Royalty and technical journal USPTO EPO Total patents to license fee

articles, patents, patents, patents, nonresidents receipts Regions and income groups 2003 2006 2005 2003 2003 2004

(percent of (percent of Regions (Number per million people) total) GDP)East Asia and the Pacific 17 0.7 0.01 37 77 0.02Europe and Central Asia 90 0.9 0.40 95 28 0.06Latin America and the Caribbean 35 0.7 0.21 46 98 0.03Middle East and North Africa 18 0.1 0.03 — — 0.02South Asia 9 0.5 0.07 1.4 60 0.00Sub-Saharan Africa 5 1.4 1.16 157 100 0.06

Income groupsWorld 111 38.6 11.4 127 41 0.27High-income countries 584 135.1 42.6 331 38 0.33Upper-middle-income countries 85 1.4 0.40 91 42 0.04Lower-middle-income countries 21 0.6 0.01 46 64 0.03Low-income countries 7 0.4 0.07 3.5 56 0.00

Source: World Development Indicators, USPTO, EPO, and World Intellectual Property Office data.Notes: EPO � European Patent Office, USPTO � U.S. Patent and Trademark Office. To reduce home bias, the total patentsgranted by the USPTO to high-income countries exclude those granted to the United States, and the total patents granted by theEPO exclude those granted to high-income European Union countries.— � not available.

granted, and the extent of licensing androyalty fees realized points to the relativelyminor role that at-the-frontier innovationplays in determining technological progress indeveloping countries and the relative impor-tance that adoption and adaptation of existingtechnologies must play. We look first at thediffusion of older technologies.

The major technological innovations of thepast two centuries—such as steam power, elec-tricity, the internal combustion engine, thetelephone, radio, and television—exist tosome degree in virtually every country in theworld. However, the extent to which they areavailable within countries varies enormously,depending both on the technical adaptivecapacity of the country (chapter 3) and on theaffordability of the technology.

Many of the most prominent technologiesare in the manufacturing sector. However, the

efficient extraction of natural resources oftenrequires advanced technology and can encour-age technological progress. Indeed, the failureto absorb new technologies is an importantreason for the slow growth of many naturalresource-based economies in Latin America(box 2.5).

Affordability limits the penetration ofelectrical networks in some countries . . .Affordability, exacerbated by fiscally con-strained governments, helps explain themodest diffusion of many technologies criticalto development. This appears to be the casefor a number of infrastructure technologiessuch as electricity (figure 2.2), rail and roadtransportation, and fixed-line telephony. Ineach of these cases, a reliance on governmentsto provide these services, coupled with weakinstitutions and a lack of domestic capacity to


62

Experience in the extractive sectors can help gener-ate new technologies that in turn can be used as

a source of global comparative advantage. The expe-rience of Brazil’s Petrobras, a majority state-ownedcompany, in exploiting that country’s considerabledeepwater oil and natural gas resources provides aninteresting example.

To exploit the Campos Basin, which lies in theAtlantic at a depth of more than 100 meters andnow accounts for nearly 84 percent of Brazil’s oilproduction, Petrobras created the anticipated pro-duction system on a floating platform. This ad-vanced system, developed with the help of foreignexperts, cut the delay between discovery and earlyproduction of deepwater fields from as long as sixyears to a mere four months and has since become amodel for the industry worldwide.

Petrobras has successfully leveraged this experi-ence, developing many patents both on its own andin conjunction with the rest of the industry, universi-ties, and research institutes. It has invested heavily inresearch and education, creating its own R&D cen-ter, to which it allocates 1 percent of its gross in-come. The center, whose staff is increasingly made upof Brazilian experts, has contributed significantly to

Box 2.5 Deepwater petroleum technology in BrazilPetrobras’s many patents and continues to help de-velop cutting-edge technology for the company.

Petrobras is now recognized as a world leader in allphases of deepwater technology—from drilling; tounderwater completion, pumping, and productionusing floating structures; to mooring and processing—with its particular expertise is in the areas of un-manned subsea installations, marine engineering, andfloating production systems. About two-thirds of itsproduction is at a depth of more than 300 meters, andat various times Petrobras has set a number ofrecords, including oil production at a water depth of1,853 meters and the then-deepest exploration well(2,853 meters) in the giant Roncador field.

Petrobras has used its advanced technology toperform exploration and production work inAngola, Argentina, Bolivia, Colombia, Nigeria,Trinidad and Tobago, and the United States and hasacquired offshore exploration blocks and interests inEquatorial Guinea, Libya, Senegal, and Turkey(Black Sea). It has also recently signed various agree-ments in China, India, Mexico, Mozambique, andTanzania.

Source: World Bank.

maintain systems, has limited their diffusionin a number of low- and middle-incomecountries.

Other factors, such as industrial structure,climate, tax policies, and preferences, are alsoat play. In the case of electricity, the way acountry organizes its power sector (the processtechnology employed) can also have a strongbearing on the diffusion of the specific tech-nology within the economy. For example,many countries in the former Soviet bloc enjoynear-universal access to electrical power andper capita consumption rates that are morethan double those in any other developingregion (table 2.3). This reflects a much earlierdecision to emphasize electrification and theprovision of subsidies under communist rule.Access to power in other regions is morespotty, with most of the population in mostlarge cities having access (or at least thepossibility of access) to the electrical powergrid, but with a large share of the rural popu-lation, particularly in the poorest countries,

having no or only limited service. In Indiafor example, only 85 percent of rural villageshave access to the power grid. In Sub-SaharanAfrica, only 8 percent of the rural populationhas access to electricity, compared with 51 per-cent of the urban population. In South Asia,only 30 percent of the rural population has ac-cess to electricity, compared with 68 percent ofthe urban population (Besant-Jones 2006).

Moreover, the reliability of the grids variesenormously, partly because of the amount ofelectricity lost through pilferage or in trans-mission. Because of electricity’s importance asan intermediate input, the reliability of theelectrical supply may be even more importantto the diffusion of other technologies than itsavailability. Many machines are sensitive to thequality of electrical power and many processesare intolerant of interruptions. As a result, un-reliable power can be an important factor inpreventing the implementation of these tech-nologies in some countries. For the world as awhole, electricity losses amount to an averageof 9 percent of the power produced. Countriesin East Asia and the Pacific and Sub-SaharanAfrica and members of the Organisation forEconomic Co-operation and Development(OECD) do better than this average, whilelosses in South Asia approach 30 percent.Furthermore, the impact of power reliabilitydiffers across countries. In Bangladesh, forexample, where transmission and distributionlosses represent only 9 percent of producedpower, some 70 percent of managers indicatethat unreliable power is a serious constraint tobusiness. In contrast, in Cameroon andMoldova, where transmission losses are muchgreater than in Bangladesh, the share of man-agers making this complaint is much lower—13 percent in Cameroon and less than 4 per-cent in Moldova (World Bank 2007e).

. . . and restricts access to efficienttransportation . . .Like the electrical network, transportationsystems are old technologies that enableother technologies, and their disseminationwithin countries has been closely affected by


63

Figure 2.2 Electrical consumption varies markedly even at similar income levels (2004)

GDP per capita PPP (current $)

Per capita electric power consumption(kilowatt hour per capita)

0

2,000

3,000

4,000

5,000

7,000

1,000

6,000

0 5,000 10,000 15,000 20,000

Europe and Central AsiaEast Asia and the Pacific

Middle East andNorth Africa

Sub-Saharan AfricaSouth Asia

Latin America andthe Caribbean

Sources: World Bank; World Development Indicators.


64

Table 2.3 Indicators of the diffusion of older technologies

Electric power transmission Electric power consumption and distribution losses Telephone mainlines

2004 2004 2004

Regions (kilowatt-hours/capita) (percentage of output) (per 100 people)East Asia and the Pacific 1,343 7 19Europe and Central Asia 3,637 12 26Latin America and the Caribbean 1,674 17 18Middle East and North Africa 1,289 16 13South Asia 414 26 4Sub-Saharan Africa 550 9 2

Imcome groups World 2,606 9 19High-income countries 9,609 6 54Upper-middle-income countries 3,454 12 23Lower-middle-income countries 1,448 10 19Low-income countries 375 23 3

Price basket for residential fixedtelephone line Road density Rail density

2004 1999 2005

(kilometers of (kilometers of (percentage of gross national road/100 square kilometers rail/100 square kilometers

Regions income/capital/month) of land area) of land area)East Asia and the Pacific — 14.2 0.42Europe and Central Asia 3.5 11.8 0.81Latin America and the Caribbean 3.2 16.1 0.31Middle East and North Africa 4.4 6.8 0.27South Asia 10.6 80.6 1.55Sub-Saharan Africa 29.3 6.4 0.18

Income groups World 2.2 22.1 0.66High-income countries 1.0 41.2 1.17Upper-middle-income countries 3.9 11.9 0.70Lower-middle-income countries 6.0 14.5 0.39Low-income countries 20.7 19.0 0.36

Agricultural machinery and tractors Irrigated land Air transport 2003 2003 2004

(number of registered (per 100 square kilometers (as a percentage carrier departures/1,000

Regions of arable land) of cropland) people)East Asia and the Pacific 93 — 1.1Europe and Central Asia 184 11 2.1Latin America and the Caribbean 123 11 2.8Middle East and North Africa 141 32 1.2South Asia 143 39 0.3Sub-Saharan Africa 13 4 0.5

Income groups World 202 18 3.7High-income countries 433 12 18.0Upper-middle-income countries 173 10 3.2Lower-middle-income countries 113 24 1.3Low-income countries 90 24 0.3

Sources: World Bank; World Development Indicators.Note: — � not available.

government regulation and affordability.Many process technologies (for example, theassembly, sorting, refrigeration, and delivery offresh fruit) depend on an effective transporta-tion network. The diffusion of railroadsamong developing countries varies widely,with the countries of the former Soviet blochaving a much more extensive rail transportsystem than other developing countries at sim-ilar income levels. This variance is explainedin part by differences in population density(figure 2.3). The cost per passenger mile of a rail

system tends to fall with population density,which helps explain the particularly low den-sity of railroads observed in Sub-SaharanAfrica (Stelling and Jensen 2001). Interestingly,with the exception of Europe and CentralAsia, per capita income does not appear to bean important factor in explaining the diffusionof either rail or road networks. Moreover, theobserved distribution of road networks is onlyweakly correlated with population density.Relative to other regions, Latin America andthe Caribbean has significantly more roads


65


Rail density

Road density

People per square kilometer

GDP per capita PPP (current $) People per square kilometer

Figure 2.3 Rail and road densities rise with income and population density (2004)

By income per capita - 2004


0

4

2

6

14

10

12

8

0 200100 300 400

By population density - 2004

0

50

150

100

200

400

300

350

250

0 5,000 10,000 15,000 20,0000

50

150

100

200

400

300

350

250

0 200 400 600 1,000800

0

4

2

6

14

10

12

8

0 5,000 10,000 15,000 20,000

East Asia and the Pacific


Latin America and the Caribbean

Sub-Saharan Africa


South Asia

than would be expected on the basis of popu-lation and income, while the high averageroad density in South Asia mainly reflects highdensities in Bangladesh and India and lowdensities elsewhere in the region.

Considerable disparities in access to roadand rail transport services are found withinmany developing countries. Rural areas in par-ticular suffer from poor access to transport ser-vices. During 1994–2001, only an estimated61 percent of the rural population in low-income countries lived within two kilometersof an all-season road (Briceno-Garmendia, Es-tache, and Shafik 2004). Poor access to trans-port facilities can cause the neglect of poten-tially productive land, limit yields of usedlands to levels below their potential, and re-duce profits from the sale of produce, all ofwhich weakens incentives for farmers to max-imize production, thereby limiting theprospects for alleviating poverty (World Bank2006). Improving road access can thus have adramatic impact on growth in remote areas.

. . . and air transport and telephones A well-developed air transport network is alsoessential for some technologies and may be a

particularly important enabling technologyfor landlocked economies with poor access toports in neighboring countries. Air transportis a newer technology, and its distributionacross countries tends to follow income at themost aggregated level. Thus high-incomecountries registered 18 carrier departures per1,000 people in 2004, compared with 0.3 de-partures for low-income countries. Althoughmiddle-income countries have a higher num-ber of carrier departures relative to populationthan do low-income countries, the cross-country correlation between income and airtransport intensity is relatively low for all de-veloping countries (figure 2.4). This suggeststhat factors such as the importance of tourismto the economy and access to alternative formsof transport—especially relevant for islandnations—are among the most important deter-minants of the intensity of air transport use.

The delivery of fixed-line telephone servicesfollows a similar pattern. On average, the in-cidence of this mature technology amongupper-middle-income countries is less thanhalf that in high-income countries, and in low-income countries falls to almost 5 percent ofdeveloped country levels. Across regions, the


66

GDP per capita PPP (current $) GDP per capita PPP (current $)

Registered air carrier departures(per 1,000 people), 2005


Fixed-line and mobile phone subscribers (per 1,000 people), 2004

0

600

400

800

1,600

200

1,200

1,400

1,000

0

10

20

60

40

50

30

0 5,000 10,000 15,000 25,00020,000




Sub-Saharan Africa


South Asia

0 5,000 10,000 15,000 35,00025,000 30,00020,000

Figure 2.4 Telephone densities are highly correlated with income, but air transport is not

incidence is once again much higher in Europeand Central Asia, reflecting the heritage of thecommunist era. Elsewhere, East Asia and LatinAmerica have fewer than 20 phone lines per100 people and South Asia and Sub-SaharanAfrica have fewer than 5 lines per 100 people.

In contrast to air transport, the cross-country correlation between the availabilityof telephones and income levels is strong (figure 2.4). The cost of residential service is sig-nificantly higher in the low-income regions ofSouth Asia and especially Sub-Saharan Africa(8.4 percent) than in the predominantlymiddle-income regions (table 2.4).

The adoption and adaptation of oldtechnologies varies by sector The diffusion of old technologies has con-tributed to rapid growth in the agriculture sec-tor in many developing countries. During1967–92, TFP (often used as a proxy forincreases in technology) is estimated to haveincreased four times as quickly in agricultureas in the manufacturing sector in both high-income and developing countries (Martinand Mitra 2001). In part this growthrepresents the exit of underemployed farmworkers to better paying jobs in other sectors,but it also represents significant improvements

in seeds; more capital intensive forms of embodied technology such as tractors, fertil-izer, and irrigation systems; and better processtechnology, such as crop rotation and man-agement techniques for disease-resistant crops(box 2.6).

At the same time, the diffusion of medicaltechnologies within low-income countries hasbeen slow. Some of the most important tech-nological developments of the past 100 yearshave been medical, including the discoveryand widespread distribution of antibiotics andthe eradication and effective treatment of awide range of previously deadly or debilitatingviruses, including retroviruses such as thosethat cause HIV/AIDS.

The diffusion of knowledge about treat-ments is generally relatively speedy and effi-cient within the medical community, but theirdiffusion and application within the popula-tion of the developing world is much slower. InEurope and Central Asia, Latin America andthe Caribbean, and the Middle East and NorthAfrica, the average share of children immu-nized for measles, diphtheria, pertussis(whooping cough), and tetanus is 89 percent orbetter, bringing them close to the immuniza-tion rates in high-income countries. East Asiaand the Pacific also posts immunization ratesabove 80 percent. However, immunization


67

Table 2.4 Affordability of fixed-line phones falls rapidly with lower incomes(cost of fixed-line phone service as a percentage of monthly income in dollars and PPP)

Per capita income

Monthly price

Price as a % of monthly income

GNI PPP fixed-line phone GNI PPP

RegionsEast Asia and the Pacific 1,630 5,194 5.9 4.3 1.4Europe and Central Asia 4,143 9,152 9.5 2.8 1.2Latin America and the Caribbean 4,045 8,116 10.0 3.0 1.5Middle East and North Africa 2,198 6,084 7.3 4.0 1.4South Asia 692 3,142 5.1 8.8 1.9Sub-Saharan Africa 746 2,004 14.0 22.5 8.4

Income groupsWorld 7,011 9,424 11.7 2.0 1.5High-income countries 35,264 32,550 27.6 0.9 1.0

Sources: World Bank; World Development Indicators.Note: GNI � gross national income; PPP � purchasing power parity.

rates in South Asia and Sub-Saharan Africaaverage 59 to 63 percent (table 2.5). In part,this reflects particularly low immunizationrates in some of the larger countries in these re-gions, notably India (less than 60 percent) andNigeria (less than 35 percent), which outweighthe better performance of some of the smallercountries, for example, in Sri Lanka, 99 percentof children aged 12 to 23 months are im-

munized. The disappointing failure to de-liver this basic technological service arisesdespite the intense involvement of the interna-tional community in assisting, and in someinstances taking full responsibility for, thisprocess. Moreover, the pace at which theserates are rising is disappointingly low as coun-tries continue to struggle to implementeffective delivery systems. Partly as a result,


68

The green revolution is an example of the dramaticeffects that modest technological advances can

have in boosting the incomes of the poor. The greenrevolution was a decades-long effort, guided primar-ily by public sector and nonprofit institutions, tocreate and disseminate agricultural technologies todeveloping countries. The principal technologiesinvolved were pesticides, irrigation, and syntheticnitrogen fertilizer, which had long been available inindustrial countries, along with the development ofhigh-yielding varieties of maize, wheat, and rice.Asia’s green revolution doubled cereal production be-tween 1970 and 1995 while increasing the land areadevoted to cereals by only 4 percent (World Bank2007b). Even though the impact of the green revolu-tion on the poor was initially a source of controversy,

Box 2.6 The green revolutionby the late 1990s it was clear that poor people hadreaped substantial benefits from higher incomes, lessexpensive food, and increased demand for their labor.The public sector was critical to this effort, becausethe development of new seed technologies has someaspects of a public good: developers cannot capturethe full benefits, because once the seed is widelyavailable, it can be easily reproduced. The green revo-lution also demonstrates some of unintended effectsthat can accompany the adoption of new technolo-gies: the excessive use of agrochemicals has pollutedwaterways, wasteful irrigation has contributed towater scarcity, and high livestock concentrations nearurban areas have contributed to the spread of disease.

Source: World Bank 1998, 2007b.

Table 2.5 Immunization rates lag significantly in South Asia and Sub-Saharan Africa(children aged 12–23 months immunized)

DPT Measles DPT Measles

1993 2003 1993 2003 2003 2003

(ratio to (ratio to (percent (percent high-income high-income

Regions immunized) immunized) countries) countries)East Asia and the Pacific 83 83 79 83 0.87 0.90Europe and Central Asia 80 89 84 91 0.94 0.99Latin America and the Caribbean 78 90 82 93 0.95 1.01Middle East and North Africa 85 91 84 92 0.96 1.00South Asia 59 63 59 61 0.66 0.66Sub-Saharan Africa 49 59 51 61 0.62 0.66High-income countries 88 95 83 92 1.00 1.00World 71 76 71 75 0.80 0.82

Sources: World Bank; World Development Indicators.Note: DPT � diphtheria, pertussis, and tetanus.

child mortality rates remain elevated in theseregions.

The health benefits of clean drinking waterand sanitation facilities have been understoodfor centuries. Nevertheless, one in five peopleliving in developing countries lack access toimproved water sources and only half haveaccess to improved sanitation facilities(table 2.6). In South Asia and Sub-SaharanAfrica, only some 37 percent of the popula-tion has access to improved sanitation ser-vices, while only slightly more than half of theSub-Saharan African population has access to

improved drinking water (this share rises to65 percent if Nigeria, where only 35 percentof the population has access to improvedwater, is excluded). The rest of the developingworld does much better on these measures.For example, close to 90 percent of the popu-lation in Europe and Central Asia has accessto improved water (91.7 percent) and sanita-tion sources (85 percent). Nevertheless, thediffusion of these basics technologies is weakin rural parts of all developing areas, reflectingmore intense affordability issues and therelative scarcity of basic technological literacy


69

Table 2.6 Diffusion of both water and sanitation technology is low in rural areas

Improved water sources

Total population Rural population Urban population

1990 2004 1990 2004 1990 2004

Regions (percent of population with access)East Asia and the Pacific 71.8 78.5 61.4 69.8 97.3 91.9Europe and Central Asia 91.7 91.7 83.4 79.8 97.0 98.7Latin America and the Caribbean 82.8 91.0 60.0 73.0 92.6 96.0Middle East and North Africa 87.5 89.5 78.9 80.8 96.1 96.3South Asia 70.6 84.4 64.9 81.3 88.6 93.6Sub-Saharan Africa 48.9 56.2 36.1 42.4 81.9 80.1

Income groupsWorld 76.4 82.7 63.2 72.2 95.2 94.5High-income countries 99.8 99.5 99.1 98.5 99.8 99.8Upper-middle-income countries 88.1 92.7 73.5 77.8 94.8 97.7Lower-middle-income countries 74.2 80.8 62.9 70.9 96.4 93.1Low-income countries 64.3 75.0 56.7 69.4 87.0 88.1

Improved sanitation facilities

Total population Rural population Urban population

1990 2004 1990 2004 1990 2004

Regions (percent of population with access)East Asia and the Pacific 29.7 50.6 15.3 36.1 65.5 72.4Europe and Central Asia 86.1 85.0 72.0 70.3 93.7 93.0Latin America and the Caribbean 67.4 77.1 35.4 48.7 80.7 85.7Middle East and North Africa 69.9 76.2 52.0 57.9 87.1 92.3South Asia 17.4 37.2 6.3 26.6 50.3 62.7Sub-Saharan Africa 31.5 37.2 23.8 28.2 52.4 53.3

Income groupsWorld 44.4 57.0 22.8 37.7 77.2 79.4High-income countries 100.0 100.0 100.0 100.0 100.0 100.0Upper-middle-income countries 76.7 81.4 52.6 59.9 87.1 88.6Lower-middle-income countries 37.3 55.4 19.7 38.8 72.9 76.2Low-income countries 21.3 38.3 11.6 28.5 49.6 60.5


and the competencies necessary to install andmaintain such systems (see the discussion onbasic technological literacy in chapter 3). Forexample, in China and India, only 44 and 33percent, respectively, of the rural populationhave access to improved sanitation.

Older technologies have become widelydiffused in many countries, but largedisparities remainOlder technologies have penetrated less com-pletely into developing countries than into de-veloped countries, but the gap is much lesspronounced than the gap for indicators of sci-entific innovation and invention. Moreover,the relationship between income levels andthe diffusion of older technologies within thedeveloping world is relatively weak, suggestingthat the efficiency of the regulatory environ-ment and the diffusion of basic skills withincountries are more important than incomes indetermining the actual level of diffusion ofthese technologies. Countries with the highestachievement in each income group find them-selves at about the median level of achievementof the next highest income group. Once again,the level of diffusion of the older technologiestends to be higher for countries of the formerSoviet bloc than for other countries at the sameincome level, while both the upper-middle-income and lower-middle-income countries ofLatin America and the Caribbean tend to re-port lower levels of diffusion than other coun-tries at similar income levels.

The striking differences between Europeand Central Asia on the one hand and LatinAmerica and the Caribbean on the other handin the diffusion of older technologies may re-flect differences in income distribution and inthe nature of R&D activities (box 2.7). Europeand Central Asia had more equal access toeducation combined with greater governmentinvestment in infrastructure, which facilitatedmore rapid diffusion of technologies than inLatin America and the Caribbean. In addition,whereas R&D activity was clearly linked tothe industrial strategy of Soviet-era firms inEurope and Central Asia, R&D in Latin

America was concentrated in universities, wasoriented toward research at the global frontier(but generally not of cutting-edge quality), andhad few links to firms (Maloney 2006).

Penetration of recent technologiesThe relatively slow diffusion of many oldtechnologies in developing countries contrastssharply with the relatively rapid penetrationof newer technologies (table 2.7). Macro-economic turmoil, civil strife, and fiscal con-straints limited the within-country diffusion ofmany older technologies, but more hospitablecircumstances—including low inflation, lowgovernment deficits, and a technical and regu-latory environment that has better harnessedprivate sector financing of new technologies—have contributed to the spread of more recenttechnologies. In a few cases, newer technolo-gies have leapfrogged over older ones, for ex-ample, mobile phones now have higher pene-tration rates in some countries than fixed-linetelephones.

Distinguishing between old and recenttechnologies is necessarily arbitrary. To acertain extent, road infrastructure is an an-cient technology, and yet the technology em-bodied in producing a kilometer of Germanautobahn is completely different from that re-quired to construct a kilometer of dirt track inSomalia. Similarly, exports that are currentlyclassified as high-tech are in some cases evolu-tionary developments from relatively old tech-nology (mobile phones, for example, evolvedfrom radios and fixed-line telephones).7

Nonetheless, the distinction is useful, becausein many cases the factors that have impededthe diffusion of old technologies within devel-oping countries are qualitatively differentfrom those that impede the distribution ofmore recent technologies. For instance, thediffusion of many of the older technologiesdepended upon the creation and maintenanceof expensive government infrastructure at atime when many governments were grapplingwith severe budget constraints and weak tech-nical and governance capacity. Not only aretoday’s technologies being exploited in a more


70

relaxed and stable regulatory environment,but also many of them are being financed andbuilt by private sector investors with access toample funds and outside expertise.

Exports of high-tech goods are onlyloosely related to incomesOne frequently used indicator of the diffu-sion of recent technology is the share ofhigh-tech goods in total merchandise ex-ports. To be sure, the informational contentof this measure has decreased with the pro-liferation of relatively low-tech assembly op-erations of high-tech goods, which in turnhas reduced the level of technological com-

petence associated with a given level of ex-port of high-tech goods. Nevertheless, theshare of high-tech exports is generally posi-tively correlated with other indicators oftechnological achievement.8

Middle-income countries as a group have amuch higher share of high-tech exports thanlow-income countries. Within the middle-incomegroup, however, the lower-middle-incomecountries average a higher share of high-tech ex-ports than the upper-middle-income countries(table 2.8). The East Asian countries have muchhigher shares of high-tech exports than theother regions, and the Middle East and NorthAfrica region has much lower shares than the


71

While much of the value added from extractiveindustries, such as crude oil production and

mining, is a return to land, the technology employedin these activities is often very sophisticated. Someeconomies, such as Australia, Canada, and Sweden,achieved rapid rates of growth over the 20th centurythrough the efficient exploitation of natural re-sources. By contrast, Latin America’s naturalresource-based economies achieved relatively limitedgrowth; until recently, substantial mineral depositshave gone unexploited. Two central reasons explainthe failure to capitalize on Latin America’s naturalresource opportunities.

First, the region had low levels of human capitaland weak institutions that slowed the adoption andcreation of new technologies. Latin American countriesinvested much less than other regions in promoting ed-ucation systems, with the result that by 1870, the liter-acy rate was only one-third to one-fourth as high as inCanada and the United States. Early industrializationreflected the cumulative impact of numerous small ad-vances made by many individuals, but in Latin Amer-ica, the lack of access to education translated into lim-ited innovation and slower technological progress,because colonial institutions deemphasized technicaleducation and universities failed to produce sufficientengineers and scientists through the 19th century.

Box 2.7 Technology and growth in Latin America’snatural resource-based economies

Second, innovation was discouraged and firmentry was inhibited by anticompetitive guilds, labormarkets that were excessively protective of insiders’rights, concentrated credit markets that only lent toinsiders, explicit trade barriers that impeded knowl-edge spillovers from trade interactions, and barriersto FDI. The concentration of wealth also discour-aged innovations by newcomers. Rights to organizecorporations and financial institutions were rationedto protect the value of rights held by powerful inter-ests and the costs associated with filing patents wereexorbitant. After the Great Depression, attempts toforce rapid industrialization through import substitu-tion policies led to sectors that were out of line withcomparative advantage, that were walled off fromcompetition and sources of innovation, and that re-quired substantial subsidies to survive. Natural re-source sectors, the likely source of Latin America’scomparative advantage, were starved of capital andworkers who were drawn to the heavily subsidizedand inefficient manufacturing enterprises.

The combination of inefficient industrializationwith the stifling of natural export sectors left manycountries in the region vulnerable to balance of pay-ments crises and severely constrained growth.

Source: Lederman and Maloney 2007.

other predominantly middle-income regions. Inpart these differences reflect the impact of long-standing policies in several East Asian countriesthat emphasized exports of increasingly sophis-ticated products and these countries’ proximityto transport corridors that facilitated their par-ticipation in international production networks.

Although well below half the level of EastAsia and the Pacific, Latin America and theCaribbean’s share of high-tech exports relativeto the total of manufactured exports of 13 per-cent in 2004 was larger than that of the otherregions (table 2.8). This mainly reflects a highshare of high-tech exports in Mexico (20 per-

cent). The average share for individual coun-tries in Latin America and the Caribbean is8.6 percent, with high-tech exports represent-ing 7 percent or less of the total merchandiseexports of Argentina, Colombia, Honduras,Nicaragua, and Paraguay.

Personal computers have diffused relatively slowly . . .Personal computers (PCs) are among the re-cent technologies for which data exist for awide number of countries. PCs are a relativelynew technology that, despite their present-dayubiquity in high-income countries, have


72

Table 2.7 Diffusion of recent technologies

Internet users Internet bandwidth Broadband subscribers

Percentage change Percentage change Percentage change2005 1999–2005 2004 1999–2004 2005 2001–05

Regions (per 1,000 people) (megabytes/second) (per 1,000 people)East Asia and the Pacific 89 48 8,735 149 26 236Europe and Central Asia 190 48 6,670 132 21 208Latin America and the Caribbean 156 41 4,513 121 16 89Middle East and North Africa 89 64 899 91 — —South Asia 49 66 2,249 114 1 131Sub-Saharan Africa 29 42 114 62 — —

Income groupsWorld 137 20 43,856 108 42 59High-income countries 527 14 121,433 107 163 45Upper-middle-income countries 196 36 5,611 126 21 147Lower-middle-income countries 95 50 5,533 134 23 187Low-income countries 44 72 708 120 1 143

Personal computers Cellular subscribers Digital cellular subscribers

Percentage change Percentage change Percentage change2004 1997–2004 2004 1995–2004 2004 1999–2004

Regions (per 1,000 people) (per 100 people) (per 1,000 people)East Asia and the Pacific 38 26 24 58 257 54Europe and Central Asia 98 20 44 79 512 43Latin America and the Caribbean 88 17 32 51 337 42Middle East and North Africa 48 17 13 73 142 70South Asia 12 29 4 87 40 88Sub-Saharan Africa 15 11 8 61 83 47

Income groupsWorld 130 14 28 37 284 29High-income countries 579 12 77 28 768 19Upper-middle-income countries 113 18 48 58 521 37Lower-middle-income countries 45 23 24 61 255 54Low-income countries 11 25 4 92 43 88

Sources: World Bank; World Development Indicators.Note: Period growth rates are compound annual growth rates.— � not available.

actually diffused relatively slowly throughoutthe world since their introduction in the1980s, at least compared with the speed atwhich the use of mobile phones and the Inter-net has spread. Thus in 1995, France had justunder 145 computers per 1,000 inhabitants,fewer than half as many as in the United Statesat the time (325) and roughly the same as inHungary today. France now has 575 comput-ers per 1,000 inhabitants, compared with 762in the United States, and many developingcountries in Europe and Central Asia andLatin America have PC ownership rates simi-lar to Hungary’s. Indeed, the regional averagefor Europe and Central Asia is brought downby low penetration rates in Turkey andUkraine (the second and third most populouscountries in the region, respectively), whichhave only 52 and 28 computers per 1,000 in-dividuals, respectively. If we use the simple av-erage of the penetration rate in individualcountries in the region, there are about 150computers per 1,000 people, with many coun-tries posting penetration levels close to the un-weighted average for high-income countries(460 PCs per 1,000).9

Nevertheless, three-quarters of low-incomecountries have 15 or fewer PCs per 1,000 peo-ple and one-quarter have fewer than 5 per1,000 people. Yet, several low-income coun-tries have substantially more. Mongolia, forexample, reports having 133 PCs per 1,000people, illustrating that even though the densityof PC ownership is correlated with income,substantial variations exist across countries atsimilar income levels.

. . . while diffusion of the Internet andmobile phones has been extremely rapid The penetration of Internet use, a more recenttechnology,10 offers an interesting comparison(figure 2.5). Internet bandwidth consumptionand the number of broadband subscribersmore than doubled from 1999 to 2004 in bothmiddle- and low-income countries. High-income countries have almost as many PCsper capita as there are Internet users in devel-oping countries, which have twice as many Internet users as PCs. The ratio rises as percapita incomes decline, with four times asmany Internet users as PCs in the Middle East


73

Table 2.8 Share of high-tech products in total exports(high-tech exports as a percent of manufactured exports)

2004 1999–2004

Percentage Regions point changeEast Asia and the Pacific 33.4 2.7Europe and Central Asia 8.7 �0.7Latin America and the Caribbean 13.1 �1.4Middle East and North Africa 3.2 �0.4South Asia 4.1 0.5Sub-Saharan Africa — —

Income groups World 21.3 �0.4High-income countries 22.3 �0.4Upper-middle-income countries 16.2 �3.1Lower-middle-income countries 22.2 4.0Low-income countries — —

Sources: Centre d’Etudes Prospectives et d’Informations Internationales; World Bank.Note: — � not available.

Internet users per 1,000 people, 2004


0

200

400

300

600

100

500

0 5,000 10,000 15,000 20,000



Middle East and North AfricaLatin America and the Caribbean

Sub-Saharan Africa


South Asia

Figure 2.5 The incidence of Internet usevaries widely across countries

and North Africa and South Asia. The capac-ity to share an Internet connection, either for-mally through a commercial venture such asan Internet café or informally, makes Internetuse much more affordable than owning a PCand lies at the root of this difference.

A lack of infrastructure helps explain weakpenetration rates in some low-income coun-tries. For example, even though Internet pene-tration rates rose by 41 percent in Sub-Saharan Africa from 1999 to 2005, Internetpenetration in the region remains the lowestamong developing regions, in part because nohigh-speed, low-cost backbone exists to con-nect eastern and central Africa to the rest ofthe world. As a result, Internet transactionsmust be made via satellite, which provideslower bandwidth at higher cost than fiber op-tics (Kenyan call center operators pay $7,000per megabyte of bandwidth compared witharound $500 for operators connected by fiberoptic cable in India). As a result, prospects forSub-Saharan Africa are expected to improvefollowing the recent installation of a fiberoptic backbone along the western coast of thecontinent and the expected completion of asimilar backbone along the eastern coastin 2008.

Technology is also providing solutions forovercoming infrastructure costs. In a numberof countries, wireless broadband connectionsare outpacing digital subscriber line (DSL) andcable as a mechanism for distributing Internetaccess to customers. So-called 3G mobilephones already provide reasonable bandwidthin many countries, while more advanced stan-dards offer hope for even faster implementationand diffusion. Some 23 developing countries areplanning to, or already have begun to, deployWiMax systems, a wireless, broadband Internetstandard touted as the successor to today’s WiFiand 3G systems. Those with existing WiMaximplementations include the DominicanRepublic, Pakistan, South Africa, and Uganda.

The ability to share the fixed costs of a mo-bile phone and its monthly subscription costs,along with its portability, have facilitated thediffusion of this technology in developing

countries. Although lack of competition anddifficulties innovative entrepreneurs encoun-tered in getting licenses slowed the initial dif-fusion of mobile phone technology, much haschanged in recent years (Sullivan 2007).Mobile phone ownership rates in developingcountries—even in the poorest countries—arerising rapidly, having almost doubled in low-income countries between 2000 and 2004. In-deed, new subscribers are signing up at such afast pace that the data in table 2.7 are alreadybroadly out of date.

Because the market is evolving so rapidly,with new applications for mobile phonetechnology being developed on a regular basis,evaluating its overall impact is difficult. Pene-tration rates in Europe and Central Asia andLatin America and the Caribbean are alreadyhigh, rivaling those observed in high-incomecountries less than 10 years ago. Penetrationrates in East Asia and the Pacific are somewhatlower on average; however, looking at only themiddle-income countries in the region and ex-cluding small island economies, the averagepenetration rate in East Asia and the Pacificis higher than in Latin America and theCaribbean. Penetration rates in low-incomecountries are much lower, on average, al-though some countries have reached levelscomparable to those in middle-income coun-tries. As of 2005, six Sub-Saharan Africancountries (Botswana, Gabon, Mauritius, theSeychelles, Sierra Leone, and South Africa) hadmobile phone penetration rates above 30 per-cent. Although penetration rates in South Asiaare also low, the large populations of thesecountries and the pace at which firms areadding customers means that globally, a sub-stantial proportion of new mobile phonesubscribers comes from developing countries.11

The rapid penetration of mobile phonetechnology reflects in part the process bywhich it has been financed. Unlike most fixed-wire telephone systems, railroads, and electri-cal grids, mobile phone technology has beenintroduced into most developing countries bywell-funded private operators working withina relatively competitive environment. As a


74

result, the creation of the necessary infrastruc-ture for these systems has not been held backby the government financing and bureaucraticconstraints that slowed the diffusion of oldertechnologies. Moreover, microfinance tech-niques have facilitated expansion of the de-mand side of the business (Sullivan 2007).

The technological and economic implica-tions of the rising penetration of mobile phonesare only now being assessed. In poor, ruralareas, where the transportation of goods andpeople is heavily constrained by poor infra-structure, the introduction of cheap, personalcommunications may be of great value both asa substitute for moving people and to assurethat the movement of people or goods is worth-while. In particular, the availability of relativelycheap and efficient communications has re-duced informational asymmetries in a number

of sectors, increasing producers’ revenues andlowering consumers’ costs (albeit at the ex-pense of middlemen). In addition, this technol-ogy is increasingly being used to enable a de-gree of arm’s-length financial intermediationthat many argue is critical to development, butthat has largely been unavailable in the pastbecause of a lack of infrastructure (box 2.8).

The diffusion of new technologies hasencouraged rapid growth in businessservicesThe Internet, greater availability of comput-ers, and faster communications have com-bined to greatly expand the potential for de-veloping countries to supply services from adistance in a process called offshoring. Ini-tially offshoring services were concentrated onlower-end software services and business


75

The poor confront considerable challenges in gain-ing access to well-functioning savings and pay-

ments services. Financial institutions do not exist inmany rural areas, and those that do often imposehigh minimum balance requirements (reflecting highunit transaction costs for small accounts) that arewell beyond the reach of poorer households. How-ever, the adaptation of technology has allowed someinnovative financial institutions in Sub-SaharanAfrica to extend financial outreach to the poor.

For example, the Equity Bank in Kenya has out-fitted a series of vans with laptops and telecommuni-cations facilities to act as mobile banking units. Ithas also designed flexible savings mechanisms withemergency loan facilities. Teba Bank of South Africahas developed a smart card that uses existing mobilephone technology to provide low-cost, electronicbanking services (savings and payments) for low-income customers. The program was originally de-veloped to handle wage payments for migrant work-ers. The value of the cards can be topped up or thecards can be used to make purchases at any of thesimple wireless terminals that have been placed inshops frequented by low-income clients. Remote

Box 2.8 Innovative use of communications technologyis improving financial access for the poor

Transaction Systems in Uganda is introducing a similar, but more sophisticated scheme. A systemdeveloped by Celpay allows clients in the DemocraticRepublic of Congo and Zambia to use their mobilephones to pay bills. The client establishes an accountwith Celpay and can then make purchases by textinga request to Celpay, which will transfer money to themerchant’s account. Security is provided by the useof a personal identification number, which is neededto complete the transaction.

In a series of surveys of banking services in threemiddle-income and four low-income countries, Bank-able Frontier Associates (2007) found that eventhough only 1.5 percent of the adult population inSouth Africa was using mobile phone banking, thepotential for the service was large. Between 7 and41 percent of the unbanked population of the coun-tries surveyed (Botswana, Kenya, Namibia, SouthAfrica, Tanzania, Uganda, and Zambia) has access(including shared access) to a mobile phone, andthese penetration rates are rising.

Source: Bankable Frontier Associates 2007; World Bank 2007c.

processes as well as call centers. More re-cently, offshoring has moved into such areasas investment and financial services, humanresources, health services, retail functions,logistics, and customer support functions(World Bank 2005). In addition to increasingdemand for labor and boosting export rev-enues, offshoring of services to developingcountries can improve their incentives to pro-vide education and training, help improve thequality of services provided domestically, en-courage technology and knowledge transfers,and minimize (compared with manufacturing)the environmental consequences of economicgrowth. By one account, the most attractivelocations for offshoring global services (basedon costs, the availability of workers with appro-priate skills, and the overall business environ-ment) include Brazil, Chile, China, the CzechRepublic, India, Malaysia, the Philippines,and Thailand (A. T. Kearney 2007). As thecomplexity of services offshored increases,geographic proximity to major markets hasbecome more important and has providedgreater opportunities, for example, for theCzech Republic to supply Western Europe andfor Mexico to supply Canada and the UnitedStates. The advantages of fluency in Englishand French, along with shared time zones, haveincreased the potential for African countries tosupply services to the European market.12

While India has dominated the outsourcingmarket, rapid expansion of the business maybe running into capacity constraints as thepool of unemployed and underemployedskilled workers dries up and wages are bidup.13 Eventually, rising labor costs may partlyerode the advantage of the current major off-shore centers, providing greater room forcompetition from poorer countries. A recentsurvey found that the relative cost advantageof the leading offshore destinations fell in2006 (A. T. Kearney 2007). However, partlybecause of learning by doing, these countries’scores along other dimensions, including peo-ple skills and the business environment, haveincreased. As a result, they have been able tomove up the value added ladder by increasing

the sophistication of the services they delivereven as their costs rise. This development un-derlines a major message for policy makersand businesses in developing countries: im-proving the skills of the labor force by devot-ing more resources to education and training,along with improving the overall climate forinvestment, is essential for competing in tech-nologically sophisticated markets.

Logistics represent an important process technologyAs noted earlier, the spread of modern com-munications technology and the diffusion ofcomputers, coupled with quality improve-ments in transportation services, have com-bined to greatly improve the rapid and effi-cient delivery of goods and services, enablingjust-in-time inventory processing and more ef-ficient supply chain management (this subsec-tion is based on Arvis and others 2007). TheWorld Bank’s logistics performance index pro-vides an overall evaluation of the perceived so-phistication with which countries are able todeliver goods and services (figure 2.6). It con-tains several subindexes that measure servicescritical to logistics, including customs, infra-structure, ability to track shipments, and busi-ness processes (competence) along with thetimeliness and cost of deliveries of domesticlogistics companies (table 2.9).

The overall quality of logistics services isclearly correlated with income. The top per-formers are high-income countries (Singapore,with an index of 4.19, ranks number 1), whilethe worst performers are the poorest countriesthat are landlocked or that suffer from severegovernance problems or conflict (Afghanistan,with an index of 1.21, ranks last). On average,low-income countries score significantly lowerthan middle-income countries.

Nevertheless, index levels show consider-able dispersion among countries with similarincome levels. Countries where trade hasplayed a significant role in promoting growth(for example, Chile, China, India, Malaysia,South Africa, Thailand, and Vietnam) tend toscore high relative to their income level.


76

Vietnam, a low-income country, ranks 53rdamong 150 countries, or slightly above the av-erage for upper-middle-income countries. Incontrast, countries where growth has been gen-erated by oil and mineral assets, for example,Algeria, Bahrain, and Saudi Arabia, score lowrelative to income. The absence of a strong

manufacturing sector in these latter countriestends to reduce the political impetus for the re-forms that would improve logistics. Ultimately,countries that achieve high scores on the logis-tics performance index are those that havevigorously pursued reforms to improve the ef-fectiveness of public sector institutions and to


77

Table 2.9 The quality of logistics services in 2005 varies by income(score on logistics performance index)

International Logistics Tracking and DomesticOverall Customs Infrastructure shipments competence tracing logistics costs Timeliness

Regions (Index)East Asia and the Pacific 2.58 2.41 2.37 2.64 2.54 2.53 3.04 3.01Europe and Central Asia 2.59 2.39 2.39 2.61 2.53 2.55 2.97 3.04Latin America and the Caribbean 2.57 2.38 2.38 2.55 2.52 2.58 2.97 3.02Middle East and North Africa 2.42 2.24 2.27 2.44 2.33 2.35 2.95 2.88Sub-Saharan Africa 2.35 2.21 2.11 2.36 2.33 2.31 2.98 2.77South Asia 2.30 2.06 2.07 2.28 2.32 2.32 3.12 2.73

Income groupsHigh-income countries 3.67 3.45 3.66 3.52 3.64 3.71 2.58 4.05Upper-middle-income countries 2.85 2.64 2.70 2.84 2.80 2.83 2.94 3.31Lower-middle-income countries 2.47 2.31 2.27 2.48 2.40 2.45 3.01 2.93Low-income countries 2.29 2.12 2.06 2.32 2.29 2.25 2.99 2.71

Source: Arvis and others 2007.Note: The maximum score attainable is 5; the minimum is 1.

encourage the efficiency of private sector insti-tutions through competition.14

Among the components of the index, assess-ments of the quality of infrastructure, the qual-ity of services, and the ease of customs clear-ance processes are highly correlated acrosscountries. In contrast, the cost of services variesless across countries (with the exception ofmarkedly high road freight rates in Sub-Saharan Africa) and thus makes a more limitedcontribution to cross-country differences in theoverall index. This highlights the importance ofthe speed and reliability of shipping in globallyintegrated production networks. Interestingly,the gap between the best and worst performersin relation to the overall assessment of the reli-ability of the supply chain is twice the averagegap across various dimensions of supply chainperformance. The reliability of the supply chaintends to be determined by its weakest link.

Evaluating overall technologicalprogress

The preceding sections of this chapter havediscussed technological achievement in de-

veloping countries along three dimensions: sci-entific innovation and invention, the diffusionof old technologies, and the diffusion of newtechnologies. In this section we calculate sum-mary indexes of technological achievementalong each of these dimensions, as well as anoverall index that combines these subindexeswith additional information about the extentto which countries are exposed to technologythrough trade and FDI, issues that are dis-cussed in more detail in chapter 3.

Summary indicators for scientificinnovation and technologypenetrationA statistical approach to summarizingtechnological progress

In creating the summary indexes, a statisticaltechnique, principal components analysis, isused to combine subindicators in a flexiblemanner.15 This approach has been widely usedin health economics (Gwatkin and others

2000a, 2000b, 2000c; McKenzie 2003;Montgomery and others 2000; Vyas andKumaranayake 2006), in poverty analysis(Sricharoen and Buchenrieder 2005), in regu-latory policy analysis (Nicoletti, Scarpetta,and Boylaud 1999), in constructing cross-country measures of capital controls (Chinnand Ito 2006) and in the analysis of e-readinessin India (Government of India 2006). Itcontrasts with most existing efforts toconstruct overall indexes of technologicalachievement, which tend to aggregatesubindexes using arbitrary weights with aweak theoretical or empirical basis, by usingthe statistical properties of the underlying datato determine the weights used in calculatingthe summary and overall indexes. Principalcomponents analysis is used to generateaggregate indexes at two points in time,the early 1990s and the early 2000s,16 forscientific innovation and invention, thepenetration of older technologies, andthe penetration of newer technologies. Thesesummary indexes are then combined with anindex of the extent to which countries areexposed to foreign technologies (throughtrade and FDI), which is developed inchapter 3, to generate an aggregate index oftechnological achievement. Table 2.10 lists theindicators that are summarized in both theoverall index and each of the summarysubindexes. The technical annex to thischapter explains the steps taken to calculatethese weights in more detail.

The relationship between technologicalachievement and income varies dependingon the dimension observedFigure 2.7 reports the distribution of the sum-mary subindex for each of the three dimensionsof technological achievement discussed in thischapter (the summary index of the extent of ex-posure to foreign technologies is presented inchapter 3). A quick glance reinforces the earlierconclusion that, by and large, developing coun-tries are not participating in scientific innova-tion at the technological frontier. Indeed, only ahandful of countries, eight of which are former


78

The distribution of technological achieve-ment across the other indicators (diffusion ofold innovations and of new innovations) isalso skewed toward high-income countries, butmuch less so. Thus the intensity with whichupper-middle-income countries exploit botholder and newer technologies is between 50and 60 percent of the level in high-incomecountries. This ratio is between 30 and 40 per-cent for lower-middle-income countries and isabout 23 percent for low-income countries(table 2.11). However, the dispersion of thesummary indicator of the penetration of oldertechnologies within income groups is verywide. Many low-income countries reporthigher utilization rates for older technologiesthan do many upper-middle income countries.This report suggests that other factors—such as


79

Table 2.10 Indicators included in summary indexes of technological achievement

Indicator Measure Source

Scientific innovation and inventionScientific and technical journal articles population World Development IndicatorsPatents granted by the U.S. Patent

and Trademark Office population Lederman and Saenz 2005Patents granted by the

European Patent Office population Lederman and Saenz 2005

Penetration of older technologiesGroup AElectrical power consumption kilowatt-hours/capita World Development Indicators International outgoing telephone traffic minutes World Development Indicators Air transport, registered carrier

departures worldwide % of GDP World Development IndicatorsAgricultural machinery: tractors per 100 hectares of arable land World Development Indicators

Group BMain lines per 100 inhabitants World Development Indicators Exports of manufactures % of merchandise exports World Development Indicators Medium-tech exports % of merchandise exports CEPII BACI database

Penetration of recent technologiesInternet users per 1,000 people World Development IndicatorsPersonal computers per 1,000 people World Development Indicators Cellular subscribers per 100 inhabitants World Development Indicators Percentage of digital mainlines World Development Indicators High-tech exports % of total exports CEPII BACI database

Exposure to external technologyFDI net inflows % of GDP World Development Indicators Royalties and license fee payments % of GDP World Development IndicatorsImports of high-tech goods % of GDP CEPII BACI databaseImports of capital goods % of GDP CEPII BACI databaseImports of intermediary goods % of GDP CEPII BACI database

Source: World Bank.Note: BACI � Banque analytique de commerce internationale, CEPII � Centre d’Etudes Prospectives et d’Informations Internationales, EPO � European Patent Office, FDI � foreign direct investment, GDP � gross domestic product, USPTO � United States Patent and Trademark Office.

Soviet bloc countries, have anything like thesame level of at-the-frontier scientific activityas the high-income countries. While this mayreflect an innate bias in the indicators used(number of journal citations and patent appli-cations), the results are consistent with the viewthat most technical progress in developingcountries occurs through the adaptation andadoption of new-to-the-market or new-to-the-firm technologies rather than through thecreation of new-to-the-world technologies.Moreover, notwithstanding that some firms—and even some cities—in developing countriesdo participate actively at the technologicalfrontier, when viewed from the national level,not even the most advanced developing coun-tries participate at levels comparable to thoseprevalent in high-income countries.


80

Figure 2.7 Distribution of technological achievement by dimension

Index

Source: World Bank.

0High-income



countriesLow-income

countries

0.05

0.20

0.25

0.15

0.10

Scientific innovation and invention (2000–03)

Index

0High-income



countriesLow-income

countries

0.20

0.35

0.40

0.15

0.10

0.05

0.30

0.25

Penetration of older technologies (2000–03)

Index

0High-income



countriesLow-income

countries

0.50

0.80

0.90

0.40

0.30

0.20

0.10

0.70

0.60

Penetration of recent technologies (2000–03)

history, effectiveness with which governmentshave delivered some public sector technologicalservices, and past turmoil—may have had agreater influence than income in explaining theintegration of these technologies into theireconomies.17 In contrast, the diffusion of recenttechnologies is more correlated with incomeand shows both much less variation and lessoverlap across income groups. These results areconsistent with the view that nonfinancial im-pediments to technological diffusion have con-strained the diffusion of more recent technolo-gies by less than they have for older ones.

In terms of scientific innovation and inven-tion, middle-income countries have been catch-ing up, at least in relative terms, but as alreadynoted, the gap between them and high-incomecountries remains large. In addition, the gapbetween most low-income countries and thetechnological frontier has widened furtherboth in relative and in absolute terms.

The story on the diffusion of technology ismore encouraging. On average, in middle-income countries older technologies are dif-fusing at 2.5 times the rate as in high-incomecountries and more than four times as fast asin low-income countries (table 2.12). Whilethis result appears to be robust for middle-income countries (figure 2.8), the variance ismuch higher among low-income countries.Several low-income countries have recordedsubstantial increases in technological progress,for example Benin, Ghana, and Togo recordmore than 100 percent improvements. In manyothers, however, progress has been slower

than in high- and middle-income countries,implying that the technology gap for thesecountries is either stable or widening.

Not surprisingly, the most rapid increases intechnological achievement recorded over thepast decade or so are for more recent technolo-gies, whose starting points are relatively loweven in high-income countries (figure 2.8).Clear indications of catch-up are evident fornewer technologies, with the penetration ratesin upper-middle-income countries increasing1.5 times as quickly as in high-income countries.The pace of increase among low-income coun-tries was more than four times as rapid, but thisreflects, to a significant degree, very large per-centage improvements in a few countries thatstarted off with very low levels. Notwithstand-ing these caveats, most developing countries aremaintaining pace with high-income countriesand many, especially among the upper-middle-income countries, are gaining ground.

New technologies are not as diffused asold technologies, but the gap betweenincome groups is smallerOverall, the penetration of recent technologiesin the economic life of developing countries isless extensive than for older technologies,which is entirely understandable given thelength of time that has passed since the oldertechnologies were introduced. Nevertheless,the gap between countries at different incomelevels is not as striking as one might expect.Many upper-middle-income countries haveachieved levels of technological achievement


81

Table 2.11 Technological achievement indeveloping countries relative to that inhigh-income countries(percent of level in high-income countries)

Scientific Penetration Penetrationinnovation and of older of recent

invention technologies technologies

High-income countries 100.0 100.0 100.0Upper-middle-incomecountries 3.3 58.4 49.6

Lower-middle-income countries 0.6 41.6 31.8

Low-income countries 0.1 23.7 22.7

Source: World Bank.

Table 2.12 Increase in technologicalachievement in developing countriesrelative to that in high-income countries(index, percent increase in high-income countries = 100)

Scientific Penetration Penetrationinnovation and of older of recent

invention technologies technologies

High-income countries 100.0 100.0 100.0Upper-middle-incomecountries 191.6 220.8 162.3

Lower-middle-income countries 157.1 251.8 145.8

Low-income countries 63.7 480.4 411.3

Source: World Bank.


82

Figure 2.8 Increase in summary technological achievement subindexes, 1990s–2000s

Percent change in index, 2000–03 over 1990–93

Source: World Bank.

Note: Countries are ranked according to the level of technological achievement in 2000.

�200High-income



countriesLow-income

countries

800

1,400

1,600

600

400

200

0

1,200

1,000

Scientific innovation and invention


�100High-income



countriesLow-income

countries

200

350

150

100

50

0

�50

300

250

Penetration of older technologies


0High-income



countriesLow-income

countries

20

50

10

40

30

Penetration of newer technologies

similar to those in high-income countries, and substantial overlap is apparent betweenupper- and lower-middle-income countriesand between lower-middle-income and low-income countries. Interestingly, the clear ad-vantage countries in Europe and Central Asiaenjoy along other dimensions of technologicalachievement is less marked here, with EastAsian countries performing better than mightbe expected.

Overall technological achievementTo understand overall technological achieve-ment, two alternative summary indexes weregenerated. The first combines the three sum-mary subindicators of achievement discussedearlier, while the second includes an additional

summary indicator that measures the extent towhich economies are using imported technol-ogy in their production processes. The under-lying components and their recent evolutionare discussed in more detail in chapter 3 in thecontext of the channels by which externaltechnology is transmitted to developing coun-tries. These overall summary indexes are cal-culated using the same basic technique used tocalculate the subindicators (see the technicalannex to this chapter for details), with onedifference: rather than using the raw data asinputs, the previously calculated summary in-dicators are used.

Figure 2.9 reports levels of technologicalachievement in 2000 according to these twosummary indicators. The country coveragediffers somewhat between the two indexes. To


83

Figure 2.9 Alternative summary indexes of technological achievement (2000s)

Technological achievement, excluding external channels

Index

Index

Source: World Bank.

0

0.15

0.10

0.05

0.20

0.30

0.35

0.25

Upper-middle-incomecountries

High-income OECDcountries

High-incomenon-OECD countries

Lower-middle-incomecountries

Low-incomecountries

Overall technological achievement index

0

0.15

0.10

0.05

0.20

0.25



High-incomenon-OECD countries


Low-incomecountries

be included, each country must have data forall variables, and as additional variables areadded, some countries are lost from the sam-ple. Reflecting this requirement, only 20 low-income countries are included in the firstindex and 16 in the second, down from 46 inthe case of the subindicator with the best cov-erage.18 Although these are relatively diversegroups of countries, they cannot be consideredrepresentative of all developing countries. Par-ticular care should be taken in extrapolatingresults derived from these countries to all low-income countries.

Countries with similar income levels canhave very different levels of technologicalachievementWhile the influence of income on technologi-cal achievement is well established, consider-able variation occurs within income groups(figure 2.10). The top performers within de-veloping country income groups achieve atechnology rating about equal to that of themedian country in the next highest income

group, and scores for countries at roughly thesame income level show substantial disper-sion. Overall, the relationship between tech-nological achievement and income per capitais nonlinear, with the rise in technologicalachievement tending to flatten out for coun-tries with per capita incomes between $10,000and $25,000, a group that includes upper-middle-income countries and some of the lesswealthy high-income countries such as Greeceand Portugal. Countries in Latin Americahave weak technology scores given their in-come levels. Despite the perceived technologi-cal prowess of countries in East Asia, exceptfor Malaysia, the highest-scoring country inthe region, the developing countries in the re-gion do not particularly distinguish them-selves, in part because technological diffusionin these countries remains concentrated in afew urban centers and has not diffused widelyelsewhere.

Focusing only on developing countries inthe second panel of figure 2.10, the correla-tion with income remains, but the same


84

Figure 2.10 Technological achievement rises with income levels

All countries

Index

Source: World Bank.

�0.10

�0.05

0.30

0.25

0.20

0.10

0.15

0.05

0

05,

000

10,0

00

15,0

00

30,0

00

35,0

00

40,0

00

25,0

00

20,0

00

Developing countries only

Index

�0.10

�0.05

0.25

0.20

0.10

0.15

0.05

0

02,

000

4,00

06,

000

12,0

00

14,0

00

20,0

00

18,0

00

16,0

00

10,0

008,

000

All countries

Europe and Central AsiaEurope and Central Asia



Middle East and North AfricaLatin America and the Caribbean

Sub-Saharan Africa


South Asia

High-income OECD countries

High-income other countries

Per capita income (PPPs) Per capita income (PPPs)


tendency for a flattening in the relationship isstill observable. For countries in Latin Amer-ica and the Caribbean, the relationship be-tween technological achievement and incomeper capita flattens out at even lower incomelevels, while for countries in Europe and Cen-tral Asia, the pattern follows more closely thatof the overall sample, which includes high-income countries. Among developing countries(excluding those in the Europe and CentralAsia region), technological achievement flat-tens out at an index level of around 0.15 forcountries with per capita incomes around

$5,500. While not conclusive, these results areconsistent with a view that other factors, suchas technological absorptive capacity (see chap-ter 3) limit the level of technological achieve-ment that some developing countries canattain even as incomes continue to rise.

Technological convergence appears to beconstrained by weak absorptive capacityin some regionsFigure 2.11 reports values for the two overallsummary indexes at two points in time,


85

Figure 2.11 Comparison of levels of technological achievement, early 1990s and early 2000s

Technological achievement, excluding external channels

Source: World Bank.

0

0.15

0.10

0.05

0.20

0.35

0.30

0.25



High-incomeother countries


Low-incomecountries

Overall technological achievement index

0

0.15

0.10

0.05

0.20

0.25



High-incomeother countries


Low-incomecountries

Index

Index

1990s 2000s

roughly the early 1990s and the early 2000s.Both indexes offer a broadly consistent viewof technological achievement. They both con-firm that the absolute size of the increases intechnological progress over the decade islarger among higher-income countries thanlower-income countries, but that the relativeimprovement in developing countries has out-paced that in high-income countries, implyingthat catch-up is occurring.

The extent of apparent catch-up is strongestwhen considering the narrower definition oftechnological achievement that includes onlyscientific inputs and the penetration of oldand new technologies. According to this mea-sure, low-income countries for which dataare available have shown the largest percent-age improvement. If changes in the extent towhich countries are making use of externaltechnologies through imports and FDI areincluded, the extent of convergence declinesfor all developing country groups except theupper-middle-income countries. This findingreflects that high-income countries have alsoincreased their imports of high-tech goodsand have also benefited technologically fromthe operation of technologically sophisticatedforeign-owned firms on their soil. However,trade may contribute more to technologicalimprovement in the South than in the North(Lumenga-Neso, Olarreaga, and Schiff2005).

Among upper-middle-income countries,catch-up is particularly strong in Chile, Hun-gary, and Poland, where the level of techno-logical achievement rose by more than125 percent during the 1990s. For mostcountries the pace of convergence was muchslower. As indicated earlier, relatively weakdata coverage across low-income countriesmakes generalizing about their progress diffi-cult. Only 16 low-income countries (10 in thecase of the summary index that includes im-ports of technology and FDI) have sufficientdata for both the early 1990s and the early2000s to permit an estimate of their rate oftechnological progress. Although generaliza-tions to all low-income countries are not

possible, some commonalities do emergefrom this subsample, namely:

• The absolute increase in the overallindex of technological achievement forlow-income countries is about the sameas for lower-middle-income countries(and thus the percentage increase in low-income countries is much greater),strongly suggesting a catch-up effectrelative to the lower-middle-incomecountries (table 2.13).

• The percentage increase in achievementalong the scientific innovation and in-vention dimension for all eight Sub-Saharan African countries for which dataare available, along with Bangladeshand India, lies well below the averagefor middle-income countries. Only Viet-nam approaches the performance ofmiddle-income countries.

• The picture for the diffusion of old inno-vations is decidedly more mixed, with 5 of


86

Table 2.13 Overall technological progressin absolute and relative terms

Technologicalachievement excluding Overall technological

external channels achievement

(Percent change in the index)High-income 94 77Upper-middle-income 127 109Lower-middle-income 137 103Low-income 227 161Low-income

(excluding Sudan) 160 124

(Percent change relativeto high-income countries)

High-income 100 100Upper-middle-income 135 141Lower-middle-income 146 133Low-income 241 208Low-income

(excluding Sudan) 170 160

(Absolute change in the index)High-income 0.096 0.068Upper-middle-income 0.057 0.046Lower-middle-income 0.036 0.028Low-income 0.024 0.022Low-income

(excluding Sudan) 0.024 0.022

Source: World Bank.

16 low-income countries showing stagna-tion or declines in the index between theearly 1990s and early 2000s, while 6recorded large percentage increases, com-parable to those displayed by the moresuccessful middle-income countries.

• Even for the diffusion of recent innova-tions, where successful examples of mo-bile phone and Internet diffusion havebeen much publicized, only 5 of 25 low-income countries (Guinea, Mongolia,Pakistan, Sudan, and Zimbabwe)achieved increases in the index thatexceeded the middle-income average.However, low-income countries did in-crease penetration rates for recent inno-vations more quickly than high-incomecountries.19 Given that new technolo-gies sometimes substitute for older tech-nologies, such as mobile phones forfixed-line telephones, transmission of in-formation over the Internet for trans-mission involving travel and telephones,the previous finding suggests that theoverall pace of convergence in low-income countries may be accelerating,particularly as data for 2005 and 2006suggest continued high growth in thediffusion of mobile telecommunicationsand Internet technologies.20

Econometric evidence supports the viewthat the relationship between income andtechnological diffusion follows an S-curve:technological diffusion is slow at very lowincomes, in part because of difficulties in af-fording new technologies, in part because lowlevels of human capital severely constraintechnological progress. As incomes rise, tech-nological diffusion increases rapidly, particu-larly in percentage terms, because of the lowbase level. At some level of income, however,the pace of technological diffusion slows. Oneexplanation for this slowdown at higher in-come levels is the slow pace of improvement inan economy’s ability to absorb new technolo-gies (its technological adaptive capacity), asdetermined by the level of human capital, the

governance structure, and the infrastructure(Howitt and Mayer-Foulkes 2005; Klenowand Rodriques-Clare 2004; Lederman andSaenz 2005). According to this view, technol-ogy in a country tends to converge toward alevel consistent with the country’s technologi-cal adaptive capacity. As a result, countriesmay experience relatively rapid technologicalprogress for a period, but may subsequentlystagnate at a given level unless they take stepsto further raise their technological adaptivecapacity. Chapter 3 develops the componentsof technological adaptive capacity and de-scribes trends in developing countries.

Technological diffusion over the long term

So far, we have emphasized the technologicalperformance of countries at different in-

come levels over the recent past. Thanks to anew data set developed by Comin and Hobijn(2004), we can now analyze the process oftechnological diffusion over the longer term.This data set traces the extent of diffusion ofsome 100 technologies in 157 countries dur-ing the period 1750–2003.21 For each tech-nology, only countries for which publisheddata exist are included, implicitly restrictingthe sample to countries (and technologies)where a significant degree of diffusion hasoccurred. The data analyzed here are furtherrestricted to include only those country-technology pairs (a data set with one countryand data for 7 technologies would have 7country-technology pairs) where the intensityof use has reached at least 5 percent of the av-erage level of the 10 countries with the highestrecorded level of diffusion. Under this restric-tion, there are 1,181 country-technologypairs, 699 of which correspond to developingcountries, heavily weighted to technologiesdiscovered in the late 19th or early 20thcenturies.22

Two important points emerge from thisanalysis. First, the diffusion of technologyacross the globe has accelerated over time.


87

Table 2.14 reports, for several old and newtechnologies, the number of years that elapsedbetween the discovery of the technology andthe time it reached 80 percent of the countriescurrently reporting data for that technology.23

The acceleration in the pace at which tech-nologies spread across countries is particularly

striking in the communications field, forwhich data are relatively good and countrycoverage is extensive. Thus telephone andtelegram services were invented in the middleof the 19th century, and more than 90 yearspassed before those services reached 80 per-cent of the countries that currently report data


88

Table 2.14 Successful diffusion has accelerated

Period technology was initially discovered

Technology 1750–1900 1900–50 1950–75 1975–2000 Number of countries

(years following discovery until technology reached 80 percent of reporting countries)TransportationShipping (steam) 83 21Shipping (steam motor) 180 57Rail (passenger) 126 93Rail (freight) 124 99Vehicles (private) 96 153Vehicles (commercial) 63 123Aviation (passenger) 60 109Aviation (freight) 60 103

CommunicationsTelegram 91 77Telephone 99 156Radio 69 154Television 59 156Cable television 50 98PC 24 134Internet use 23 151Mobile phone 16 150

ManufacturingSpindle (ring) 111 50Steel (open hearth furnace) 125 50Electrification 78 155Steel (electric arc furnace) 92 91Synthetic textiles 36 75

Medical (OECD only)Cataract surgery 251 19X-ray 93 27Dialysis 33 29Mammography 33 18Liver transplant 28 29Heart transplant 28 27Computerized axial

tomography (CAT) scan 18 29Lithotriptor 15 26

Average (excluding medical) 106.9 60.9 23.5 16.0Average (including medical) 118.9 61.3 25.7 15.5

Source: Calculations from CHAT database (Comin and Hobijn 2004).

for them. In contrast, mobile phones, comput-ers, and cable television reached 80 percent ofthe countries that currently report data in lessthan 25 years. The same sort of accelerationcan be observed for the dissemination oftransportation, manufacturing, and medicaltechnology.

Second, and consistent with the first point,technological diffusion appears to accelerateabove a certain threshold. Table 2.15 consid-ers only those country-technology pairs thathave reached a level of penetration equal to25 percent of the average level observed in the10 countries where the technology is employedmost intensively. Looking at the results for theworld as a whole, the amount of time requiredto go from the 5 percent level to the 25 percentlevel (averaged across country-technologypairs) is much smaller than the time requiredto reach the 5 percent level. For example, inthe first half of the 20th century, for a technol-ogy to reach the 5 percent threshold took, onaverage, 52 years, but only an additional 13

years to reach 25 percent. Moreover, the paceof acceleration has increased over time. Fortechnologies introduced since 1975, a groupdominated by electronics and informationtechnologies, on average, it took 16 years fromits invention for a technology to reach the 5percent threshold in a given country, but onlyanother 3 years to reach the 25 percent thresh-old. Although the pace of diffusion was some-what slower in developing countries than inhigh-income countries, it too follows the samepattern. Because table 2.15 excludes country-technology pairs where the 5 percent thresholdhas been reached, but not the 25 percentthreshold, the recorded diffusion times areprobably lower-bound estimates.24 This pat-tern is consistent with the existence of signifi-cant economies of scale and barriers to entryamong these technologies, such that once thebarriers are overcome and the technology is inplace, scaling up occurs relatively quickly.

While diffusion has occurred relativelyrapidly among successful diffusers, successful


89

Table 2.15 The pace at which technology diffuses has picked up among successful adaptors

1800–99 1900–50 1950–75 1975–2000

Threshold Threshold Threshold Threshold5% 25% 5% 25% 5% 25% 5% 25%

(years from discovery until threshold reached)RegionsEast Asia and the Pacific 60 69 23 28 18 21Europe and Central Asia 91 117 47 57 25 30 18 21Latin America and the Caribbean 71 105 54 72 30 35 18 21Middle East and North Africa 97 118 58 67 25 29 18 21South Asia 52 62 — — — —Sub-Saharan Africa 85 109 56 69 — — 18 21

Income groupsHigh-income OECD countries 63 91 46 60 20 24 13 17Other high-income countries 95 112 57 65 20 25 15 18Upper-middle income countries 83 110 51 64 26 31 18 21Lower-middle-income countries 86 114 57 69 — — 20 22Low-income countries 56 68 — —

World 76 102 52 65 22 26 16 19Developing countries 84 111 54 67 26 31 18 21

Source: World Bank calculations using the CHAT database (Comin and Hobijn 2004).Note: The sample is restricted to only those 567 country-technology pairings where the 25 percent threshold was reached andthat were below 10 percent when they appeared in the database; — � no data.

diffusion is the exception rather than the rule.For example, of 102 country-technologypairings first recorded in 1975–2000, only 56(55 percent) have reached the 25 percentthreshold and only about 35 (34 percent) havereached the 50 percent threshold (table 2.16).For developing countries, the pace (and ex-tent) of diffusion is significantly slower(lower) than in high-income countries, withonly 24 (36 percent) developing countries hav-ing reached the 25 percent threshold and only6 (9 percent) having reached the 50 percentthreshold. This slower diffusion is true evenfor extremely old technologies, a result consis-tent with the idea that affordability and com-petency issues are binding constraints on thefurther diffusion of technologies in thesecountries. This result is broadly consistentwith the observation that for some groups ofcountries, overall technological achievementappears to stop increasing after a given level isreached, and many developing countries maythus face severe barriers to achieving acceler-ated technological progress.

Slow diffusion within countries reflects a nonlinear process As noted earlier, the surprisingly low level ofoverall technological achievement in countriessuch as China and India contrasts with popularperceptions, which are based on the relativetechnological sophistication of some of the twocountries’ major cities and trading centers.However, the same kind of technological diver-sity observed across countries is visible withincountries as well (see box 2.9 for the case ofIndia). For example, although one might haveexpected India to have scored substantially bet-ter than many Sub-Saharan African countriesin overall technological diffusion, in fact, itdoes not. Several technologically advancedcities in India notwithstanding, technologieshave not penetrated deeply in many parts of theIndian countryside. Here the challenge is to putin place a basic infrastructure in the country-side that can support the kind of sophisticatedtechnologies that the elites in the country arecapable of supporting. As one observer put it,an energy technology revolution must precede

90

Table 2.16 Slow diffusion means that many developing countries never reach the 25 or 50 percent threshold

1800s 1900–50 1950–75 1975–2000

Threshold Threshold Threshold Threshold5% 25% 50% 5% 25% 50% 5% 25% 50% 5% 25% 50%

(number of country-technology pairs that have reached threshold)Regions

East Asia and the Pacific 18 0 0 38 9 3 7 2 1 6 2 0Europe and Central Asia 56 19 6 47 23 6 40 18 3 23 13 6Latin America and the Caribbean 80 11 1 95 34 8 31 3 0 19 4 0Middle East and North Africa 28 4 1 44 16 6 9 1 0 6 2 0South Asia 7 0 0 11 3 3 0 0 0 1 0 0Sub-Saharan Africa 27 4 0 83 21 8 11 0 0 12 3 0

Income groupsHigh-income OECD countries 150 114 75 134 93 55 96 87 75 28 26 23Other high-income countries 25 16 7 28 23 14 14 10 8 7 6 6Upper-middle-income countries 90 30 6 112 53 16 61 24 4 29 19 6Lower-middle-income countries 109 8 2 130 38 12 33 0 0 33 5 0Low-income countries 17 0 0 76 15 6 4 0 0 5 0 0

Total number of country-technology pairsWorld 391 168 90 480 222 103 208 121 87 102 56 35Developing countries 216 38 8 318 106 34 98 24 4 67 24 6

Source: World Bank calculations using the CHAT database (Comin and Hobijn 2004).Note: Sample restricted to only those 1951 country-technology pairings that were below 10 percent when they appeared in the database.


any information technology revolution (Fried-man 2007). The rise in China’s index of diffu-sion of new technologies is almost double thatof India, in part because the more technologi-cally backward regions in China have madeprogress in closing the gap with the more tech-nologically advanced regions on the coast (Jef-ferson, Rawski, and Zhang 2007).

The technology employed by firms withinsectors in individual countries also exhibitstremendous variation. In India, most firms,

especially small ones, tend to use low levels oftechnology, and only a few operate near thenational technological frontier. In most sectors,productivity at the national technological fron-tier is about five times the mean level for allfirms (Dutz 2007). For small formal enter-prises, average productivity is even lower:about one-sixth of the level at the technologicalfrontier for each sector and only one-eighththat of top local performers. Smaller informalenterprises are likely to be even less productive.


91

Large segments of the Indian economy are technol-ogy sophisticated. Its high-tech industries are im-

portant global players, its premier education andR&D institutions are recognized internationally.Along with China, it has the largest pool of skilled manpower, including those with degrees inengineering and other technical disciplines. More-over, India has become one of the world's largestmarkets for telecommunications technology, is aleader among developing countries in exports of soft-ware and information technology-enabled services,and has demonstrated its potential to be a majorplayer in biotechnology-pharmaceuticals and the automobile and engineering sectors, with Bangalorehaving emerged as a major international center fortechnological production and innovation.

Nevertheless, on a per capita basis, India contin-ues to lag behind middle-income countries in the rate

Box 2.9 The technological divide within Indiaof technological diffusion, R&D expenditures, attain-ment levels of basic and higher education, availabilityand quality of logistics services, and size of revenuesand employment in software and other high-tech in-dustries. In addition, India does not score substan-tially better than many Sub-Saharan African countriesin terms of the overall penetration of technologies.

The juxtaposition of India's increasing technologi-cal prowess and relatively poor access to technologyin per capita terms largely reflects the limited penetra-tion of technology in rural areas, which account formore than 70 percent of the population, but less than30 percent of GDP. For example, in June 2007 tele-density—the number of subscribers (wired andwireless combined) per 100 individuals—was52.3 percent for urban dwellers compared with 6.5 forrural inhabitants (see the box figure). Although the gapremains large (especially in terms of quality and relia-bility), and indeed, has widened, a surge in rural mo-bile phone access means that by mid-2007, tele-densityin rural India was equal to the level recorded for urbanareas in 1998. Older technologies, such as radio, tele-vision, bicycles, and motorized two-wheelers, tend tobe more evenly diffused than newer ones, such as mo-bile phones, computers, and the Internet, given thelonger time since the former were introduced.

The digital divide between rural and urban areaspromises to narrow over the long term, particularlyin high-income states and near major cities. How-ever, in some states, in the more remote rural areas,and among tribal and other linguistic groups that lagbehind in economic development, the gap may wellincrease over time.

Source: Mitra 2007.

Cities

Rural areas

Number of subscribers per 100 people

Urban and rural teledensity(fixed and mobile), India, 1998–2007

60

50

40

30

20

10

01998 1999 2000 2001 2002 2003 2004 2005 2006 2007a

Source: Telecommunications and Regulatory Authority of India.

a. Estimated.

The skewed distribution of enterprise produc-tivity implies potentially huge productivity andoutput increases are possible, if already existingwithin-country knowledge were to diffuse fromtop performers to the rest of the economy.Assuming that domestic competencies wereavailable (or created) to efficiently use the tech-nologies employed by enterprises at the na-tional frontier, Indian GDP could be 4.8 timeshigher if those technologies were successfullyapplied by their less productive rivals. Similarly,in Brazil, the productivity of innovative firmswith more than 10 employees, which accountfor 26 percent of total sales, is, on average,6.5 times higher than that of similarly sizedfirms classified as weakly innovative (whichaccount for 11 percent of sales, but 38 per per-cent of employment).

Conclusion

All told, the evidence reviewed in this chap-ter suggests that for most developing coun-

tries technological progress is mainly a processof adaptation and adoption of technologiesfrom abroad rather than the creation of new-to-the-world technologies. The pace of technol-ogy dissemination across countries has pickedup considerably over the past 100 years, andmost technologies are available at some level inmost countries, but the extent to which tech-nologies are available differs enormously.Many developing countries made progress inclosing the technology gap with advancedcountries during the 1990s. However, despitemore rapid improvement in technologicalachievement among the poorest countries,enormous gaps in technological achievementremain. Even upper-middle-income countrieshave less than one-third of the level of TFP ofhigh-income OECD countries, and low-incomecountries have only 7 percent. The gap in TFPlevels between high-income countries and LatinAmerica and the Caribbean, the Middle Eastand North Africa, and Sub-Saharan Africa haswidened since 1990. Moreover, the gap be-tween major centers and lesser cities and ruraleconomies remains large even in the most suc-cessful countries.

At the same time, income is not the onlydeterminant of technological progress. Al-though innovation at the technological frontier(as measured by patents and scientific journalarticles) drops off quite sharply as income lev-els decline across countries, considerable over-lap among income groups exists in the extentof diffusion of old technologies. Thus the mostadvanced middle-income countries demon-strate greater technological achievement in oldtechnologies than the least sophisticated high-income countries, while the more advancedlow-income countries rate higher than thelowest-ranking middle-income countries.

The technological gap between high-income and developing countries is more pro-nounced for new technologies; however, manydeveloping countries are acquiring new tech-nologies at a more rapid pace than older tech-nologies. Given that some new technologies,such as mobile phones and to some extentcomputers, are substitutes for old technolo-gies, the rapid diffusion of new technologiesholds promise for a substantial, widespreadadvance in technological achievement. Thisprogress likely reflects several factors: the re-duction of regulatory constraints on economicactivity that has occurred over the past15 years in many developing countries; the en-abling of private sector investors, who are freeof local government budget constraints), totake the lead in implanting many of these tech-nologies; the growing incomes in developingcountries that have improved the affordabilityof new technologies; and the improvements inthe technological absorptive capacity of devel-oping countries and the increased exposure tointernational technology through trade flows,FDI, and a growing international diaspora.This last issue is the subject of chapter 3.

Technical Annex: Construction of the summary indexes

The summary indexes, the overall index oftechnological achievement, and the tech-

nological adaptive capacity index reported inchapter 3 were calculated by aggregating some


92

34 separate variables, with the weights used inthe aggregation calculated by principal compo-nents analysis (see below). This approachdistinguishes these indexes from most of thosereported in the literature, which even thoughthey are based on similar underlying base data,use arbitrary weighting schemes with limitedtheoretical or empirical bases (see Archibugiand Coco 2005 for a review).

A number of existing measures of technolog-ical achievement or technological progress em-phasize inputs into technological advancement(numbers of scientists and engineers, R&D ex-penditure, or levels of R&D personnel), includ-ing, in some cases, even more indirect inputs,such as the general level of education of the pop-ulation and governance factors that facilitatethe absorption of technology (see, for instance,UNCTAD 2005). Other measures focus on out-puts, that is, on indicators of technologicalperformance, such as the shares of high-tech in-dustries in exports and in manufacturing valueadded (UNIDO 2002). Still others focus moreon the mechanisms by which technologicalprogress is achieved (Sagasti 2003) or techno-logical learning occurs (Soubattina 2006). Anoncomprehensive list of prominent technologyindicators includes the following:

• The index of innovation capability ispublished by the United Nations Confer-ence on Trade and Development (UNC-TAD 2005) and consists of an un-weighted average of an index of humancapital (calculated as a weighted averageof tertiary and secondary school enroll-ment rates and the literacy rate) and atechnological activity index (calculatedas an unweighted average of three indi-cators: R&D personnel, U.S. patentsgranted, and scientific publications, allper million population).

• The index of competitive industrial per-formance is published by the UnitedNations Industrial Development Orga-nization (UNIDO 2002) and is calcu-lated as a simple average of four basicindicators: manufacturing value added

per capita, manufactured exports percapita, share of medium- and high-techactivities in manufacturing value added,and share of medium- and high-techproducts in manufactured exports;

• The technology achievement index ispublished by the United Nations Develop-ment Programme (UNDP 2001) and com-bines (a) the indicators of human skills(mean years of schooling in the popula-tion age 15 and older and enrollment ratiofor tertiary-level science programs); (b) thediffusion of old innovations (electricityconsumption per capita and telephonesper capita) and of recent innovations (In-ternet hosts per capita and high- andmedium-tech exports as a share of all ex-ports); and (c) the creation of technology(patents granted to residents per capitaand receipts of royalties and license feesfrom abroad). The index is constructed assimple averages of these indicators withinsubgroups and then across groups.

• The national innovative capacity index(Porter and Stern 2003) focuses on gov-ernment- and firm-level policies associ-ated with successful innovation. It is com-posed of four subindexes: proportion ofscientists and engineers in the population,innovation policy, innovation linkagesand what they call the cluster innovationenvironment. The overall index is calcu-lated as an unweighted sum of the foursubindexes, but the weights assigned toeach indicator in the subindexes are de-termined by the coefficients obtainedfrom a regression of the number of U.S.Patent and Trademark Office patents onthe relevant indicators controlling fortotal population, the proportion of scien-tists and engineers employed, and thestock of international patents generatedby the country between 1985 and 1994.

Estimating weights for variablesusing principal componentsAll the measures discussed above assignessentially arbitrary weights to the different


93

indicators included in the indexes, or in the caseof regression analysis, use weights derived fromspecific assumptions about functional formsand the data generating process. This report hasfollowed a statistical approach, principal com-ponents analysis, to weighting variables. Thisapproach is widely used in health economics(Gwatkin and others 2000a, 2000b, 2000c;McKenzie 2003; Montgomery and others 2000;Yvas and Kumaranayake 2006) and in povertyanalysis (Sricharoen and Buchenrieder 2005). Ithas also been used in regulatory policy analysis(Nicoletti, Scarpetta, and Boylaud 1999) and inconstruction of cross country measures of capi-tal controls (Chinn and Ito 2006). Most recentlyin the technology field, it has been used in a gov-ernment of India study of e-readiness (Govern-ment of India 2006). Principal componentsanalysis permits the calculation of weights foreach indicator included in the overall index in anobjective manner, with the weights being deter-mined by the data—not by subjective judgment.

Principal components analysis is a statisticaltechnique for reducing the dimensionality ofdata, thereby summarizing the informationalcontent in a large set of data by calculating or-thogonal linear combinations of the originaldata series. Essentially, it is a procedure thathelps to reduce the number of variables in theanalysis by calculating combinations of the un-derlying series that contain most of the informa-tion in the larger data set. It involves an exami-nation of the correlation matrix for the variablesand the extraction of the principal componentsof the data obtained from the eigenvectors of thecorrelation matrix whose eigenvalues are largest.

Intuitively, the procedure followed here isakin to an unobserved variable problem. It isassumed that there is some unobserved vari-able called T (technology), and that thisvariable is correlated with a number of othervariables (X), such as R&D expenditures,share of high-tech goods in total manufactur-ing, and so on. These correlated variables aregrouped together into a single data set and theeigenvalues of its correlation matrix are exam-ined to identify a limited number of linearcombinations (principal components) of the

originating data that explain most of the vari-ance in the original data set (at the limit, if allthe eigenvalues were used, all the variancewould be explained).

The first principal component is the linearcombination of the underlying data that ac-counts for the largest amount of variability inthe sample, the second principal component isthe one that accounts for the next largestamount of variance, and so on successivelyuntil all the variance is explained. All compo-nents are orthogonal to (uncorrelated with)each other; therefore each can be interpretedas an underlying force (visible in varyingdegrees in each original variable). By selectingthe n eigenvectors that explain a large share ofthe total variance, the overall dimensionalityof the data set can be reduced to n.

It is assumed that the main correlate in theunderlying data reflects some form of techno-logy and therefore that the principal compo-nents can be used as an index of technology. Thisis essentially the same process as taking a simpleaverage of several indicators, but with theattached weights being determined by the datarather than being imposed by the researcher.

As a large set of indicators are used thatreflects a wide array of country characteris-tics, more than one principal component is re-quired to adequately capture the informationin the overall data set. One approach wouldbe to calculate each of these purely data-driven, but necessarily arbitrary, componentsand use them to calculate the overall index.An alternative approach that uses a multistageprocedure and subdivides the data into groupsthat are economically or statistically highlycorrelated or both was adopted. A principalcomponents analysis on these subgroups canbe used to create a subindex for those vari-ables, which can then be used in a secondstage to calculate an overall summary index.

Two methodologies for determining thesubgroups were employed. The first was basedon an ex ante grouping of the indicators fol-lowing an economic rationale, and the secondconsisted of an ex post grouping of indicatorsbased on an analysis of the correlation matrix


94

and a graphical analysis of component loadingsto identify groups of indicators that are highlycorrelated with each other. Because the overalltechnological achievement indexes obtained bythese two methodologies are very similar, withcorrelation coefficients of 0.99, the remainderof the discussion is limited to the results ob-tained from the ex ante grouping given its morestraightforward economic interpretation.

Data preparationTo maximize the economic comparability ofthe underlying cross-county data, all datawere scaled, that is, expressed as a percentageof population, a percentage of GDP, apercentage of exports or imports, or a per-centage of arable land, as relevant. Soresearchers could minimize the influence ofoutliers and one-off events, the scaled datawere averaged over two time periods for eachcountry: 1990–93 and 2000–03.25 All datawere converted into an index bound between

0 and 1 by subtracting from each variable theminimum observed value in the sample(across countries and time periods) and divid-ing by the difference between the maximumvalue in the sample and the minimum value.26

Hence the value for indicator j for country iand time t is given by

xijt � (Xijt � Min Xj) (Max Xj � Min Xj)

Applying principal componentsanalysis to technologyFor the purposes of this study, 34 variables wereidentified that bore an ex ante relationship withtechnology and for which adequate countrycoverage existed over the 1990–2006 period tosupport the calculation of two indexes, one forthe early 1990s and the second for the early2000s. The variables related to technologicalachievement and their sources are reported intable A2.1 and those related to technologicalabsorptive capacity are reported in table A2.2.


95

Table A2.1 Indicators used to calculate the summary indexes and overall index related to technological achievement

Scientific innovation and invention

Scientific and technical journal articles by population World Development IndicatorsPatents granted by the United States Patent and Trademark Office by population Lederman and Saenz 2005Patents granted by the European Patent Office by population Lederman and Saenz 2005

Penetration of older technologies

Electrical Power Consumption kilowatt-hours/capita World Development IndicatorsInternational outgoing telephone traffic percent of GDP per 1,000 people World Development IndicatorsMain lines per 100 inhabitants World Development IndicatorsAir transport, registered carrier departures worldwide percent of GDP per 1,000 people World Development IndicatorsAgricultural machinery: tractors per 100 hectares of arable land World Development IndicatorsExports of manufactures percent of merchandise exports World Development IndicatorsMedium-tech exports percent of total exports CEPII BACI database

Penetration of recent technologies

Internet users per 1,000 people World Development IndicatorsPersonal computers per 1,000 people World Development IndicatorsCellular subscribers per 100 inhabitants World Development IndicatorsPercentage of digital mainlines World Development IndicatorsHigh-tech exports percent of total exports CEPII BACI database

Exposure to external technology

FDI net inflows percentage of GDP World Development IndicatorsRoyalties and license fee payments percent of GDP World Development IndicatorsImports of high-tech goods percent of GDP CEPII BACI databaseImports of capital goods percent of GDP CEPII BACI databaseImports of intermediary goods percent of GDP CEPII BACI database

Source: World Bank. Note: BACI � Banque analytique de commerce internationale, CEPII � Centre d’Etudes Prospectives et d’Informations Interna-tionales, EPO � European Patent Office, FDI � foreign direct investment, GDP � gross domestic product, USPTO � UnitedStates Patent and Trademark Office.

An initial analysis of the two data setsrevealed the existence of two principalcomponents that explained 10 percent or moreof the overall variance and three eigenvaluesthat exceeded 1—a widely used rule of thumbfor determining the underlying dimensionalityof a data set—in each of the data sets beingused (tables A2.3 and A2.4). Bartlett’s test forsphericity confirms that the basic indicators arecorrelated for both indexes, which confirmsthe meaningfulness of applying principal com-

ponents analysis to this data. The Chi-squarestatistics are 1,520.88 (p-value of 0.00) and1,572.10 (p-value of 0.00), respectively. Thosestatistics indicate a strong rejection of the nullhypothesis that variables are not correlated.

Subsequently, as outlined earlier, principalcomponents analysis was performed on ex anteeconomically motivated subgroups of the datafor each summary index. In most cases, the firstprincipal component from these subgroupingsexplained more than 60 percent of the total


96

Table A2.2 Indicators used to calculate the summary indexes and overall indexof technological absorptive capacity

Macroeconomic environmentGeneral government balance as percentage of GDP IMF/WEO and World BankAnnual CPI inflation rate Thomson Datastream and World BankReal exchange rate volatilty J.P. Morgan, IMF and World Bank

Financial structure and intermediationLiquid liabilities percent of GDP Beck, Demirgüç-Kunt, and Levine 2000Private credit percent of GDP Beck, Demirgüç-Kunt, and Levine 2000Financial system deposits percent of GDP Beck, Demirgüç-Kunt, and Levine 2000

Human capitalPrimary educational attainment percent of population aged 15 and over Barro and Lee 2000Secondary educational attainment percent of population aged 15 and over Barro and Lee 2000Tertiary educational attainment percent of population aged 15 and over Barro and Lee 2000

GovernanceVoice and accountability Kaufmann, Kraay, and Mastruzzi 2007Political stability Kaufmann, Kraay, and Mastruzzi 2007Government effectiveness Kaufmann, Kraay, and Mastruzzi 2007Regulatory quality Kaufmann, Kraay, and Mastruzzi 2007Rule of law Kaufmann, Kraay, and Mastruzzi 2007Control of corruption Kaufmann, Kraay, and Mastruzzi 2007

Source: World Bank.

Table A2.3 Share of total variance explainedby principal components, technologicalachievement index

Technological achievement index (2000–03)

Cumulative Share of share of

Component Eigenvalues variance explained variance explained

1 9.91 0.52 0.522 3.59 0.19 0.713 1.49 0.08 0.794 0.92 0.05 0.845 0.85 0.04 0.88� � � �

19 0.00 0.00 1.00

Bartlett’s test: Chi-sq (171) 1520.88 (p-value 0.00)

Source: World Bank.

Table A2.4 Share of total variance explainedby principal components, technologicalabsorptive capacity index

Technological absorptive capacity index (2000–03)

Cumulative Share of share of


1 8.76 0.58 0.582 1.95 0.13 0.713 1.27 0.09 0.804 0.89 0.06 0.865 0.66 0.04 0.90� � � �

15 0.01 0.00 1.00


Source: World Bank.

variance for the subgroup,27 suggesting that itadequately summarized the information in theoverall grouping (table A2.5). Bartlett’s testfor sphericity rejects the null hypothesis of nocorrelation between variables in all cases at the1 percent level.

Indicators of the penetration of old tech-nologies constitute a notable exception. Whenall seven indicators were included, two eigen-values exceeded unity. In addition, while thefirst principal component explained 55 percentof the variance, the second component


97

Table A2.5 Share of total variance explained by principal components for each sub-groupof indicators

CumulativeShare of share of


Penetration of old innovations

1 1.46 0.73 0.732 0.54 0.27 1.00



1 3.43 0.69 0.692 0.80 0.16 0.853 0.54 0.11 0.964 0.22 0.04 1.005 0.01 0.00 1.00


Technological achievement index (2000–03)



Scientific invention and innovation

1 1.49 0.75 0.752 0.51 0.25 1.00


Penetration of recent innovations

1 3.28 0.66 0.662 0.92 0.18 0.843 0.57 0.11 0.954 0.19 0.04 0.995 0.05 0.01 1.00


Source: World Bank.

Technological absorptive capacity index (2000–03)



Human capital

1 1.87 0.62 0.622 0.77 0.26 0.883 0.37 0.12 1.00


Governance

1 5.30 0.88 0.882 0.33 0.05 0.933 0.23 0.04 0.974 0.09 0.02 0.995 0.03 0.01 1.006 0.02 0.00 1.00




Macroeconomic environment

1 2.36 0.79 0.792 0.55 0.18 0.973 0.09 0.03 1.00


Financial structure and intermediation

1 2.69 0.90 0.902 0.30 0.09 0.993 0.02 0.01 1.00


accounted for a non-negligible 22 percent ofthe variance. The group was therefore furtherdivided into two subgroups: industrializationand penetration of other old innovations. Foreach subgroup only the first principal compo-nent presented an eigenvalue greater than 1and accounted for more than 70 percent of thevariance in both instances. Subsequently, an-other stage of principal component analysiswas performed to calculate an aggregatesubindex for the penetration of old innova-tions, and the results are presented in the tableA2.9 on the following page.28 The overall con-clusions obtained do not change substantiallywhen the two groups are entered separatelyinto the overall technological achievement

index (rather than combined into onesubindex).29

As expected, the principal componentsanalysis assigned each subindex unequalweights in the overall index. Table A2.6 sum-marizes the weights assigned to each indica-tor series in the summary subindex oftechnological achievement for each of theeconomic variables. Table A2.7 reports thesame data for the components used to calcu-late the overall index of technologicalabsorptive capacity.

In a second stage, we conducted a princi-pal components analysis on the subindexes.In the case of the technological achievementand the technological absorptive capacityindexes, the first principal component explained79 percent and 65 percent, respectively, of theoverall variance (table A2.8), suggesting thatthe first eigenvector of the correlation matrixprovided a satisfactory summary of the infor-mation included in each of the subindexes.


98

Table A2.6 Factor loadings and variableweights for technological achievementsubgroups (2000–03)

Factor Variable loadings weights (%)

Scientific innovation and development

Patents 0.7071 68.42Scientific and technical articles 0.7071 31.58

Penetration of old innovations

Industrialization 0.7071 67.43Main telephone lines per 100

inhabitants 0.5815 30.82Exports of manufactures 0.5597 42.67Exports of medium-tech goods 0.5904 26.51

Other old innovations 0.7071 32.57Electric power 0.4067 32.66International telephone traffic 0.5137 21.00Air transport 0.5396 20.06Agricultural machinery 0.5288 26.28

Penetration of recent innovations

Internet users 0.5282 25.45Personal computers 0.5214 24.52Cellular subscribers 0.5060 28.25Percentage of digital mainlines 0.1846 6.39Exports of high-tech goods 0.3987 15.40


Net FDI inflows 0.4969 28.55Royalties and license fee payments 0.4384 16.35Imports of high-tech goods 0.4624 17.59Imports of capital goods 0.4914 28.20Imports of intermediary goods 0.3248 9.30

Source: World Bank.Note: FDI � foreign direct investment.

Table A2.7 Factor loadings and variableweights for technological absorptivecapacity subgroups (2000–03)

Factor Variableloadings weights (%)

Human capital

Primary educational attainment 0.4648 24.92Secondary educational attainment 0.6153 39.27Tertiary educational attainment 0.6367 35.82

Governance

Voice and accountability 0.3918 19.14 Political stability 0.3760 15.42 Government effectiveness 0.4224 17.98 Regulatory quality 0.4140 11.85 Rule of law 0.4250 18.39 Control of corruption 0.4179 17.22

Macroeconomic stability

General government balance 0.4990 28.95Annual consumer price index

inflation rate 0.6100 34.40Real exchange rate volatility 0.6156 36.66

Financial structure and intermediation

Liquid liabilities 0.5910 30.79Private credit 0.5424 38.58Financial system deposits 0.5971 30.63

Source: World Bank.

Table A2.9 reports the implicit weights at-tached to each subindex in the two summaryindexes.

A similar process was undertaken, withsimilar results, for the data from the 1990s. Incalculating the percentage changes in eachsubindex and in the overall index, the factorloadings from the 2000s estimation procedurewere used to ensure comparability of the datasets.


99

Table A2.8 Share of total varianceexplained by main principal componentsof technological achievement andtechnological absorptive capacity usingthe sub-indexes (2000–03)

Share of Cumulative share ofComponent Eigenvalues variance explained variance explained

Technological achievement1 3.17 0.79 0.792 0.60 0.15 0.943 0.14 0.04 0.984 0.09 0.02 1.00


Technological absorptive capacity1 2.61 0.65 0.652 0.80 0.20 0.853 0.32 0.08 0.934 0.27 0.07 1.00


Source: World Bank.

Table A2.9 Factor loadings and variable weights obtained from second-stage principalcomponents analysis (2000–03)

Scientific innovation Penetration of old Penetration of Exposure to and invention innovations recent innovations external technology

Technological achievement

Factor loadings 0.5272 0.5404 0.4409 0.4855Subindex weights (%) 21.74 23.99 34.79 19.48

Financial structureHuman capital Governance Macroeconomic environment and intermediation

Technological absorptive capacity

Factor loadings 0.5392 0.5579 0.3493 0.5254Subindex weights (%) 25.29 36.98 10.66 27.06

Source: World Bank.

Notes 1. TFP simply measures all influences on GDP

growth other than increases in capital and labor. Thuschanges in TFP could reflect changes in the composi-tion of output (for example, a shift from agriculture tomanufacturing), changes in the quality of labor or cap-ital not reflected in the data (for example, educationlevels), or any other variable that is an important de-terminant of growth but whose influence is not explic-itly accounted for in growth equations.

2. See http://www.itto.or.jp/live/PageDisplayHandler?pageId=217&id=280.

3. See, for example, Haskel and Slaughter (2002)and Krugman (2000). The rise in the global supply of

goods produced by unskilled and semiskilled labor, andthe influx of low-skilled immigrants, are also cited ascontributing to earnings inequality in high-incomecountries.

4. This is an important conclusion of Global Eco-nomic Prospects 2007. Although intercountry inequal-ity (where each country is accorded equal weight) hasworsened, weighting country observations by popula-tion shows an improvement in income distribution.Taking into account within-country inequality, globalinequality has remained roughly constant since the late1980s.

5. Anecdotal evidence indicates that access tomobile phones improved returns to producers at the

expense of middlemen for fishermen in Porto daManga, Brazil, and Moree, Ghana, and for farmers in awholesale market in Sri Lanka (de Silva and Zainudeen2007). The advent of the Internet and automated tellermachines had a similar effect in the United States, over-coming the anticompetitive effects of state banking reg-ulation and strong lobbying in state legislatures.

6. The focus on patents and scientific publications re-flects academic research on technology. Patents have theadvantage of being more clearly associated withprocesses rather than products (by definition, a patent isnot granted on a product, but rather on the method bywhich it is produced). The disadvantage is that patentsexclude a number of important forms of innovation, no-tably software (until recently) and processes for manag-ing multinational production and distribution networks.

7. The definition of a high-tech export used hereincludes products with high R&D intensity, such asaerospace-related items, computers, pharmaceuticals,scientific instruments, and electrical machinery. Assuch, it excludes a number of services such as softwareengineering that may, by their nature, be even moretechnologically intensive.

8. For example, the correlation coefficient of theshare of high-tech exports in total foreign sales withadult literacy was 43 percent in 2005 and withexpected years of schooling was 22 percent.

9. Regional and income group data in table 2.7 areweighted averages of individual countries, with theweights given by their populations. The simpleaverages cited in the text give equal weight to everycountry independent of the size of its population.

10. The core technology for the Internet can traceits history back to the early 1960s and a networkdeveloped by the U.S. Defense Department’s AdvancedResearch Projects Agency. However, the Internet as it isunderstood by most people today—the World WideWeb and HTML web pages—was first introduced inthe early 1990s, with the first web browser, Mosaic,being released in 1993.

11. According to the Cellular Operators Associa-tion of India, the country had more than 121 millionsubscribers in March 2007.

12. For example, South Africa is beginning to at-tract companies for business process outsourcing. OneBritish executive (Ranger 2006) noted that hisU.K. customers were more comfortable with the SouthAfrican accent than with the Indian accent and citedthe advantages of working at similar times as the par-ent company.

13. A 2005 report from Gartner Inc. stated thatIndia had captured 80 to 90 percent of total offshoreoutsourcing revenue (Tucci 2005).

14. The scores in large countries may be biased,better reflecting the scores in coastal trading cities and

less the score in the interior. Thus the Russian Federa-tion’s weak overall score in comparison with China’smay be explained by the relative absence of large portcities close to major Russian manufacturing centers,whereas in China, port cities and manufacturing cen-ters tend to be in close proximity.

15. Principal components analysis involves exam-ining the eigenvectors of the correlation matrix of aset of related data and extracting from it a weightingscheme that describes as much of the information con-tained within the data set as possible using a mini-mum number of orthogonal linear combinations ofthe original data. By construction, a data set that con-sists of 100 series will have 100 of these eigenvectorsthat fully describe all the information in the data set.However, the first five of these eigenvectors (five dif-ferent linear combinations of the initial 100 series)may describe 90 percent of the total variance. In sucha case, principal components analysis would involvecalculating an overall index based on these fivesubindexes. In a two- or three-stage procedure such asthe one used here, the data are divided into subgroupseither based on the ex ante characteristics of thesubindexes or on the basis of statistical correlations.Then a separate principal components analysis is doneon each of these subindexes, which are subsequentlycombined in a second or third round to determine theoverall index.

16. To minimize the influences of outliers, the com-ponent indicators of the subindexes for each time pe-riod are calculated as the four-year average of valuesfor the period 1990–93 and 2000–03. To maximizecountry coverage, missing data are gap-filled by usingmore recent or older data generally from within theanalytical period. For the transition economies of theformer Soviet bloc, data as recent as 1995 are used forthe early 1990s data point in cases where data do notexist or are unreliable.

17. The relationship between income and techno-logical achievement is complex. The level of income af-fects the ability to gain access to technology, while thelevel of technology helps to determine income levels(see the earlier section on “The role of technology indevelopment”).

18. Low-income countries included in the firstindex are Bangladesh, Benin, Côte d’Ivoire, Ethiopia,Ghana, India, Kenya, the Kyrgyz Republic, Mozam-bique, Nepal, Nigeria, Pakistan, Senegal, Sudan,Tanzania, Togo, Vietnam, the Republic ofYemen, Zambia, and Zimbabwe. The second indexincludes all of these except Nepal, Vietnam, theRepublic of Yemen, and Zimbabwe.

19. The penetration of new technologies during1990–2000 increased by 102 percent in high-incomeOECD countries, 256 percent in upper-middle-income


100

countries, 219 percent in lower-middle-income coun-tries, and 123 percent in low-income countries.

20. For low-income countries, the number of Inter-net users per 1,000 people rose by 172 percent between1999 and 2005, the number of secure Internet serversper 1 million people increased by almost 30 percent ayear between 2004 to 2006, and the number of mobilephone subscribers per 1,000 people rose by 92 percentbetween 1995 and 2005 (World Bank 2007d).

21. Measurements of the extent of diffusion differby technology, but generally involve available statisticson technology flows per country scaled by income orpopulation, whichever is more appropriate. For exam-ple, the diffusion of electricity is measured by kilowatthours consumed per person and the diffusion of rail-roads is measured by tonnage moved divided by grossnational product.

22. Of the 699 country-technology pairs related todeveloping countries, 216 refer to a technology thatwas first recorded in the 19th century, 318 date fromthe first half of the 20th century, 98 come from thethird quarter of the 20th century, and 67 from the finalquarter of the 20th century.

23. The data set includes an estimate of the date ofdiscovery for each technology.

24. That is, there are country-technology pairs, notconsidered in this analysis, that have reached the 5 per-cent level but are taking a long time to reach the25 percent level.

25. When missing data issues occurred over thistime period, the earliest available observation for agiven indicator in the period from 1988–96 and1998–2006 was used so as to expand country cover-age. The budget balance indicators and the real ex-change rate volatility indicators were averaged over theperiods 1990–96 and 2000–06 to purge out cyclicaleffects. The real exchange rate volatility series is theyearly average of the absolute value of monthly changein the real effective exchange rate.

26. For this operation the authors used a combineddata set including observations from both the 1990sand 2000s to ensure that data for each period had thesame underlying scaling.

27. Frequently more than 70 percent of the vari-ance was explained by the first principal component.

28. A data-driven grouping of indicators was alsoconducted and used to generate a two-step index simi-lar to the one reported in the main text. As the indexesderived from this procedure did not differ materiallyfrom the one reported here, they are not reported.

29. Furthermore, results for the overall technologi-cal achievement index remain unchanged when the en-tire old innovations indicator group is included, that is,when no division into subgroups is made and thesecond principal component is ignored.

ReferencesAcemoglu, Daron. 2002. “Technical Change, Inequality,

and the Labor Market.” Journal of EconomicLiterature 40: 7–72.

Alderman, Harold, John Hoddinott, and Bill H.Kinsey. 2006. “Long Term Consequences of EarlyChildhood Malnutrition.” Oxford EconomicPapers 58 (3): 450–74.

Arbache, Jorge Saba, Andy Dickerson, and FrancisGreen. 2004. “Trade Liberalization and Wages inDeveloping Countries.” Economic Journal 114:F73–F96.

Archibugi, Daniele, and Alberto Coco. 2005. “Mea-suring Technological Capabilities at the CountryLevel: A Survey and a Menu for Choice.”Research Policy 34: 175–94.

Arvis, Jean-François, Monica Alina Mustra, JohnPanzer, Lauri Ojala, and Tapia Naula. 2007. Con-necting to Compete: Trade Logistics in the GlobalEconomy. Washington, DC: World Bank.

Bailey, Martha J. 2006. “More Power to the Pill: TheImpact of Contraceptive Freedom on Women’sLabor Supply.” Quarterly Journal of Economics121 (1): 289–320.

Bankable Frontier Associates. 2007. “Financial ServiceAccess and Usage in Southern and East Africa:What Do FINSCOPE™ Surveys Tell Us?”Bankable Frontier Associates. http://www.bankablefrontier.com. Accessed October 2007.

Behrman, Jere, and Mark Rosenzweig. 2004. “Returnsto Birthweight.” Review of Economics and Statis-tics 86 (2): 586–601.

Besant-Jones, John. 2006. “Reforming Power Markets inDeveloping Countries: What Have We Learned?”Energy and Mining Sector Board Discussion Paper19. World Bank, Washington, DC.

Briceno-Garmendia, Cecilia, Antonio Estache, andNemat Shafik. 2004. “Infrastructure Services inDeveloping Countries: Access, Quality, Costs andPolicy Reform.” Staff Working Paper 3468.World Bank, Washington, DC.

Chandra, Vandana. ed. 2006. Technology, Adaptationand Exports: How Some Developing CountriesGot It Right. Washington, DC: World Bank.

Chandra, Vandana, and Shashi Kolavalli. 2006.“Technology, Adaptation and Exports: HowSome Developing Countries Got It Right.” InTechnology, Adaptation and Exports: How SomeDeveloping Countries Got It Right, ed. VandanaChandra, 1–48. Washington, DC: World Bank.

Chinn, Menzie D., and Hiro Ito. 2006. “What Mattersfor Financial Development? Capital Controls,Institutions and Interactions.” Journal of Devel-opment Economics 81 (1): 163–92.


101

Comin, Diego, and Bart Hobijn. 2004. “Cross-Country Technology Adoption: Making theTheories Face the Facts.” Journal of MonetaryEconomics 51 (1): 39–83.

de Silva, Harsha, and Ayesha Zainudeen. 2007.“Teleuse on a Shoestring: Poverty Reductionthrough Telecom Cccess at the ‘Bottom of thePyramid’.” Paper prepared for the Centre forPoverty Analysis Annual Symposium on PovertyResearch in Sri Lanka, Colombo, December 6–7.http://www.lirneasia.net.

Dutz, Mark A., ed. 2007. Unleashing India’s Innova-tion: Toward Sustainable and Inclusive Growth.Washington, DC.

Friedman, Thomas. 2007. “If IT Merged with ET.”International Herald Tribune, October 31.

Glewwe, Paul, Hanan G. Jacoby, and Elizabeth M.King. 2001. “Early Childhood Nutrition and Aca-demic Achievement: A Longitudinal Analysis.”Journal of Public Economics 81 (3:) 345–68.

Government of India. 2006. E-Readiness AssessmentReport of States/Union Territories 2005. NewDelhi: Ministry of Communication and Informa-tion Technology.

Gwatkin, Davidson R., Shea Rustein, Kiersten John-son, Eldaw Suliman, Adam Wagstaff, and AgbessiAmouzou. 2000a. Socio-economic Differences inBrazil. Washington, DC: World Bank. http://www.worldbank.org/poverty/health/data/index.htm#lcr.

______. 2000b. Socio-economic Differences in Ethiopia.Health, Nutrition, and Population in Ethiopia.Washington, DC: World Bank.http://worldbank.org/poverty/health/data/index.htm#lcr.

______. 2000c. Socio-economic differences in Nigeria.Health, Nutrition, and Population in Nigeria.Washington, DC: World Bank. http://worldbank.org/poverty/health/data/index.htm#lcr.

Hanushek, Eric A., and Ludger Woessmann. 2007.“The Role of Education Quality in EconomicGrowth.” Working Paper 4122. World Bank,Washington, DC.

Haskel, Jonathan E., and Matthew J. Slaughter. 2002.“Does the Sector Bias of Skill-Biased TechnicalChange Explain Changing Skill Premia?” Euro-pean Economic Review 46: 1757–83.

Howitt, Peter, and David Mayer-Foulkes. 2005.“R&D, Implementation, and Stagnation:A Schumpeterian Theory of Convergence Clubs.”Journal of Money, Credit, and Banking 37 (1).

Jefferson, Gary H., Thomas G. Rawski, and YifanZhang. 2007. “Productivity Growth and Conver-gence Across China’s Industrial Economy.” Draft.Brandeis University, Waltham, Mass.

Kaufmann, Daniel, Frannie Leautier, and MassimoMastruzzi. 2005. “Governance and the City:An Empirical Exploration into Global Deter-minants of Urban Performance.” PolicyResearch Working Paper 3712. World Bank,Washington, DC.

Kearney, A. T. 2007. “Labor Cost Advantages of‘Offshore’ Locations Decline in 2006.” A. T.Kearney. http://www.atkearney.com. AccessedMarch 2007.

Klenow, Peter J., and Andrés Rodriguez-Clare. 2004.“Externalities and Growth.” Working PapersSeries 11009. National Bureau of Economic Re-search, Cambridge, MA. http://www.nber.org/papers/w11009.

Krugman, Paul. 2000. “Technology, Trade and FactorPrices.” Journal of International Economics 50:51–71.

Lederman, Daniel, and William F. Maloney, eds. 2007.Natural Resources, Neither Curse nor Destiny.Palo Alto, CA: Stanford University Press.

Lederman, Daniel, and Laura Saenz. 2005. “Innova-tion and Development Around the World,1960–2000.” Policy Research Working Paper3774. World Bank, Washington, DC.

Lumenga-Neso, A., M. Olarreaga, and M. Schiff.2005. “On ‘Indirect’ Trade-Related R&DSpillovers and Growth.” European EconomicReview 49 (7): 1785–98.

Maloney, William. 2006. “Missed Opportunities: In-novation and Resource-Based Growth in LatinAmerica.” Unpublished report. World Bank,Washington, DC.

Martin, W., and D. Mitra. 2001. “Productivity Growthand Convergence in Agriculture and Manufactur-ing.” Economic Development and CulturalChange 49 (2): 403–23.

McKenzie, D. J. 2003 “Measure Inequality with AssetIndicators.” Working Paper 042. Cambridge,MA: Harvard University, Center for InternationalDevelopment, Bureau for Research and EconomicAnalysis of Development.

Miller, Grant. 2005. “Contraception as Development?New Evidence from Family Planning inColombia.” Working Paper 11704. NationalBureau of Economic Research, Cambridge, MA.

Mitra, Raja M. 2007. “India’s Emergence as a GlobalR&D Center.” Working Paper R2007:012,Swedish Institute for Growth Policy Studies,Stockholm.

Montgomery, M. R., K. Gragnolati, A. Burke, andE. Paredes. 2000. “Measuring Living Standardswith Proxy Variables.” Demography 37: 155–74.


102

Nicoletti, Giuseppe, Stefano Scarpetta, and OlivierBoylaud. 1999. “Summary Indicators of ProductMarket Regulation with an Extension to Employ-ment Protection Legislation.” Economics Depart-ment Working Papers 226. Organisation forEconomic Co-operation and Development, Paris.

Poncet, Sandra. 2006. “The Long Term GrowthProspects of the World Economy: Horizon2050.” Working Paper 2006–16. Centre d’EtudesProspectives et d’Informations Internationales.

Porter, M., and S. Stern. 2003. “Ranking NationalInnovative Capacity: Findings from the NationalInnovative Capacity Index.” In Global Competi-tiveness Report, World Economic Forum, Geneva.

Ranger, Steve. 2006. “South Africa Touts Its OffshoreCharms.” http://www.silicon.com. AccessedDecember 12, 2006.

Sagasti, Francisco. 2003. Knowledge and Innovationfor Development: The Sisyphus Challenge for the21st Century. Cheltenham, U.K.: Edward Elgar.

Schultz, T. Paul. 2007. “Population Policies, Fertility,Women’s Human Capital, and Child Quality.”Discussion Paper 954. Yale Economic GrowthCenter, New Haven, CT.

Soubattina, Tatanya. 2006. “Generic Models of Tech-nological Learning by Developing Countries.”Draft. World Bank, Washington, DC.

Sricharoen, Thitiwan, and Gertrud Buchenrieder.2005. Principal Component Analysis of Povertyin Northern Thailand. Paper presented at theConference on International Agricultural Re-search for Development, Stuttgart-Hohenheim,Germany, October 11–13.

Stelling, Petra, and Arne Jensen. 2001. “Train Operators’Economies of Scale and Business Strategies.” Work-ing paper. Handelshogskolan, Gotteborg, Sweden.

Sullivan, Nicholas P. 2007. You Can Hear Me Now:How Microloans and Cell Phones Are Connect-ing the World’s Poor to the Global Economy.San Francisco: Jossey-Bass.

Tucci, Linda. 2005. “Offshore Comparison Shopping:Best Buys Abroad.” http://www.searchcio.com.Accessed September 28. 2005.

UNIDO (United Nations Industrial Development Or-ganization). 2002. Industrial Development Re-port 2002/2003: Competing through Innovationand Learning. Vienna: UNIDO.

______. 2005. Industrial Development Report 2005—Capability Building for Catching-Up: Historical,Empirical and Policy Dimensions. Vienna:UNIDO.

UNCTAD (United Nations Conference on Tradeand Development). 2005 World Investment Re-port: Transnational Corporations and the Indus-trialization of R&D. New York and Geneva:UNCTAD.

______. 2007. The Least Developed Countries Report2007: Knowledge, Technological Learning andInnovation for Development. New York andGeneva, UNCTAD.

UNDP (United Nations Development Programme).2001. Human Development Report 2001:Making New Technologies Work for HumanDevelopment. New York: Oxford UniversityPress.

Vyas, Seema, and Lilani Kumaranayake. 2006. “Con-structing Socio-economic Status Indices: How toUse Principal Components Analysis.” HealthPolicy and Planning 21 (6): 459–68.

Watkins, Alfred and Anubha Verma. 2007. “Govern-ment of Rwanda-World Bank, Science, Technol-ogy and Innovation Capacity Building AssistanceProgram: Developing Practical Solutions to Prac-tical Problems.” Internal report. World Bank,Washington, DC.

World Bank. 1998. World Development Report:Knowledge for Development. New York: OxfordUniversity Press.

______. 2005. Sri Lanka—Offshoring ProfessionalServices: A Development Opportunity. Washing-ton, DC: World Bank.

______. 2006. Transport for Development. Washing-ton, DC: World Bank.

______. 2007a. World Development Indicators 2007.Washington, DC: World Bank.

______. 2007b. World Development Report 2008:Agriculture for Development. New York: OxfordUniversity Press.

______. 2007c. Making Finance Work for Africa.Washington, DC: World Bank.

Zhu, Susan Chun, and Daniel Trefler. 2005. “Tradeand Inequality in Developing Countries: A Gen-eral Equilibrium Analysis.” Journal of Interna-tional Economics. 65: 21–48.


103

105

Determinants of TechnologicalProgress: Recent Trends andProspects

3

As discussed in the previous chapter, the paceat which technologies spread between andwithin countries has picked up. As a result,most developing countries are narrowing thetechnological divide that separates them fromhigh-income countries. Nevertheless, the tech-nology gap remains large; for many, includingseveral low-income countries, it is wideningrather than closing, in part because of theslowness with which technologies spreadwithin countries. For virtually all developingcountries, the domestic pace of technologicalprogress is determined mainly by the speedwith which already existing technologies areadopted, adapted, and successfully applieddomestically, and done so throughout theeconomy, not just in the main cities.

This chapter explores some of the majordeterminants of this kind of within-countrydiffusion of technology. It adopts an analyti-cal framework that distinguishes between thefactors that dictate the extent to which aneconomy is exposed to external technologieson the one hand and the efficiency with whichit absorbs them on the other hand. Among themost important channels through which low-and middle-income countries are exposed toforeign technologies are trade; foreign directinvestment (FDI); and contacts with highlyskilled diaspora members (nationals workingabroad) and with other information net-works, including those of academia and themedia. Maintaining an open environment tosuch flows is critical for accessing technology

at least cost. However, no matter how com-pellingly useful a technology may be, theprocess by which it spreads within a countrycan be lengthy.

The speed with which a country absorbsand adopts technology depends on many fac-tors, including the extent to which a countryhas a technologically literate workforce and ahighly skilled elite; promotes an investmentclimate that encourages investment and per-mits the creation and expansion of firmsusing higher-technology processes; permitsaccess to capital; and has adequate public sec-tor institutions to promote the diffusion ofcritical technologies where private demand ormarket forces are inadequate.

The process of technology absorption is alsosubject to virtuous circles. Scale economiesin technologically sophisticated sectors and inlearning by doing tend to make the acquisitionof technology a nonlinear process, character-ized initially by slow penetration until somethreshold is reached, followed by a period ofrapid acceleration, and finally by a period ofslower diffusion as saturation is achieved. Asa consequence, while gaps in technologicalachievement create opportunities for acceler-ated growth and convergence in laggingeconomies, they can also lead to divergence ifthe conditions for technology adoption inlagging countries are insufficient. Many tech-nologies operate synergistically to reinforcethe demand for each other and the effective-ness and capacity of supply.

Although the process of technological dif-fusion has a clear logic, the process is by nomeans mechanical. It occurs through interac-tions among individuals, entrepreneurs, firms,and governments. The role of government isboth direct, as a supplier of many technologi-cal services, and indirect. In particular, theefficiency with which firms can diffuse tech-nology within the domestic economy dependson the overall political and economic context,the level and distribution of human capital,the quality of the macroeconomic environ-ment, and the rules and regulations governingthe conduct of business, all of which are heav-ily influenced by governments.

This chapter discusses the principal chan-nels through which developing countries areexposed to advanced technology, analyzes themain determinants of domestic absorptivecapacity, and indicates likely future trends intechnological diffusion. The following sixmain messages emerge from this analysis.

The principal channels by whichdeveloping countries are exposed toexternal technology—which includetrade, FDI, and a highly skilled dias-pora—have increased substantially overthe past several decades. The share ofimported high-tech products in grossdomestic product (GDP) has risen bymore than half in both low- and middle-income regions since the mid-1990s,that of imports of capital goods by 37percent, that of imports of intermediategoods by 26 percent, and that of FDI in-flows by sixfold since the 1980s. Finally,the size and sophistication of global di-asporas has increased markedly, alongwith substantial improvements in thetechnology by which migrants can trans-mit their know-how and interact withtheir home economies.

The ability to absorb foreign technol-ogy, which depends on domestic policiesand institutions, has also improved inmany developing countries. Reflectingrising school enrollment, literacy rates inlow-income countries have increased

from less than 50 percent in 1990 tomore than 62 percent, and among youththey now exceed 74 percent. In addition,the macroeconomic, governance, and in-vestment climate that innovative firmsand entrepreneurs need to operate is im-proving. More countries are operatingin a context of close to stable prices (me-dian inflation in low-income countriesdeclined from 9.2 to 4.2 percent be-tween 1990 and 2006) and flexible ex-change rate regimes, while governmentfinances are better balanced.

Technological diffusion amongmiddle-income countries has benefitedfrom the reorientation of global produc-tion processes. Advances in communica-tions and transport technology havegiven rise to the growth of global pro-duction networks, facilitating increasedtrade and technological advances inmany developing countries, particularlymiddle-income developing countries.Until recently, low wages and a solid, iflow, level of basic technological literacyhave been sufficient to capture a signifi-cant role in global networks in manycountries. As wages rise, however, thesecountries will need to make substantialadditional investments in human capitalto maintain their share of global produc-tion and continue the technological con-vergence of the past few years. They willalso need to adopt a more proactive ap-proach to developing local competenciesand to using research and development(R&D) and outreach programs to bol-ster the diffusion process.

For low-income countries, poor tech-nological adaptive capacity and limiteddissemination of often simple technolo-gies to the countryside are severely con-straining technological progress. Despiteprogress in basic technological literacy,extremely low levels of income, weakgovernance structures, and, in somecases, ongoing conflict continue tostymie the ability of low-income, and


106

especially Sub-Saharan African, coun-tries to obtain and absorb new tech-nologies. Nevertheless, the potential fortechnological progress through thegreater dissemination of relatively sim-ple technologies is huge.

The absence or low quality of somebasic technologies that governments his-torically provide hinders technologicaldiffusion by the private sector. Thesebasic technologies often represent essen-tial complementary technologies whoseabsence can prevent the successful adop-tion of a new-to-the-market technology.

The relatively rapid disseminationof new communications technologiesthroughout the developing world, in-cluding in low-income countries, offersa ray of hope. These potentially trans-formational technologies are enabling,often for the first time, the kinds ofarm’s-length transactions that may becritical to firm development and thespread of technology in these countries.New technologies are frequently intro-duced and promoted by members ofnational diasporas, both directly throughnetworks and indirectly through invest-ments financed from remittances.

The rest of this chapter is devoted toexploring how developing countries absorbexternal technology. The next section presentsan analytical framework for technologicalprogress. This is followed by a discussion ofhow trade, FDI, and migration expose coun-tries to new technologies and can promote in-ternal diffusion of those technologies. The sec-tion also discusses trends in these flows andthe potential magnitude of associated techno-logical progress and how this may havechanged over time. The chapter then turns toan analysis of the domestic factors that facili-tate the absorption of new technologies. Thissection first examines how government poli-cies, and the business environment in general,facilitate the creation and expansion of innova-tive firms. It then looks at how levels of humancapital—from literacy rates to R&D capacity—

affect countries’ ability to absorb technologyfrom abroad. The chapter concludes with aspeculative view of the prospects for technolog-ical progress and some policy messages.

Drivers of technologicalprogress: A framework

The process by which countries adopt,adapt, and absorb external technologies is

complicated. The overall framework followedin this chapter (depicted in figure 3.1) draws onprevious work done at the World Bank, in par-ticular, the 1998 World Development Reporton the knowledge economy (World Bank 1998)and several regional and country-specific policyanalyses of technology and technological com-petencies. In addition, it relies upon the acade-mic literature on technology diffusion, includ-ing an excellent review article by Keller (2004);several articles by Coe and Helpman concern-ing the role of FDI; case studies of the processof technology diffusion by Chandra andKolavalli (2006); empirical work on the tech-nological influence of imports by Lumenga-Neso, Olarreaga, and Schiff (2005); and thediscussion by Rodrik (2004) on the impact ofmarket failures on innovation incentives.

Exposure to external flows interacts with domestic capacity to diffusetechnologyFor the purposes of analytical simplicity, theframework presents technological progressin developing countries as a process wherebyan economy is exposed to higher-technologybusiness processes, products, and servicesthrough foreign trade; FDI; and contacts withits diaspora and other communication chan-nels, including academic and international or-ganizations (the large arrows at the top of thefigure). Exposure to new ideas and techniquesis, however, not sufficient to ensure techno-logical progress on the ground. The extent towhich these flows are translated into techno-logical progress depends on the technical ab-sorptive capacity of the economy (representedby the ringed drum). This in turn depends on

D E T E R M I N A N T S O F T E C H N O L O G I C A L P R O G R E S S

107

the extent to which the business and macro-economic climate fosters an environment inwhich firms—the main mechanism for tech-nological diffusion within a country—are ableto form, grow, and expand. Absorptive capac-ity also depends on the levels of basic techno-logical literacy and advanced skills found inthe country, which together dictate the coun-try’s capacity to implement technologies onthe one hand and to do the research necessaryto understand, implement, and adjust im-ported technologies on the other hand. Alsoimportant are government actions designed tohelp overcome market failures that mightlimit the financing of innovative activity, plusactions that focus technology policy on adapt-ing and adopting those existing technologiesfor which there is a market and for which ad-equate domestic competencies exist. Criticalhere are outreach and dissemination policies,

which need to serve as a two-way conduit,both informing the population about techno-logical solutions and providing feedback toproviders concerning the usability of and de-mand for proposed solutions. Taken together,these factors act as filters (the rings in thedrum) that dictate how much of the potentialtechnological flow is actually absorbeddomestically.

The overall process is, of course, morecomplicated, with both technological flowsand technological adaptive capacity influenc-ing each other. For example, internationaltrade is perhaps the most important vector forthe transmission of technology, but the extentof a country’s openness to trade depends sig-nificantly on the amount of FDI that has oc-curred, the existence of a vibrant and techno-logically literate diaspora, and the domesticbusiness climate. Similarly, the quantity of FDI


108

Returns toscale

Spillovereffects

Pro-active policies

Finance of innovative firms

Basic technological literacy

Governance and the business climate

Source: World Bank.

Figure 3.1 Domestic absorptive capacity both conditions and attracts external flows

Technological frontier

Domestic technological achievement

Tech

nolo

gica

lab

sorp

tion

Dynamiceffects

magnifytechnology

transfer

Diasporaand othernetworks

FDITrade

Tran

smis

sion

chan

nels

Tech

nolo

gica

l abs

orpt

ive

capa

city

Policies to - create competencies - build infrastructure - foster an innovation- friendly business climate

and its overall effectiveness depends on thequality and technological literacy of the laborforce.

Increasing returns and spillover effectscan magnify these effects . . .Both domestic and external determinants oftechnological transfer are affected to varyingdegrees by increasing returns to scale andspillover effects that can magnify the absorp-tive impact of these flows. Access to foreignmarkets may allow domestic firms to grow andexploit economies of scale associated withsome technologies, raising the overall wealthand technological sophistication of an econ-omy. Meanwhile, the technological spilloversthat can be expected from FDI, includingdemonstration effects and the transfer of busi-ness process and human capital to domesticfirms through employee turnover, are likely tobe greater the more qualified is the labor force.Both FDI and trade can contribute to clustereffects and networking externalities that in-crease the potential for spillovers and to tech-nological diffusion from individual sectors andfirms to the rest of the economy. Alternatively,economies of scale and agglomeration effectsmay prevent entry by new firms in some mar-kets, cutting off otherwise promising opportu-nities for technological learning. In addition,imitators may limit entrepreneurs’ ability tocapture the returns to new-to-the-marketinnovations, thereby reducing incentives fortechnological progress.

. . . but a lack of financing can stymieinnovation Affordability issues can influence both the sizeof initial inflows and a country’s technologicalabsorptive capacity. Even if profitable invest-ments in technology are available and the do-mestic environment encourages the absorptionof new technologies, low incomes may makenew technologies unaffordable to individualsand firms in developing countries. At the ex-treme, individuals near subsistence levels maybe unwilling to risk adopting a new-to-the-market technique, may be unable to generate

adequate savings to invest in a new technol-ogy, or may lack the collateral required to bor-row. Thus poverty is a major cause, as well asa result, of low levels of technology.

Affordability is also an issue at the macrolevel. Low incomes constrain government fi-nance, limiting the government’s capacity toput into place both the level of physical infra-structure and the investments necessary todevelop a level of domestic human capitalcapable of supporting and exploiting evensimple technologies.

Policy should not impede innovative firmsFinally, firms, entrepreneurship, and govern-ment action that actively supports the diffu-sion of economically relevant and profitabletechnologies are the grassroots mechanisms bywhich technologies diffuse within countries.Firms must be able, and entrepreneurs permit-ted, to profit from the exploitation of new-to-the-market-technologies if those technologiesand products are to diffuse. This means thatpolicy must be welcoming of such profits andthat both R&D and dissemination efforts notonly need to focus on creating or adaptingproducts and ideas (domestic or foreign) tothe local market, but also must give priority toassisting firms to exploit them.

External transmission channels

This section presents data and describes re-cent trends concerning the external chan-

nels through which developing countries areexposed to foreign technologies. Where rele-vant, it also draws on the literature to com-ment on how specific elements in a country’stechnological absorptive capacity (discussed inmore detail later in this chapter) interact withthese external flows to determine the extent towhich these channels translate into technolog-ical achievement.

TradeTrade is one of the most important mechanismsby which embodied technological knowledge(in the form of both capital and intermediate


109

goods and services) is transferred across coun-tries. Imports of technologically sophisticatedgoods help developing countries raise the qual-ity of their own products and the efficiencywith which they are produced. Countries canalso absorb new technology by exporting tocustomers who implicitly or explicitly provideguidance in meeting the specifications requiredfor access to global markets. For developingcountries with low R&D intensity, trade open-ness and exposure to foreign competitionprovide powerful inducements to adopt moreadvanced technology in both exporting andimport-competing firms and are likely to pro-duce large technology spillovers and produc-tivity gains (Schiff and Wang 2006a).

However, the extent to which exposure toforeign technologies is reflected in the exportand import patterns of individual countriesdepends on, among other things, the absorp-tive capacity of individual countries. As Soub-botina (2006) discusses, countries with rela-tively weak domestic scientific capacities tendto follow a more passive approach to technol-ogy absorption that is characterized by limitedefforts to leverage the technology imported byforeign firms operating on their soil. For thesecountries, most technology transfers takeplace either through imports of high-techgoods, or perhaps through an apparentlyhigh-tech export sector that is, in reality, dom-inated by assembly operations associated withelevated imports of high-tech goods. Wheresophisticated domestic capacities coexist witha significant degree of basic technological lit-eracy in the population, technology diffusionis enhanced and, in general, the technologicalcontent of exports is higher.

Following Soubbotina (2006) classifica-tions, “traditionalist slow learners,” such asBangladesh and Burkina Faso, which have lowlevels of technological competency and tech-nological literacy, tend to rely to a large extenton imports of machinery and equipment.Other countries, such as Malaysia, Mexico,and the Philippines, appear to follow a “pas-sive FDI-dependent” learning style. For thesecountries, the share of high- and medium-tech

goods in their manufactured exports is higherthan these goods’ share in manufacturingvalue added, reflecting the dependence oftheir high-tech exports on imports of techno-logically sophisticated components and therelatively low technological complexity ofdomestic manufacturing operations. By con-trast, “active FDI-dependent” countries, suchas Chile and Hungary, strike a better balancebetween the share of high-tech goods and ser-vices in overall exports and domestic valueadded, reflecting greater domestic technologi-cal competencies. For the Russian Federationand some of the other countries that belongedto the former Soviet Union, a strong techno-logical base and relatively low import sharesof high-tech goods reflect their more advancedlearning style, which places greater emphasison domestically developed technologies.Nevertheless, these technologies mainly feedinto products that serve the local market,because both high costs and quality concernskeep these sectors from being internationallycompetitive.

The potential for technology transferthrough imports has risenImports improve domestic technology becauseembodied technology both allows firms to em-ploy more efficient production processes andaffords the possibility that firms can copymore advanced products and processes. At thesame time, competition from technologicallysuperior imports may boost domestic produc-tivity.1 Developing countries that have a largeshare of imports from high-income countrieswith large R&D expenditures have signifi-cantly higher productivity than developingcountries that import from advanced coun-tries with lower R&D expenditures (Coe,Helpman, and Hoffmaister 1997).2 There alsois evidence for a positive relationship betweenaccess to imported intermediate goods andperformance (Handoussa, Nishimizu, andPage 1986). More recent literature highlightsthe indirect benefits for developing countriesfrom North–North trade in R&D. The ex-change of high-tech goods and services among


110

high-income economies contributes to an in-crease in the global stock of knowledge andeventually becomes available to developingcountries through North–South trade. Finally,technology diffusion, like trade, tends to be re-gional, with the largest transfers coming fromnatural trading partners, for example, Jordanbenefits more from the European Union andMexico benefits from Canada and the UnitedStates (Schiff and Wang 2006b).

However, the extent to which importedtechnology boosts the sophistication of do-mestic technological activity either directly orindirectly through spillovers depends on thequality of a country’s technological absorptivecapacity. Thus while using an imported capitalgood can lift the technological content of ac-tivity in a country, to the extent that importerspay competitive prices for the technology,there may be no net gain to the country (Eatonand Kortum 2001). Moreover, the businessclimate may be too weak or the technologicalliteracy of the local labor force may be toolow to successfully adapt the machinery tolocal conditions (Dahlman, Ross-Larson, andWestphal 1987; Rosenberg 1976). As a result,the country may not realize the potential pro-ductivity improvements available from im-ported technology (Pack 2006).

Developing countries’ high-tech importshave increasedTo the extent that a developing country canmake use of or imitate sophisticated goods, itslevel of technological achievement should in-crease in line with the quality and technologicalsophistication of imported goods. Since the mid-1990s, the share of imported high-tech productsinGDPhas increasedbymore than50 percent inlow-income countries and by 70 percent in mid-dle-income countries (table 3.1).3 Among devel-oping regions, East Asia and the Pacific hasthe highest share of high-tech imports in GDP(8.4 percent), with the highest share being forMalaysia (37 percent) and the Philippines(18 percent), but Europe and Central Asia hasexperienced the largest increase, reflecting thetransition of many of the region’s countries tomarket economies and their improved access tohigh-tech products following the relaxation ofCold War export restrictions. Among regionsdominated by middle-income countries, high-tech imports represent 3.8 percent of GDP inLatin America and the Caribbean and 3.6 per-centofGDPintheMiddleEastandNorthAfrica,less than in Sub-Saharan Africa (4.5 percent).

Low-income countries have also improvedtheir exposure to high-technology embeddedin foreign products. After hovering around


111

Table 3.1 Trade in technology goods has increased in developing countries

Share of high-tech exports Imports of high-tech goods Imports of capital goods in world high-tech exports

1994–96 2002–04 % change 1994–96 2002–04 % change 1994–96 2002–04 % change

(% of GDP) (% of GDP) (% of GDP)

RegionsEast Asia and the Pacific 5.9 8.4 42 11.6 12.8 10 9.9 19.0 93Europe and Central Asia 3.2 7.2 125 7.1 14.7 107 1.0 2.7 163Latin America

and the Caribbean 2.4 3.8 61 5.4 7.2 32 2.1 3.4 61Middle East and North Africa 2.5 3.6 44 6.3 8.9 42 0.1 0.2 29South Asia 1.4 2.1 53 3.1 3.8 22 0.2 0.3 58Sub-Saharan Africa 3.2 4.5 39 9.3 10.5 14 0.1 0.1 4

Income groupsHigh-income countries 3.4 4.7 38 5.5 7.0 27 86.5 74.3 �14Upper-middle-income countries 4.2 7.2 71 8.7 13.1 51 6.6 9.6 47Lower-middle-income countries 3.2 5.4 70 6.9 9.2 33 6.7 15.7 137Low-income countries 1.8 2.7 53 4.9 5.7 17 0.3 0.4 53

Source: World Bank calculations using Centre d’Etudes Prospectives et d’Informations Internationales’ database, Banque Analytiquede Commerce International.

1.8 percent between 1994 and 2001, the aver-age share of high-tech imports in low-incomecountries’ GDP began to rise in 2002, reaching3.2 percent in 2004 (figure 3.2). Both SouthAsia and Sub-Saharan Africa have enjoyed sig-nificant increases, although the ratio of high-tech imports to GDP remains extremely low inSouth Asia, less than 3 percent of GDP. In mostcountries in Sub-Saharan Africa, the share ofimported high-tech goods fluctuates between2 and 5 percent of GDP from year to year.Mauritius and South Africa import the mosthigh-tech goods relative to the size of theireconomies, between 6 and 8 percent of GDP inany given year, while Somalia imports theleast, less than 1 percent of GDP.

Despite developing countries’ increased ex-posure to foreign technology through trade, itsdistribution across regions within countriestends to be extremely uneven, with foreigntrade concentrated in a few major cities or re-gions. For example, 70 percent of high-techtrade (both imports and exports) in China orig-inates in four regions and is highly correlatedwith R&D intensity and foreign firms (OECD

2007). As a result, the benefits of exposure totrade also tend to be unevenly distributed.

Capital goods imports have also increasedAlthough imports of high-tech goods providean indication of an economy’s exposure totechnology, this indicator does not distinguishbetween imports of technology for consump-tion and imports for production, nor does itindicate the extent to which these imports im-prove the technological content of a country’seconomic activities. Technological contentdepends importantly on the structure of theeconomy and the nature (assembly or highvalued added transformation) of the workdone with the imports (box 3.1).

Imports of capital goods, such as machineryand equipment, which enable the production ofhigher quality and more technologically sophis-ticated goods, have a less ambiguous impact ona country’s technological capacity. For coun-tries operating within the technological fron-tier, a higher share of imported capital goodsin GDP can reflect the presence of strong in-vestment activity; a process of technological


112

Figure 3.2 Rising share of high-tech imports

Percent of GDP

Source: World Bank calculations using Centre d’Etudes Prospectives et d’Informations Internationales database, Banque Analytique deCommerce International.

0

2

12

4

10

8

6

1995

1996

1997

1998

1999

2000

2001

2002

2003

2004

By region

1994

Percent of GDP

0

1

9

2

6

4

3

1995

1996

1997

1998

1999

2000

2001

2002

2003

2004

8

7

5

By income group

1994


Low-income countries

Lower-middle-income countries



Europe and Central Asia Sub-Saharan Africa



South Asia

upgrading; and, over the longer term, arelatively sophisticated structure of produc-tion.4 As a result, relatively technologically so-phisticated middle-income countries importmore capital goods (as a share of GDP) than lesssophisticated low-income countries (table 3.1).

Overall, the share of capital goods in theGDP of developing countries has risen substan-tially over the past decade. Upper-middle-income countries saw a 51 percent increase inthe share of these goods in GDP, while lower-middle-income countries have boosted their re-liance on such goods by 33 percent. The formergroup of countries continues to have the largestshare of capital goods in GDP, about 13 per-cent, more than double the share of low-income countries. As a consequence, the gapbetween low-income countries (especially theLeast Developed Countries [UNCTAD 2007])and other developing countries has widened.Europe and Central Asia saw the biggest re-gional increase in imports of capital goods,reflecting the substantial economic recovery inthese countries following the recession that ac-companied the transition to market economies.As was the case for high-tech imports, SouthAsia has the lowest level of imports of capitalgoods and has shown little improvement,

presumably reflecting the relatively autarchicpolicies that governments in the region havefollowed until recently. By contrast, Sub-Saha-ran Africa imports substantially more capitalgoods, although the ratio of capital goods im-ports to GDP has increased less than in otherdeveloping regions since the mid-1990s, withthe exception of East Asia and the Pacific. Thestory is particularly varied in Latin America,where some countries, such as Costa Rica andMexico, increased their imports of capitalgoods following liberalization policies in theearly 1990s, while others did not. The increasein the share of capital goods in GDP was par-ticularly notable in Costa Rica, where it rosefrom about 7 percent in the mid-1990s to about18 percent in 2002–04.

Exports of technological goods have also expandedParticipation in high-tech export markets hasalso been identified as a channel through whichtechnology is diffused within developing coun-tries. Many case studies suggest that exportingfirms in developing countries benefit from im-plicit and explicit technological transfers thatoccur as a result of their interactions with for-eign buyers. Benefits accrue because foreign


113

Based on a breakdown of trade flows by technol-ogy level and by production stage, Lemoine and

Ünal-Kesenci (2003) highlight the following divergentintegration paths adopted by China, India, and Turkey,which all started from a relatively similar degree ofindustrial specialization about 10 years ago:

• China has become an assembly country,strongly integrated into the international seg-mentation-of-production processes in Asia.Most of China’s imports of high-tech productsare parts and components. These high-tech im-ports are predominantly incorporated into theproduction of exports and are not used to mod-ernize domestic production capacities. Given its

Box 3.1 Technology imports: Different paths fordifferent countries

level of development, China’s exports display anoutstandingly strong high-tech content.

• India is characterized by limited participation ininternational division-of-production processesand by a low level of imports in high-tech prod-ucts. These high-tech imports are evenly distrib-uted among the different stages of productionand the different sectors, while high-tech exportsare concentrated in chemical industries.

• Turkey’s high-tech imports consist mainly of cap-ital goods and correspond to a classical form oftechnology transfer aimed at upgrading indige-nous industrial capacities. Turkey’s foreign tradeis strongly structured by its traditional comple-mentaries with Europe.

buyers may have higher quality standards thandomestic buyers. Foreign buyers may also as-sist with process improvements and provide in-formation about and experience with foreignmarkets. Moreover, the additional demandthat foreign markets provide may allow for theexploitation of economies of scale that justifymore capital-intensive production (see, for ex-ample, Hobday 1995 and Rhee, Ross-Larson,and Pursell 1984 on the effects of learning byexporting in East Asia). Technology transfersthrough exports may be most important inproduction networks with clearly articulatedsupply chains (Gereffi 1999; Hobday 1995).5

Spillovers may also be common in labor-inten-sive sectors where production processes are rel-atively simple and the relevant knowledge iswidely available in industrial countries (Enosand Park 1987; Hou and Gee 1993).

Unfortunately, these efficiency benefits fromexporting are not confirmed by econometricstudies (Keller 2004), which generally find thatthe positive relationship between exportingand productivity results largely from the self-selection of firms into the export market.6

Whatever the magnitude of spillovers fromexports and of concerns surrounding the re-export nature of some high-tech exports, theexport of technology products is neverthelessan important indicator of technologicalachievement. High-tech exports offer betterprospects for future growth than lower-techgoods because their market has been expandingmore rapidly and because they offer superiorspillover potential by transmitting skills andgeneric knowledge that can be used in other ac-tivities (Guerrieri and Milana 1998). High-techexports are also less vulnerable to easy entry bylower-wage competitors, substitution by tech-nical change, and market shifts (Lall 2001).

Although high-income countries continue todominatetheworldmarketforhigh-techgoods,7

middle-income countries have substantially in-creased their market share since the mid-1990s.Lower-middle-incomecountriesmore thandou-bled their global share of high-tech exports, in-creasing themfrom6.6percent in the mid-1990sto 15.7 percent in 2002–04 (3.4 to 5.0 percent

if China is excluded). Moreover, the share ofhigher-technology goods in the total merchan-dise exports of these countries has also been in-creasing (figure 3.3). Much of this increase re-flected the transfer of manufacturing processesfrom high-income countries to developingcountries, notably those in East Asia and the Pa-cific. China was a major beneficiary of thisprocess, increasing its global market share ofhigh-techexports fromabout3percent to11per-cent, but so too were other countries. ThePhilippines, for example, increased its marketshare from 0.9 percent to about 2.0 percent.Upper-middle-income countries made less spec-tacular progress, increasing their global marketshare from 6.5 percent to 10.5 percent, althoughsome countries, such as Costa Rica, the CzechRepublic, and Hungary, were able to increasetheir high-tech markets significantly.8 In Sub-Saharan Africa, South Africa is the largest ex-porter of high-tech products, but its share has re-mained small and stable at about 0.08 percent.

The relative performance of differentmiddle-income regions reflects different levels


114

Figure 3.3 Exports of low-, medium-, andhigh-technology goods

Percent of total merchandise exports

Source: World Bank calculations using Centre d’EtudesProspectives et d’Informations Internationales database,Banque Analytique de Commerce International.

0

20

40

10

25

15

5

35

30

1994–96 2002–04

High-techexports

1994–96 2002–04

Medium-techexports

1994–96 2002–04

Low-techexports


Low-income countries

High-income countries


of technological capabilities and learningstyles. Since the early 1990s, Latin Americahas almost doubled its share of high-techproducts in world markets and now has thesecond largest market share after East Asiaand the Pacific. However, in contrast with Eu-rope and Central Asia and the Middle Eastand North Africa, it has done so with rela-tively little input from imported technology(imports of high-tech and capital goods as ashare of GDP in Latin America and theCaribbean are half the rate of Europe andCentral Asia). Partly as a consequence, Europeand Central Asia has gained market share inhigh-tech goods much more quickly thanLatin America, and based on recent perfor-mance, is poised to overtake that region soon.

Low-income countries remain marginalplayers in the world market for high-techgoods, and even though their global share ofexports of medium-tech goods has doubled, be-tween the mid-1990s and 2002–04, it remainslow at 0.8 percent. The share of low-incomecountries in world exports of low-tech goodsis more substantial and has increased from3.5 percent to 5.2 percent. Among these coun-tries, Vietnam has improved its global marketshare of low-tech products from 0.2 to 0.8 per-cent, a 250 percent increase. India remains themost important exporter of low-tech productsin this income group with 2 percent of theworld market, second after China among de-veloping countries, which accounts for about17 percent of the world’s low-tech exports.

Overall exposure to foreign technologieshas increasedOverall, the increased participation of devel-oping countries in global trade has substan-tially increased their exposure to foreign tech-nologies. For middle-income countries, thisexposure and the attendant expansion of high-tech exports have likely yielded important sidebenefits that are reflected in the sustainedacceleration in developing country growthrates over the past 15 years. While the increasein trade openness has been generalized, theextent to which countries have been able to

translate that into improved export perfor-mance and an increase in the technologicalsophistication of their own exports has varied,with countries like those in Europe and CentralAsia that have relatively well-educated popula-tions and a strong institutional structure hav-ing extracted the greatest benefits. Amongother countries, notably low-income countries,weak absorptive capacity may be restrictingthe extent to which their economies have ben-efited from the increased exposure.

Foreign direct investmentLike trade, FDI can be a powerful channel forthe transmission of technology to developingcountries by financing new investment, bycommunicating information about technologyto domestic affiliates of foreign firms, andby facilitating the diffusion of technology tolocal firms.

Foreign investors bring both equipmentand know-howMeasuring the technological contribution ofFDI is particularly difficult, in part becausethe standard measure from the balance of pay-ments includes both physical (brownfield andgreenfield) investments and financial invest-ments (mergers and acquisitions). This said,FDI inflows to developing countries rose from$10 billion in 1980 to an estimated $390 bil-lion in 2007, or from 0.4 to 2.9 percent ofGDP, with the bulk of the increase occurringduring the late 1990s in response to the liber-alization of FDI policies.9 Assuming that for-eign firms employ a higher level of technologythan the average domestic firm, then this ris-ing trend will have increased the average levelof technology in these countries, as well astheir exposure to higher technologies.

FDI as a share of GDP has risen in all de-veloping regions and income groups since the1980s, but the increase has been concentratedin middle-income countries, where FDI rose toalmost 3 percent of GDP (table 3.2). East Asiaand the Pacific had the highest ratio during the1990s, but it has since declined, in part be-cause of a collapse in FDI inflows to a few


115

countries affected by the East Asian financialcrisis in the late 1990s (particularly Indonesia),and in part because FDI inflows to China, al-though still high, have failed to keep pace withthe rapid growth of output. Meanwhile inLatin America and the Caribbean, efforts to in-crease trade openness resulted in a boom of FDIinflows, which reached an average of 3.5 per-cent of GDP in 2000–06. In Europe and Cen-tral Asia, FDI rose from next to nothing beforethe breakup of the Soviet bloc to 2.2 percent ofGDP in 2000–06. In the Middle East andNorth Africa and South Asia, FDI remains lowat around or less than 1 percent of GDP. Mostrecently, FDI inflows to Sub-Saharan Africahave surged, reflecting substantial investmentin oil and mineral production and a moregeneralized interest in the region stemmingfrom increased political stability, liberalizationof FDI policies over the past 15 years, andimproved growth performance.

One way that FDI boosts technologytransfer is by financing new machinery andequipment purchases. The share of FDI indeveloping countries’ fixed capital formationincreased from 2.9 percent in the 1970s to 10.7percent in this decade (table 3.3), with the in-crease in middle-income countries being morepronounced than in low-income countries.However, FDI includes mergers and acqui-sitions that may involve no additional physicalinvestment, and the share of mergers and ac-quisitions, including privatization transactions,

in total FDI has been rising. Nevertheless, theforeign component of aggregate investment indeveloping countries has likely been risingalong with the extent of technological transferthrough this channel. Moreover, the transfers inknow-how, business process technology, andmarket knowledge associated with mergers andacquisitions can occur whether or not any asso-ciated physical investment is involved, and mayeven be more important.

Foreign firms may also improve the tech-nological capacity of developing countries by


116

Table 3.3 Foreign direct investment as apercent of fixed capital formation

1970s 1980s 1990s 2000s

RegionsEast Asia and

the Pacific 1.9 3.0 12.2 8.4Europe and

Central Asia 2.8 — 8.1 15.5Latin America

and the Caribbean 3.3 0.1 11.3 13.1Middle East

and North Africa 2.8 2.1 3.7 6.5South Asia 0.3 0.4 2.5 4.1Sub-Saharan Africa 3.6 2.2 9.3 18.6Income groupsLow-income countries 2.2 1.3 5.6 6.7Middle-income countries 2.9 1.9 11.0 11.3All developing

countries 2.9 1.8 10.4 10.7OECD countries 13.5 11.3 3.5 11.5

Source: World Bank 2007a.Note: — � not available; OECD � Organisation for Economic Co-operation and Development.

Table 3.2 Foreign direct investment as a percent of GDP

1970–79 1980–89 1990–99 2000–06

All developing countries 0.5 0.4 1.5 2.7

By regionEast Asia and the Pacific 0.5 0.5 2.8 2.3Europe and Central Asia 0.3 0.1 0.9 2.2Latin America and the Caribbean 0.6 0.7 1.5 3.5Middle East and North Africa 0.5 0.5 0.6 1.1South Asia 0.0 0.1 0.4 0.7Sub-Saharan Africa 0.7 0.4 1 2.5

By income groupsLow-income countries 0.3 0.2 0.8 1.1

excluding India 0.6 0.4 1.4 1.8India 0.0 0.0 0.3 0.7

Middle-income countries 0.5 0.4 1.6 2.9

Source: World Bank 2007a.

financing R&D. Multinational corporationsundertake most of their R&D activities intheir home country or in other high-incomecountries.10 Nevertheless, the role of develop-ing countries appears to be rising. R&Dspending in developing countries by majority-owned foreign affiliates of U.S. parent compa-nies increased from $0.9 billion in 1999 to$1.6 billion in 2003 (Bureau of EconomicAnalysis 2007). The contribution of multina-tionals’ R&D to total measured R&D activityin developing countries varies from more than60 percent in Hungary to less than 5 percent inIndia (figure 3.4).

FDI may generate technology spillovers In addition to its technological impact on thefirm directly touched by the investment, FDImay also affect the level of technology indomestic firms.11 Spillovers can arise whenworkers receive training or accumulate experi-ence working for multinationals and thenmove to domestic firms or set up their own en-terprise (Fosfuri, Motta, and Ronde 2001;Glass and Saggi 2002). For example, withinsix years of the beginning of FDI-led export

growth in Mauritius, 50 percent of all firmsoperating in export-processing zones were lo-cally owned, founded, managed, and staffed—in many cases by employees who had receivedon-the-job training in foreign enterprises andhad left to set up their own companies (Rhee,Katterback, and White 1990).

Already existing local firms may also ob-serve the actions of foreign firms and learnabout new products, equipment, marketingtechniques, and management practices. For ex-ample, 25 percent of the managers of Czechfirms and 15 percent of the managers of Latvianfirms report that they learned about new tech-nologies by observing foreign firms as they en-tered their industry (Javorcik and Spatareanu2005). In Morocco and Tunisia, domesticallyowned international call centers have risen inimitation of foreign firms (box 3.2).

If multinational entry leads to an increasein demand for intermediates, it may result inthe expansion of upstream domestic industries(see the experience with Zambian supermar-kets in box 3.3). Downstream industries mayalso benefit from the increased competitionand added variety of inputs created by the for-eign investment. In addition, foreign investorsmay provide advice, designs, direct produc-tion assistance, or marketing contacts to sup-pliers, which the latter can then deploy morebroadly than simply providing cheaper ormore reliable inputs to the foreigners.12

The entry of multinationals is likely to in-crease competition for the domestic firmswithin the industry, potentially forcing themto improve their efficiency and introducenew technologies or business strategies(Blomstrom, Kokko, and Zejan 2000), as inWal-Mart’s joint venture in Mexico (box 3.4).Such competition can make surviving domes-tic competitors stronger, but other domesticfirms may be driven out of business, lose mar-ket share, and experience a loss of high-skilledworkers and higher costs for intermediategoods resulting from increasing demand fromthe foreign-owned firms.13 These effects mayvary by industry depending on factors suchas the market structure before the entry of


117

Figure 3.4 Share of foreign affiliates inbusiness R&D expenditure

0

Hunga

ry

Brazil

Lithu

ania

Mex

ico (2

001)

Thaila

ndChin

a

Latvi

a

Slovak

Rep

ublic

Estonia

India

(199

9)

Argen

tina

Poland

Turk

ey (2

000)

Chile

(200

2)

10

20

30

70

50

40

60

Sources: OECD; Activity of Foreign Affiliates database.http://www.sourceoecd.org; UNCTAD 2005; Eurostat;R&D statistics.

Note: The year is 2003 unless otherwise indicated.

Percent


118

The liberalization of the Zambian distribution sec-tor allowed the inflow of foreign retail companies,

which replaced some of the traditional retail sectorwith modern supermarkets and added both upstreamand downstream benefits to the domestic economy.Consumers benefited from lower costs, while localsuppliers learned new production and marketing tech-niques that enabled them to improve the quality, effi-ciency, and revenues of their operations.

When Shoprite of South Africa first opened super-markets in Zambia, it found that the quality andquantities provided by individual local smallholderfarmers were too low and too unreliable, and there-fore imported 90 percent of its fresh produce from

Box 3.3 South African investment in Zambia’s retailsector has improved the quality of local produce andfarmers’ earnings

South Africa. Farmers’ cooperatives that benefitedfrom donor-funded technical assistance have sincemanaged to improve the quality of their products andservices, and Shoprite is now sourcing 90 to 95 percentof its fresh produce from Zambian farmers. In the caseof one cooperative, farmers’ cash income has increasedfrom $2 to $3 a month to $50 to $70 a month, andlocal access to health care and education services hasimproved. The supermarkets themselves are providinglocal farmers with technical assistance. Agricultural ex-perts from the retailer’s subsidiary visit farms, give ad-vice on crop sequencing, and provide inputs.

Source: Mattoo and Payton 2007.

Leading call center companies from France andSpain have paved the way for domestically

owned and export-oriented call centers in Moroccoand Tunisia. Although call centers existed inMorocco and Tunisia before the first European callcenters outsourced operations to the region in theearly 2000s, domestically owned firms served thelocal market and provided only basic telecommuni-cations services. Only after Atento (a Spanish-ownedsubsidiary) set up a call center in Morocco andTeleperformance (a French-owned firm) settled inTunisia to serve clients in Europe did local entrepre-neurs jump into the European market.

This transition was not without problems. Locallyowned call centers lack the international name

Box 3.2 European call centers in the Maghreb have inspired local entrepreneurs and prompted aspecialization in high-value-added services

recognition of large foreign-owned firms and havehad difficulty obtaining contracts from foreigncompanies. In addition, most of them are small, witha maximum of 100 to 200 employees, compared withmore than 2,000 positions at some foreign-owned callcenters.

Interestingly, many domestically owned firms,such as Outsourcia in Morocco, have chosen todifferentiate themselves by targeting higher valueadded services. High turnover rates have also helpeddomestically owned firms to hire experienced agentswho were trained in the foreign call centers.

Source: World Bank forthcoming.

foreign multinationals, the R&D intensity ofthe products, and the links between foreignfirms and domestic firms in upstream anddownstream sectors.14

Outsourcing decisions, domestic policies,and absorptive capacity all affect spilloversEvidence indicates that spillovers to local sup-pliers are not uniform across countries oracross industries within a country.15 Multi-nationals may choose not to source inputs lo-cally because of concerns about the quality oflocal inputs or the time required to developrelationships with local suppliers or becauseof centralized sourcing arrangements that mayprovide volume discounts or access to cus-tomized inputs (UNCTAD 2001). In hostcountries with underdeveloped upstream sec-tors or in cases of FDI with very specializedinput needs, the scope for spillovers to up-stream sectors may be limited. Policies mayalso reduce the potential for spillovers. For ex-ample, in a highly protected market, foreignplants may operate at an inefficiently smallscale (Moran 2007). Requirements that for-eign firms enter into joint ventures with localcompanies may discourage use of the most ad-vanced technology to avoid leakage to poten-tial competitors (Beamish 1988).

The level of spillovers also depends on thedomestic absorptive capacity. For example,other advanced countries tend to gain fromtechnology spillovers from the activities ofsubsidiaries of U.S. multinationals, while poorcountries do not (Xu 2000). Firms using ad-vanced technology in low-income countriesfail to achieve the same level of productivity asfirms in industrial countries (Acemoglu andZilibotti 2001) or the same kinds of spilloversas in middle-income countries, in part becausethe gap between the quality and human capi-tal of the domestic workforce and that forwhich the equipment was originally designedis too large (Borensztein, de Gregorio, and Lee1998). In general, spillovers may be morecommon when the difference in technologicallevels between the foreign multinational andthe domestic economy is not too large.

As discussed earlier, the extent to which acountry benefits from spillovers to the rest ofthe economy also depends on the country’spolicy stance on basic technological literacyand more advanced skills and on promotionof the adoption and diffusion of technologieswithin the economy. For example, some coun-tries such as Mexico and the Philippines havebenefited relatively little from FDI spilloversbecause FDI inflows, although abundant, have


119

The Aurerra–Wal-Mart joint venture, which cre-ated Walmex, followed Mexico’s reduction in

tariffs and liberalization of FDI in the late 1980s. Byexercising its bargaining power, Walmex squeezedprofit margins among the major soaps, detergent,and surfactants suppliers, offering them higher vol-umes in return.

Local firms that were not efficient enough to meetWalmex’s terms lost market share, and many failed.Those that survived grew and became more efficientand innovative. The quality and efficiency of theiroverall operations benefited from their interactionswith both their client (Wal-Mart) and with foreign-

Box 3.4 Wal-Mart’s entry in Mexico boosted theMexican soaps, detergents, and surfactants industry

owned suppliers, with many firms adopting techniquesand products first introduced into the market by theirmultinational competitors.

As a result of these spillovers and labor shedding,the surfactants sector rapidly improved its value addedper worker. The newly found competitiveness and ex-posure to the requirements of Walmex and other for-eign retailers on the domestic market has allowed theMexican surfactants sector to expand its exports andmarket share in the United States, targeting the Latinocommunity in which their brands are known.

Source: Javorcik, Keller, and Tybout 2006.

been oriented toward exploiting low wagesand have not built many links to the domesticeconomy. In contrast, countries like Singaporehave actively sought to maximize the technol-ogy spillovers from FDI by investing in thedomestic skills and competencies necessary tosupport high-skill and high value added indus-tries and by welcoming and promoting FDI insuch sectors (Lall 2003).

Spillovers might be highly concentrated in certain regions within a countryGeographic proximity also determines theextent of technological spillovers observed.The closer a local firm is to a foreign-ownedfirm, the more frequently will the firms’ em-ployees interact with each other, increasing thelikelihood that employees (and their acquiredknowledge) will move between the two firms.The spatial aspect is also important for verti-cal spillovers between foreign-owned firmsand their local suppliers, which are often lo-cated close to each other (Jaffe 1989).

The existence of such cluster effects mayexplain why FDI tends to be geographicallyconcentrated within a country mainly aroundlarge cities or coastal states. In Russia, for ex-ample, more than two-thirds of the FDI stockin 2000 was in Moscow and three surround-ing regions (Broadman and Recanatini 2005).Similarly, almost half of FDI flows in India goto the Mumbai and Delhi areas (Reserve Bankof India 2007), while in China, almost 90 per-cent of FDI flows go to the western coastal re-gion (Kui-Yin and Lin 2007). This said, be-cause of data limitations and conceptualproblems, econometric support for the notionthat such clusters generate important technol-ogy spillovers is limited (Lipsey and Sjohom2005), with some studies supporting their ex-istence (see Girma and Wakelin 2001 for theelectronics sector in the United Kingdom) andothers not (see Aitken and Harrison 1999 forVenezuela and Sjoholm 1998 for Indonesia).

Developing countries also purchase foreignhigh-tech firmsOutward FDI, that is, the purchase of foreignfirms by domestically owned ones, and the

licensing of technologies are two other, moredirect mechanisms by which developing coun-tries acquire foreign technologies and researchexpertise.

Over the past 20 years, firms domiciled indeveloping countries have increasingly turnedto foreign acquisitions as a means of expandingtheir market share and gaining control overtechnology. Cross-border mergers and acquisi-tions by multinational corporations located indeveloping countries increased from $400 mil-lion in 1987 (less than 1 percent of globalmerger and acquisition transactions) to almost$100 billion in 2006 (almost 9 percent of globalmerger and acquisition transactions) (WorldBank 2007a). Although technology may not bethe primary motivator in many of these pur-chases, technology transfer is associated withnearly all of them in the form of control overpatents and knowledge of manufacturingprocesses, marketing, and business process ex-pertise. Table 3.4 summarizes some of the moretechnologically important recent acquisitionsof high-tech firms by developing country firms.

Developing country firms may seek to ac-quire a brand or a marketing or distribution net-work. Examples include the Thai Union FrozenCompany’s purchase of the Chicken of the Seabrand; the South African Brewery’s purchase ofMiller Brewing (a major U.S. beer maker); andMalaysian Berjaya’s purchase of Taiga, thelargest Canadian distributor of building materi-als. Developing country firms may also purchaseforeign firms to acquire R&D capacity. For ex-ample, the Chinese company Shanghai Automo-tive Industry Corporation bought Sangyong ofthe Republic of Korea to enhance its R&Dcapabilities in sport utility vehicles. Accessingforeign technology also takes the form of estab-lishing R&D centers in developed countries. Forexample, Huawei Technologies and ZTECorporation, both Chinese companies, haveestablished R&D centers in Sweden.

Developing countries can also licenseforeign technologiesDeveloping countries can also gain access totechnology through licensing, which typically


120

involves the purchase of production or distri-bution rights for a product and the underlyingtechnical information and know-how for pro-ducing it. As measured by the payment of in-ternational royalties and fees in countries’balance of payments, licensing fees paid by de-veloping countries increased from $7 billion in1999 to $22 billion in 2006, about a fivefoldincrease when expressed as a percentage ofdeveloping country GDP.16 The increase wassharpest for oil- and mineral-exporting coun-tries, reflecting higher prices for oil andcontracts that often expressed these fees as apercentage of revenues or profits. Nevertheless,licensing fees paid by other developing coun-tries also tripled, and for both low- and middle-income countries these fees represented a largershare of overall GDP than they did for oil- andmineral-exporting countries (figure 3.5).

Licensing can be used as a substitute forFDI. Uncertainty about the policy environ-ment may lead multinationals to sell technol-ogy rather than to exploit the technologythrough foreign investment (domestic firmsmay have more information or may be betterplaced than foreigners to deal with a poor pol-icy environment). Evidence suggests that bothFDI and licensing respond to an adequate busi-ness environment, and factors such as patentprotection may shift incentives for investorsfrom FDI toward licensing (Maskus 2002).Where protection of intellectual property

rights is weak, multinationals may be less will-ing to license technology for fear of it beingcopied by domestic firms. Alternatively, theymay only be willing to license out-of-datetechnologies (Maskus 2000). Data on U.S.multinationals show that the likelihood ofentering into licensing agreements increases asdeveloping countries increase their protectionof intellectual property rights (Antras, Desai,and Foley 2007).

Some countries have pursued a licensing-based strategy of technology acquisition in the


121

Table 3.4 Selected purchases of high-tech firms by companies in developing countries,early 2000s

Acquiring company Country of acquirer Year Acquired firm Country of acquired Industry

Netcare South Africa 2006 General Health Care United Kingdom HealthTata Tea India 2006 Tetley United Kingdom Tea Videocon Industries Ltd. India 2006 Daewoo Electronics Korea, Rep. of ElectronicsChalkis China 2005 Le Cabanon France Food processingEssel Propack India 2005 Telcon Packing United Kingdom Tube packingWipro India 2005 New Logic Austria SemiconductorsOrascom Egypt, Arab Rep. of 2005 Wind Italy TelecommunicationsLenovo China 2004 IBM United States PC manufacturingTCL China 2004 Alcatel France TelecommunicationsRanbaxy India 2004 RPG France PharmaceuticalsBOW Technology Group China 2003 Hynix Korea, Rep. of PC manufacturingPKN Orlen Poland 2002 BP (500 petrol stations) United Kingdom Downstream oil

Source: World Bank 2007a.

Figure 3.5 Licensing payments have risensharply

0Oil andmineral

exporters

Other Oil andmineral

exporters

Others

1990–04

1995–99

2000–06

0.05

0.25

Percent of GDP

0.15

0.10

0.20

Sources: Balance of Payments Database (IMF) and WorldDevelopment Indicators.

Middle-income countries Low-income countries

belief that domestic firms will be able to up-grade their own technological capacities byworking with licensed technology. For exam-ple, in the 1950s and 1960s, Japan kept itseconomy relatively closed to FDI to encouragemultinationals that wished to gain from thegrowing Japanese market to license technol-ogy to domestic firms (Pack and Saggi 1997).China has also encouraged joint ventures, asopposed to FDI, to maximize technologytransfers to local firms. This strategy is likelyto work only if the country has sufficient mar-ket power. Moreover, such discriminatorypolicies run the risk of resulting in the transferof substandard technologies (Hoekman andJavorcik 2006). In contrast, several LatinAmerican countries discouraged the licensingof technology from abroad because of con-cerns about unfair pricing and competitionwith local technologies, a strategy that re-tarded or skewed technological developmentin that region (Pack and Saggi 1997).

The bulk of international royalties and feesstems from intrafirm transfers. In part, thismay reflect a preference by multinationalfirms to transfer more advanced technologiesonly to wholly owned subsidiaries rather thanto partially owned affiliates and to enter mar-kets through wholly owned subsidiaries ratherthan through joint ventures (Javorcik 2006;Mansfield and Romeo 1980). However, it mayalso mean that these fees are being used as amechanism for repatriating profits, perhapsfor tax reasons. As a result, the level of royal-ties and fees may not be a market-based re-flection of the value of technology purchasedby the local subsidiary (Robbins 2006).

Partly because of intrafirm payments and ofthe close relationship between licensing andFDI, economists have had difficulty in evaluat-ing the impact of licensing on technologytransfer. Nevertheless, a few case studies havedocumented its benefits. Brazil and Koreaachieved considerable success in absorbingnew technologies through licensing (Correa2003), and licensing agreements were animportant factor for the success of flori-culture in Kenya, maize in India, and the

electronics sector in Taiwan, China (Chandraand Kolavalli 2006). The latter study alsohighlights that even though licensing may en-able rapid acquisition of product and processknow-how, it also requires a significant level oflocal technological capability to put thelicensed technology to work.

International migrationAlong with trade and FDI, international mi-grants are another important channel for thetransmission of technology and knowledge.From the perspective of developing countries,however, the direction of technology transfercan be both outward (as migrants take awayscarce skills) or inward (through contacts withthe diaspora).

The direction and scale of technology flowsthat result from international migration areless clear than for FDI and trade. On the onehand, the brain drain associated with bettereducated citizens of developing countriesworking in high-income countries is a seriousproblem for many developing countries. Onthe other hand, the contribution that these in-dividuals would have made had they stayedhome is uncertain given the lack of opportuni-ties in some countries. Moreover, developingcountries can benefit from the immigration,albeit often temporary, of managers and engi-neers that often accompanies FDI; the returnof well-educated developing country emi-grants; and the contacts with a technologicallysophisticated diaspora.

High rates of skilled out-migration imply anet transfer of human capital and scarce re-sources (in the form of the cost of educatingthese workers) from low- to high-incomecountries (UNCTAD 2007; World Bank2006a). For some countries, the brain drainrepresents a significant problem: emigrationrates of highly educated individuals can ex-ceed 60 percent in some small countries (fig-ure 3.6), and since 1990, the highly educateddiaspora of developing countries has doubledin size.17 However, the share of developingcountry tertiary-educated individuals livingabroad remained stable and relatively low,


122

ranging from 5 to 13 percent depending on theregion (figure 3.6), because the number ofsuch individuals also doubled.

The emigration of professionals who makea direct contribution to production, such asengineers, may result in reduced rates of do-mestic innovation and technology adoption(Kapur and McHale 2005a). Emigration ratesfor scientists, engineers, and members of themedical profession tend to be higher than forthe general university-educated population.For example, in India, the emigration rate for

those with a tertiary education is 4 percent,but the rate for graduates of the elite IndianInstitutes of Technology ranged from 20 to30 percent in the 1980s and 1990s (Docquierand Marfouk 2004; Khadria 2004).

More moderate migration rates may be ben-eficial, especially when domestic opportunitiesare limited, because of technological transfersfrom the diaspora and because most migrationis not a one-way flow. For example, a majorityof foreign students from many developingcountries who earn their doctorates in theUnited States return home (figure 3.7), bring-ing with them a great deal of technological andmarket knowledge that represents an impor-tant technological transfer in favor of the de-veloping country.

The share of recent doctoral graduates fromdeveloping countries who remain in the hostcountry varies significantly across countries oforigin. In part, these cross-national differencesreflect differences in opportunity costs. Thelikelihood that a student remains in the UnitedStates after graduation falls as average percapita incomes in the home country rise. How-ever, even at a given income level, the length ofstay varies significantly across countries, withfewer graduates returning home to countriessuch as Argentina, China, India, and theIslamic Republic of Iran than would be ex-pected based on income alone. Other factorsexplaining high retention rates include thequality of living conditions and research facili-ties in high-income countries, as well as thedensity of research networks and the size of thepreexisting diaspora. Factors favoring a returninclude proximity to family, cultural affinities,and emigrants’ desire to contribute to techno-logical progress in their native country.18

The diaspora is a major source of skillsand capital Repeated waves of emigration have led to thecreation of vibrant diasporas that possesscutting-edge technology, capital, and profes-sional contacts. For example, developingcountries accounted for three-quarters (ap-proximately 2.5 million) of the 3.3 million


123

% of tertiary-educated individuals in OECD countries, 2000

Source: World Bank staff calculations based on Docquierand Marfouk 2004.

Note: OECD = Organisation for Economic Co-operation andDevelopment.

St. Kitts and Nevis

Antigua and Barbuda

Cape Verde

Grenada

Tonga

Trinidad and Tobago

Suriname

Jamaica

Guyana

Haiti

0 605040302010 70 80 100

90

86

83

82

78

74

72

71

69

67

90

Figure 3.6 The brain drain is a severeproblem in a number of small countries

% of low-, medium-, and high-skilled individuals living inOECD countries, 2000

0

10

14

4

6

12

8

2

Low skilled

Medium skilled

East A

sia

and

the

Pacific

Europ

e an

d

Centra

l Asia

Latin

Am

erica

and

the

Caribb

ean

Midd

le Eas

t

and

North

Afri

ca

South

Asia

Sub-S

ahar

an

Africa

High skilled

immigrant scientists and engineers living in theUnited States in 2003.19 Moreover, becauseout-migration rates are higher for high-skilledindividuals than for low-skilled individuals, onaverage, the diaspora is much more skilledthan the home-country population and repre-sents an important concentration of expertise(figure 3.8). Notwithstanding the size of thediaspora, relatively little rigorous empiricalresearch exists on whether and to what extentit influences technology adoption and creationin emigrants’ home countries.20 The primaryevidence of diaspora contributions to knowl-edge transfers comes in the form of case stud-ies. At a minimum, the technical, market, andmarketing knowledge of national diasporas isa huge potential technological resource.21

Returning migrants can be a major sourceof entrepreneurship, technology, marketingknowledge, and investment capital (Brinker-hoff 2006a, 2006b; Kapur 2001).22 Migrantsreturning to Egypt tend to have higher levels ofhuman capital than nonmigrants and are likelyto be more entrepreneurial the longer they

work abroad (McCormick and Wahba 2003;Wahba 2007). Returning migrants and mem-bers of national diasporas who are still abroad


124

Figure 3.7 Share of Ph.D. students still living in the United States five years after graduation

Percent No. of graduates, 1996

0

20

40

100

80

60

Percent

0

20

40

100

80

60

9.56.0 6.5 7.0 7.5 8.0 8.5 9.0

ChinaIn

dia

Argen

tina

Iran

Turk

ey

Egypt

, Ara

b Rep

. of

South

Afri

caPer

u

Taiw

an, C

hina

Colom

bia

Mex

icoChil

e

Brazil

Korea

, Rep

. of

Indo

nesia

0

400

800

1,200

1,600

Stay back rate(left scale)

Number graduating in1996 (right scale)

Log of per capita gross national income in 2001

INDCHN

IDN

IRN

EGY

TUR

PERCOL

ZAF

BRA

ARG

CHL MEX

KOR

a. Share of foreign Ph.D. graduates who remain inthe United States for five or more years

b. Higher home-country incomes are associated withlower retention rates

Source: World Bank calculations based on Finn 2003.

Note: Country mnemonics follow ISO standards.

Source: Docquier and Marfouk 2004.

Figure 3.8 High-skilled emigrantsare disproportionately representedin the diaspora

Percent of emigrants, 2000

0

40

60

10

20

50

30

East A

sia

and

the

Pacific

Europ

e an

d

Centra

l Asia

Latin

Am

erica

and

the

Caribb

ean

Midd

le Eas

t and

North

Afri

ca

South

Asia

Sub-S

ahar

an

Africa

Medium skilled

High skilled

have made major contributions to technologi-cal progress in their home countries (box 3.5).

The diaspora also contributes to technologytransfers and adoption by strengthening tradeand investment linkages. The high-skilled dias-pora of countries such as India has contributedto the growth of the information technologysector, outsourcing (Kapur and McHale 2005b;Pandey and others 2006), and FDI in theirhome countries. The flow of outward FDI fromthe United States is strongly correlated with thestock of migrants from the origin country.23

Nearly half of the $41 billion in FDI that Chinareceived in 2000 may have originated from itsdiaspora abroad (Wei 2004). Similarly, 60 per-cent of the increase in bilateral trade in differ-entiated products within Southeast Asia may beattributable to ethnic Chinese networks (Rauchand Trindade 2002).24 Moreover, technologyappears to diffuse more efficiently through

culturally and nationally linked groups, andshared ethnicity appears to counteract the kindof home bias effects that underpin the geo-graphic network or the cluster effects that givehigh-density R&D zones an innovation advan-tage (Agrawal, Kapur, and McHale 2004).

Diaspora networks and returnees helppromote technology adoptionThe diaspora’s political engagement in homecountries can also improve local technologicalabsorptive capacity, both through return andby exercising pressure on home country politi-cians from afar. Many leaders of developingcountries were educated abroad and returnedto strengthen political institutions in their coun-tries of origin (Easterly and Nyarko 2005). Inaddition, migrants have often played a valuablerole in the transfer of market-based institutions,such as venture capital, entrepreneurship, and


125

An émigré from Bangladesh working in the finan-cial sector in the United States returned to help

create the Grameen Phone network and make mobilephones available to poor people in remote villages(Sullivan 2007). Through its successful Village PhoneProgram, the network has provided business oppor-tunities to some 260,000 Village Phone operators,mostly poor rural women. Grameen Phone now has15 million customers, more than 10 times the maxi-mum potential client base initially estimated byBangladesh Telecom.

In India, Sam Pitroda, a global entrepreneur who divides his time between India and the United States,founded the Center for Development of Telematics,which developed rural automatic telephone exchangesand introduced shared public call offices all over the country, thus expanding access to cheap and reli-able domestic and international calling. The rela-tively low-cost telephone exchanges are designed tooperate without air conditioning and require signifi-cantly less maintenance than conventional ex-changes. The technology has been exported to sev-eral other developing countries.

Box 3.5 Technological transfers through the diasporaand return migrants: Some examples

Bata Shoes of the Czech Republic is an early exam-ple of technology diffusion through migration and re-turn. Faced with the threat of bankruptcy at his smallshoemaking business in the early 1900s, the founderof the company, Tomas G. Bata Sr. went to the UnitedStates in 1904 to learn more efficient mass productionmethods. Bata returned to the Czech Republic toapply these production techniques. He eventuallyexpanded production to several other countries, in-cluding India, Poland, and the former Yugoslavia, andbecame the world’s leading footwear maker by the1930s. Faced with nationalization in 1945, thefounder’s son moved his family and the company’sheadquarters to Canada, returning to the Czech Re-public once again in 1989 with a further transfer oftechnology and business know-how acquired over theintervening years. Since then, the Bata company hascontinued to spread technology. It has opened stores inCroatia, Poland, Russia, and Slovenia, as well as aproduction facility in China, and is modernizing itsBatanagar factory complex in India.

Source: http://www.bata.com; Factiva; Telenor press releases.

corporate transparency, to their countries oforigin.25 Overseas Taiwanese engineers andreturnees, for example, worked closely withpolicy makers to establish a successful venturecapital industry. This has provided local entre-preneurs with an alternative source of finance,which has helped them overcome the constraintposed by the reluctance of state-owned finan-cial institutions to lend to high-risk entrepre-neurial activities in the technology sector(Kuznetsov 2007).

Expatriate knowledge networks have beencreated to foster regular contacts; transfers ofskills; and opportunities for business with re-searchers, scientists, and entrepreneurs in thecountry of origin. Brown (2000) identified 41such networks for 30 different countries.These networks tend to be rich depositories oftalent with high concentrations of memberswith advanced degrees, many earned in thehost countries.26 Colombia’s Red Caldas net-work, set up with government assistance in1991, was one of the first diaspora networksthat succeeded in promoting collaborative re-search between domestic scientists and Colom-bian researchers abroad through workshopsand symposiums, joint research programs, vis-iting researchers, scientific events, publica-tions, and research and training opportunities(Chaparro, Jaramillo, and Quintero 2006).Less formal networks played an important rolein the transition of the Republic of Korea andTaiwan (China) from developing to high-income economies.27 Some diaspora networkshave failed, principally because they were tooambitious, particularly in cases where the pol-icy and institutional environment in the homecountry were not supportive.28 Research sug-gests that the most successful models startsmall to build up trust and credibility beforeattempting to sponsor a major research projector cooperative agenda (Kuznetsov 2006).

Remittances can promote technologydiffusion by making investments more affordableRemittances to developing countries havegrown steadily in recent years, reaching $207

billion in 2006, and are now are larger thanFDI and equity inflows in many countries, es-pecially small, low-income countries. Remit-tances can support the diffusion of technologyby reducing the credit constraints of receivinghouseholds and encouraging investment andentrepreneurship (Fajnzylber and López 2007;Puri and Ritzema 1999; Woodruff andZenteno 2007; World Bank 2006a). A surveyof self-employed workers and small firms inMexico found that remittances were responsi-ble for one-fifth of the capital invested in mi-croenterprises in urban Mexico (Woodruffand Zenteno 2001). In the Philippines, house-holds work more hours in self-employmentand become more likely to start relativelycapital-intensive household enterprises in re-sponse to an exogenous increase in remit-tances (Yang 2006).

Remittance flows have also contributedto the extension of banking services (oftenby using innovative technologies), includingmicrofinance, to previously unserved, oftenrural, sectors. This has improved the accessof households and firms to financial services(box 2.8; Gupta, Pattillo, and Waugh 2007),and their ability to purchase and invest intechnology. For example, remittance rev-enues may have helped Ghana’s ApexLinkand Mongolia’s Xac banks to expand theirnetworks and services (Isern, Donges, andSmith 2006). Cell phone money transfers,such as G-cash and Smart Padala in thePhilippines, and card-based remittances arebecoming prevalent in a number of countries,including Mozambique, South Africa, andthe United Arab Emirates, and are likely toexpand to other countries in the comingyears (Helms 2006; Jordan 2006). Remit-tances have also helped domestic banksfoster links with banks in high-incomecountries. In turn, such links have fosteredtechnology transfers as banks in high-incomecountries have helped local partners to up-grade their systems to comply with the anti-money-laundering, antiterrorism, and know-your-customer regulations in developedcountries.


126

A summary index of trends in theexposure of developing countries to external technologyThe preceding paragraphs have argued that de-veloping countries gain access to foreign tech-nology through trade, FDI, and the diasporaand that these links have been increasing overtime. This section reports on the results of aneffort to summarize these trends by applyingprincipal components analysis to five data se-ries covering trade and FDI following the samebasic methodology as used in chapter 2 to cre-ate the index of technological achievement.29

The index shows that the relationship be-tween income levels and exposure to externaltechnology is relatively weak across countries.Even though the average exposure is higheras one moves from low-income to upper-middle-income developing countries, substan-tial variation is apparent across countries. Thisvariation partly reflects issues of country size(smaller countries tend to be more open thanlarger countries), but it also stems from varyingdegrees of specialization among countries.

The index also shows that many countrieshave increased their exposure to tech-nology during the 1990s (figure 3.9). Of thedeveloping countries for which data are avail-able, 17 experienced a reduction in exposureto foreign technologies over the 1990s and 70saw their exposure rise. The average percent-age increase was highest in low-income andin upper-middle-income countries, with the in-crease among lower-middle-income countriesactually being below that observed amonghigh-income countries. These average resultsreflect a mix of strong increases in excess of100 percent in a number of countries and lessspectacular increases in others.

Much of the variation is attributable tostrong increases in the degree of opennessamong the transition economies. The averageimprovement in the index of exposure to for-eign technology for the countries of Europe andCentral Asia was 80 percent. Excluding thesecountries from the sample lowers the averageincrease in middle-income countries signifi-

cantly, resulting in a clear negative relationshipbetween the observed increase in exposure toforeign technologies and incomes (table 3.5).This is particularly encouraging for low-income countries, as the gains from such expo-sure tend to be nonlinear and enduring(Lumenga-Neso, Olarreaga, and Schiff 2005).For middle-income countries and South Asia,however, this result suggests that many coun-tries (notably those in South Asia) may bemissing out on the potential benefits to beachieved from increased openness.

Nurturing technologicaladaptive capacity

Openness to trade, FDI, and internationalcommunication through the diaspora,

other networks, and various media all serve toexpose a country to technologies and applica-tions of technologies that may not have beenexploited domestically. Exposure does not,however, guarantee that these new technolo-gies will spread and grow within the domesticeconomy. Too often technologically sophisti-cated processes or products are limited to a fewmajor centers or foreign-owned enclaves. Howfar these technologies diffuse within a countryis determined by its technological adaptive ca-pacity, that is, the quality of its labor force andthe business environment (including access tofinance) in which firms operate and are able(or unable) to start up, expand, and reap the fi-nancial rewards of their new-to-the-market in-novations. In the following sections we explorerecent trends in technological adaptive capac-ity among developing countries and the rolesfirms, governments, and individuals play increating and supporting that capacity.

Governance and the businessclimateA stable and predictable economic environ-ment reduces the risk that returns to invest-ments in technology and innovative businessactivities will be lost to conflict or widely vari-able inflation and exchange rates. A stableregulatory environment that facilitates theconduct of business by enforcing property


127

rights, by limiting corruption, and by not im-posing onerous requirements that make thecreation or expansion of firms or theiradoption of new technologies unnecessarilydifficult also contributes to an economy’s tech-nological adaptive capacity.

Political and macroeconomic stabilityhave improved countries’ ability to exploit technologyOver the past 15 years, the number of coun-tries involved in international or domestic


128

Figure 3.9 Most developing countries have increased their exposure to external technology

Index

Source: World Bank.

0High-income



countriesLow-income

countries

0.05

0.25

0.20

0.15

0.10


Percent change in index

�100High-income



countriesLow-income

countries

0

150

200

�50

100

50

Changes in exposure (2000s versus 1990s)

conflict, as measured by the InternationalCrisis Behavior Project, has declined signifi-cantly (figure 3.10). The decline has been mostpronounced in Sub-Saharan Africa, where thetotal number of countries in conflict declinedfrom a peak of 10 in 1998 to only 2 in 2004.Among its many benefits, the cessation ofconflict can provide an environment that ismore conducive to both private and publicsector investments in technology. For example,12 years after the end of hostilities, the gov-ernment of Rwanda launched an ambitious

program of technological capacity building(Watkins and Verma 2007).

Improved macroeconomic stability andgrowth in developing countries has also con-tributed to an environment that is morefriendly toward technological investment.30

High and variable inflation and high exchangerate volatility increase the risk involved ininvestments in technology and increase thereturns to financial manipulation relative to

investment. High government deficits and debtalso increase uncertainty, especially whencombined with a rigid exchange rate regime, acombination that increases the likelihood of anabrupt revaluation that would further cloudexpected future returns. The median inflationrate in developing countries fell from 19 per-cent in the early 1990s to 4 to 6 percent duringthe first half of this decade, exchange ratevolatility is down, and government deficitshave declined across the board and are nowbelow 3 percent of GDP in every developingregion except South Asia. Moreover, the accel-eration of per capita income growth over thepast 15 years, which has been most markedover the past 6 years, has improved the overallaffordability of technology (table 3.6).

A weak business environment and poorgovernance can impair technologicalprogressRegulatory restrictions that impair the econ-omy’s flexibility may limit the absorption oftechnology. Restrictions on labor mobility andrules that constrain firms’ ability to reallocateworkers within the firm can be importantbarriers to the adoption of new technologies(Parente and Prescott 1994). For example,


129

Table 3.5 Increases in exposure to externaltechnologies index, 1990s to 2000s

(excluding Europe All countries and Central Asia)

Regions (percent change)East Asia and the Pacific 13.4Europe and Central Asia 83.7Latin America and the

Caribbean 33.4Middle-East and North Africa 37.8South Asia 13.7Sub-Saharan Africa 33.4

Income groupsHigh-income countries 27.7 27.7Upper-middle-income countries 45.1 24.4Lower-middle-income countries 36.2 31.7Low-income countries 33.5 33.5

Source: World Bank.Note: Values are unweighted averages of country-specificchanges.

Figure 3.10 Number of countries in conflict worldwide

Source: International Crisis Behavior database. http://www.cidcn.umd.edu.icb.

0

1960

10

30

35

Number of countries

5

25

20

15

1965

1970

1975

1980

1985

1990

1995

2000

2005

High-income countries Upper-middle-income countries Lower-middle-income countries Low-income countries


130

Table 3.6 Macroeconomic stability has improved in developing countries

Median inflation ratea Real effective exchange rate volatilityb

1990–94 2002–06 Difference 1990–94 2002–06 Difference

World 11.1 3.6 �7.5 3.7 1.4 2.3High-income countries 3.4 2.1 �1.3 1.6 0.8 0.8Upper-middle-income countries 19.2 4.5 �14.7 2.8 1.4 1.4Lower-middle-income countries 16.3 4.7 �11.6 6.5 1.2 5.3Low-income countries 9.2 4.0 �5.2 3.9 2.0 1.9Developing countries 18.8 4.2 �14.5 4.5 1.5 3.0

East Asia and the Pacific 7.1 3.4 �3.7 1.3 1.2 0.1Europe and Central Asia 326.8 4.3 �322.5 10.6 1.2 9.4Latin America and the Caribbean 17.1 5.5 �11.6 2.9 1.3 1.6Middle East and North Africa 8.9 3.7 �5.2 1.9 1.2 0.7South Asia 9.8 5.4 �4.4 1.4 1.1 0.3Sub-Saharan Africa 9.2 4.6 �4.6 3.9 2.1 1.8

General government balancec General government debtd

1990–94 2002–06 Difference 1990–94 2002–06 Difference

World �5.0 �1.5 3.5 62.6 64.1 1.5High-income countries �4.3 0.8 5.1 58.7 55.6 �3.0Upper-middle-income countries, excluding ECA �2.3 �2.1 0.2 55.1 74.5 19.4Lower-middle-income countries, excluding ECA �3.8 �2.9 0.9 64.3 57.5 �6.8Low-income countries, excluding ECA �5.4 �1.9 3.5 97.9 106.5 8.7Developing countries, excluding ECA �4.1 �2.2 1.9 73.3 78.7 5.4

East Asia and the Pacific �2.0 �2.1 �0.1 45.6 57.0 11.4Europe and Central Asia �9.4 �1.9 7.5 38.9 37.2 �1.7Latin America and the Caribbean �2.2 �2.8 �0.6 67.5 69.7 2.2Middle East and North Africa �6.6 �2.8 3.9 71.6 71.3 �0.3South Asia �7.7 �6.1 1.5 78.2 76.3 �1.9Sub-Saharan Africa �5.3 �1.4 3.9 88.6 97.6 8.9

GDP per capita growthe

1985–94 1995–2006 Difference

World 1.8 2.5 0.7High-income countries 2.1 2.5 0.4Upper-middle-income countries, excluding ECA 2.9 2.1 �0.8Lower-middle-income countries, excluding ECA 1.1 2.1 1.0Low-income countries, excluding ECA �0.5 1.6 2.1Developing countries, excluding ECA 2.0 2.0 0.0

East Asia and the Pacific 3.7 2.5 �1.1Europe and Central Asia �2.4 4.9 7.3Latin America and the Caribbean 2.1 1.7 �0.5Middle East and North Africa 3.2 2.1 �1.1South Asia 2.6 3.3 0.7Sub-Saharan Africa 0.6 2.6 2.0

Source: DataStream, International Monetary Fund, JP Morgan, World Bank.Note: ECA � Europe and Central Asia.

a. Calculated as the mean over each indicated period of the median monthly (year over year) Consumer price index infla-tion rates of the countries in each grouping.

b. Calculated as the period average of the absolute value of the month-over-month percent change of the real effective ex-change rate of countries in each grouping.

c. Calculated as the period average of the simple mean across countries of the central government budget deficit as re-ported by the International Monetary Fund.

d. Calculated as the period average of the simple mean across countries of the government debt as reported to the WorldBank (for low- and middle-income countries and as per the IMF for high-income countries).

e. Calculated as the period average of the simple mean across countries of the growth in GDP per capita.

stringent labor market regulations in Brazilundermine productivity and technical effi-ciency (World Bank 2005), and the removal oflabor market regulations that result in invol-untary overstaffing would increase labor pro-ductivity by 7 percent in India (World Bank2004b). Similarly, restrictions on firm exit andentry can impede technological progress bypropping up inefficient firms and limiting theexpansion and creation of innovative firms.31

Rules and regulations governing firm start-up can be particularly important, because theyhave the potential to prevent a new technol-ogy or new-to-the-market product or processfrom seeing the light of day. The World Bank’sindicators on doing business suggest substan-tial room for improvement in most developingregions. On average, an entrepreneur seekingto begin a new business must undertake morethan 9 separate procedures, which can take al-most 50 days to complete (table 3.7). Amonghigh-income countries that belong to theOrganisation for Economic Co-operation andDevelopment (OECD), the equivalent figuresare 6 procedures and 17 days. Moreover, indeveloping countries the associated fees areparticularly onerous given income levels, con-suming more than an amount equivalent to1.5 years worth of per capita income for a per-son living in South Asia, compared with5 days worth of the per capita income of


131

Table 3.7 The regulatory burden is heavier in developing countries than in the OECD

Procedures Duration Cost Minimum capital requirements

(number) (days) (% of GNI per capita) (% of GNI per capita)East Asia and the Pacific 8.2 46 43 60Europe and Central Asia 9.4 32 14 54Latin America and the Caribbean 10.2 73 48 18Middle East and North Africa 10.3 41 75 745South Asia 7.9 62 163 1Sub-Saharan Africa 11.1 33 47 210Developing-country average 9.5 47.8 65 181OECD countries 6.2 16.6 5 36

Memo:Ratio of developing-country

average to OECD average 1.5 2.9 12.2 5.0

Source: World Bank; Doing Business. http://www.doingbusiness.orgNotes: Procedures required to register a firm, average time spent during each procedure, offical cost of each procedure, andminimum capital required as a percent of income per capita; GNI � gross national income.

someone in an OECD country. Minimumcapital requirements are also high comparedwith income and likely limit the size of formalsmall and medium enterprise sectors, particu-larly in the Middle East and North Africa andin Sub-Saharan Africa, where they representmore than two years of average earnings.

Ensuring timely and efficient exit by failedbusinesses also promotes technologicalprogress by freeing unemployed and underem-ployed capital and workers for more efficientuses. Developing countries, on average, requiremuch more time to resolve insolvencies (rang-ing from 2.7 years for East Asia and the Pacificto 5.0 years for South Asia) than OECD coun-tries (which require an average of 1.3 years). Inaddition, the amount recovered averages lessthan 30 cents on the dollar in all developing re-gions (and only 20 cents in South Asia and17 cents in Sub-Saharan Africa), compared with74 cents on the dollar in OECD countries.32

The quality of regulation, including its en-forcement, and of the business legal environ-ment are critical determinants of the capacityof new and innovative firms to grow and ex-pand. For example, the ability of such firmsto finance their initial operations or conductarms-length operations, both of which arecrucial for technologically advanced compa-nies that require a large customer base toexploit economies of scale, depend on the

system’s ability to enforce contracts, establishproperty rights, and enforce court decisions ina timely and cost-effective manner.

Although the number of procedures re-quired to enforce a court decision in the caseof a contract dispute in developing countries issignificantly higher than in OECD countries,the time taken to reach a decision is not toomuch greater (with the notable exception ofSouth Asia), generally less than 25 days, com-pared with 13 days in OECD countries (figure3.11). However, the time required to enforcelegal decisions approaches two years in four ofsix regions. This seriously affects firms’ (andconsumers’) ability to effect arms-length trans-actions with confidence, and is therefore animportant inhibitor to the growth of techno-logically sophisticated firms.

Corruption can also prevent entrepreneursfrom making investments in technology andexpanding their businesses in a manner thathelps extend the penetration of technologiesinto the economy, while increasing the relativereturn to activities aimed at influencing policymakers. Moreover, better governance is also

associated with improvements in processtechnology, particularly in the delivery of gov-ernment and regulated services. For example,Kaufmann, Kraay, and Mastruzzi (2007)estimate that an improvement of one standarddeviation in the summary governance indica-tor is associated, in the long term, with a two-thirds reduction in infant mortality and atripling of incomes.

Surveys suggest that developing countrieslag behind high-income countries on a widerange of governance indicators (figure 3.12).For example, government effectiveness andregulatory quality are typically considered tobe at half of OECD levels, with indicators forcorruption, rule of law, and voice and ac-countability being even lower.

These aggregate results hide a certainamount of variation across countries andregions (figure 3.13). Governance in theCaribbean and in Eastern Europe and theBaltic countries appears to be much stronger(more than 70 percent of the OECD level)than in the rest of the developing world, whilecountries in the former Soviet Union and Sub-Saharan Africa have the lowest ratings, only 34and 28 percent of OECD levels, respectively.

In contrast to other indicators of perfor-mance, such as inflation and openness, there islittle evidence that developing countries have


132

Figure 3.11 Efficiency of contractenforcement

Number of procedures Days

0

OECD

East A

sia a

nd

the

Pacific

Europ

e an

d

Centra

l Asia

Sub-S

ahar

an

Africa

South

Asia

Latin

Am

erica

and

the

Caribb

ean

Midd

le Eas

t and

North

Afri

ca

5

15

20

45 1,200

10

35

25

30

401,000

800

600

400

200

0

Procedures (left axis) Time (right axis)

Sources: World Bank; Doing Business.http://www.doingbusiness.org.

Figure 3.12 Developing countrygovernance scores relative to OECDaverage

Source: Kaufmann, Kraay, and Mastruzzi 2007.

605550

Index, OECD � 100

4540

Voice andaccountability

Regulatoryquality

Political stability

Control ofcorruption

Rule of law

Governmenteffectiveness

markedly improved their governance over thepast decade. Despite individual country im-provements and marked gains in the regula-tory environment in Europe and Central Asia,on average, the quality of governance aroundthe world has not improved much over thepast decade (see the World Bank’s GovernanceMatters series). For each country that hasdone well, one has experienced deteriorationin its governance indicators. Two countriesthat have experienced notable deteriorationare Belarus and the República Bolivariana deVenezuela. Many countries have not experi-enced any significant change in either direc-tion. On the positive side, the GovernanceMatters series shows that where countries arecommitted to reform, improvements cantake place relatively quickly. For example,during the relatively brief period of 2002–06,Kenya, Liberia, and Ukraine made significantadvances in voice and accountability, whileAlgeria and Angola made substantial progressin political stability. Thus the potential existsfor a rapid, substantial improvement in the

quality of governance in many developingcountries, an improvement that would en-courage technological progress.

Basic technological literacyWhile the policy environment is critical to theabsorption of technology, technologicalprogress also requires a literate workforce.The process by which external technologiesare absorbed, adapted, and integrated into aneconomy is not a mechanical one, but one thatdepends on the quality, quantity, and distribu-tion of human capital, that is, on the techno-logical competencies and the health of thepeople that use and implement the technology.For this reason, efforts to increase the techno-logical competency, knowledge, and under-standing of populations, firms, and govern-ments lie at the heart of the World Bank’stechnology agenda. Especially among poorcountries and in rural areas, existing deficits interms of basic skills are a binding obstacle totechnological progress and income growth.

Low incomes and poor health impair skillformation for technological progressLow income and poor health are perhapsamong the most basic constraints to techno-logical progress. Even if profitable investmentsin technology are available, inadequate incomelimits the ability to generate resources forinvestment. At the level of the economy, lowincome is both a cause and an effect of low lev-els of human capital, limited funds allocated toresearch, thin financial markets, often poorgovernance, and sometimes violence andmacroeconomic instability, all of which limitthe ability to absorb technological innovations.

Recent developments in relation to incomelevels are heartening. Growth rates of GDPper capita have picked up throughout thedeveloping world over the past 15 years(figure 3.14). The number of people living inabsolute poverty has declined by more than250 million, and their share in the populationof the developing world is expected to fallfrom 18 percent in 2004 to around 11 percentby 2015 (chapter 1).


133

Figure 3.13 Regional averages of sixgovernance indicators

Sources: World Bank; Doing Business.http://www.doingbusiness.org.

10080

Share of OECD averagePercentile rank (1–100)

6040200

Latin America


South Asia

Sub-SaharanAfrica

FormerSoviet Union

Eastern Europeand Balticcountries

OECD

East Asia

Developing-country average

Caribbean

Similarly, welcome developments havetaken place in basic health. Life expectancy atbirth has reached 70 years in middle-incomecountries and continues to converge to still-rising high-income country levels (figure 3.15).Among low-income countries, life expectancyis also converging with high-income countries.Excluding Sub-Saharan Africa, where life ex-pectancy has been declining because of HIV/AIDS, life expectancy in low-income countriesincreased from 59 years in 1990 to 64 years in2005, suggesting that in much of the world,poor health should be decreasing as a factorimpeding technological progress.

In Sub-Saharan Africa, the combination ofincomes that are still extremely low and theravages of HIV/AIDS are more problematic.The failure to control HIV/AIDS is itself anexample of poor dissemination of technologyand problems of affordability, as both theknow-how to limit the spread of the diseaseand to control its health effects are wellknown, even though implementation strate-gies are controversial and tend to be countryspecific. Some countries, such as Uganda, havesucceeded in reducing HIV infection rates.Other countries in western Africa have suc-ceeded in limiting its spread. Still others, no-tably Botswana, are doing better at treatingthose infected than in preventing new infec-tion. In too many countries, however, theepidemic continues to grow more or lessuncontrolled, with widespread societal conse-quences (World Bank 2007b). Estimates sug-gest that in Burkina Faso, Rwanda, andUganda, HIV/AIDS is likely to increase thepercentage of people living in extreme povertyby as much as 6 percentage points between2000 and 2015 (UNDP 2003). In Kenya,HIV/AIDS may reduce GDP per adult by 11percent by 2040 compared with what it wouldhave been in the absence of HIV/AIDS (Bell,Bruhns, and Gersbach 2006). In addition tothe incalculable human costs implied, contin-ued high death rates in the adult populationwill have further negative implications for theability of these countries to acquire and toapply technology, both because the experienceand technological competencies of the adultsexpected to die will be lost and because theeducational attainment and literacy of theirchildren and dependents will be impaired(Bell, Devarajan, and Gersbach 2004).

Illiteracy is declining, but still blockscountries’ ability to absorb newtechnologiesThe level of human capital is a major determi-nant of an economy’s ability to adapt and ab-sorb both sophisticated and even more basictechnologies.33 In both high-income (Eatonand Kortum 1996) and developing countries


134

Figure 3.15 Except in Sub-Saharan Africa,life expectancy is improving

40

50

South Asia

Sub-Saharan Africa

Low-incomecountries




80

1960

Life expectancy at birth (years)

60

70

Source: World Development Indicators.

1965

1970

1975

1980

1985

1990

1995

2000

2005

Figure 3.14 Per capita incomes haveaccelerated in recent years

�2

0

2

3

5

6

Low-incomecountries

Average growth in annual per capita incomes

4

1

�1

Source: World Bank.

Middle-incomecountries

Sub-SaharanAfrica

2001–06

1990s

1980s

1970s

1960s

(Caselli and Coleman 2001), the extent towhich a given technology is used within acountry depends importantly on the educa-tional attainment of the population, both be-cause such skills help individuals learn how tomake effective use of a new-to-the-firm or farmtechnique, and because they increase the likeli-hood that firms will learn of new innovationsbeyond the scope of their local communities.

Although the gap between the educationalattainment of individuals living in developingcountries and those in high-income countriesremains wide, it is closing, both in termsof the most basic indicators (literacy and

primary school completion rates) and moresophisticated indicators, such as tertiary edu-cation enrollment rates (table 3.8). Over thepast 15 years, literacy rates have increasedthroughout the developing world, with thebiggest increases recorded among low-incomecountries, particularly in South Asia. Reportedliteracy rates in Europe and Central Asia rivalthose in high-income countries, while in EastAsia and the Pacific and Latin America andthe Caribbean, literacy rates are at or close to90 percent. Elsewhere literacy lags consider-ably, with only 73 percent of the population inthe Middle East and North Africa being able to


135

Table 3.8 Educational attainment indicators

Adult literacy rate Female literacy rate Expected years of schooling

1990–2005 2005a 1990–2005 2005a 2001–05 2005

(% of population (% of female(% point aged 15 (% point population aged (years of

Regions change) and older) change) 15 and older) (% change) schooling)East Asia and the Pacific 10.7 91 14.9 87 1.9 11.2Europe and Central Asia 1.3 97 1.8 96 0.8 12.7Latin America and the Caribbean 2.3 90 2.6 89 0.5 13.1Middle East and North Africa 14.7 73 16.5 63 2.1 11.7South Asia 11.6 58 12.8 46 4.0 9.7Sub-Saharan Africa 5.1 59 5.2 50 3.5 8.0

Income groupsWorld 6.0 82 7.3 77 2.0 10.9High income 0.3 99 0.3 98 0.4 15.8Upper middle income 0.8 93 1.9 92 0.5 13.3Lower middle income 9.2 89 12.0 85 2.0 11.5Low income 9.3 61 10.1 50 3.5 9.0

Primary completion rate Secondary completion rate Tertiary completion rate

1991–2005 2005 1990–2000 2000 1990–2000 2000

(% of population

(% point (% of relevant (% point (% of population (% point aged 15Regions change) age group) change) aged 15 and older) change) and older)East Asia and the Pacific �1.7 97.7 1.5 13.7 0.8 2.5Europe and Central Asia 2.2 94.9 0.1 13.5 1.6 5.5Latin America and the Caribbean 16.9 98.5 1.1 8.7 1.2 4.9Middle East and North Africa 13.7 90.7 2.1 10.0 1.4 3.4South Asia 18.1 83.5 0.5 7.0 0.5 2.0Sub-Saharan Africa 11.5 60.8 0.5 2.3 0.4 1.0

Income groupsWorld n.a. 87.6 0.4 11.9 1.0 4.9High income n.a. 97.4 �0.7 18.6 2.9 13.3Upper middle-income 10.3 98.5 0.7 12.4 1.4 4.9Lower middle-income 2.7 96.6 1.4 12.4 0.9 2.9Low income 17.0 75.9 0.5 6.1 0.4 1.8

Source: World Development Indicators.a. Actual reference year varies by country.

read and write and about 60 percent in SouthAsia and in Sub-Saharan Africa. Moreover,there are concerns about the comparability ofthese data, as some low-income countries re-portedly define literacy as the ability to readand write one’s own name.

Divergence in the literacy rates for womenexplain much of the disparity. For example, inSouth Asia fewer than half of women age 15and older are literate. Women in the MiddleEast and North Africa and Sub-Saharan Africafare somewhat better, with literacy rates of63 and 50 percent, respectively. Althoughthe technological consequences of such wide-spread illiteracy are difficult to quantify, illiter-ate mothers are much less successful in assist-ing their children to learn (Behrman and others1999), have much more difficulty in absorbingnew techniques and instructions that are trans-mitted in written form, and are likely to be lesseffective workers than their better educatedpeers. Indeed, female illiteracy and the result-ing relative ignorance of best practices in childrearing may be a major causal factor in poorchild health care and poor female labor forceoutcomes (Rosenzweig and Wolpin 1994).

Rising primary school completion ratesshould drive further improvements in adult literacyThe rise in literacy rates is due in no small partto increased and longer participation in for-mal education. Primary school completionrates approach 100 percent in about half ofthe developing regions and have increasedsubstantially among poorer regions, notablySouth Asia and Sub-Saharan Africa. Althoughreflective of the average literacy in the geo-graphical region, these numbers hide impor-tant variations across countries. Thus the lowscore for South Asia reflects mainly low liter-acy in India, and it masks the fact that some95 percent of Sri Lankan youth can read andwrite. Similarly, China’s relatively high liter-acy rates and its large weight in East Asiaaggregate mask the less than 90 percent liter-acy rates of less populous countries such asCambodia and the Lao People’s Democratic

Republic. In particular, the gap betweenschool completion rates for girls and boys hasnarrowed significantly. Partly as a conse-quence, youth literacy rates are much higherthan adult literacy rates in South Asia andSub-Saharan Africa, which over time shouldbe reflected in better literacy scores for womenand improved transmission of knowledge andtechnology to future generations (table 3.9).

Of course, progress in providing effectivebasic education is a necessary precursor ofmore formal secondary and tertiary educa-tion. Nevertheless, the more advanced techni-cal and problem-solving skills that are taughtat the basic level can significantly increase stu-dents’ capacity to learn to work with, adapt,and maintain more technologically advancedgoods. Indeed, the main obstacle to deepeningthe use of a given technique or process in acountry is frequently a lack of sufficient num-bers of individuals trained to maintain andinstall systems. For example, in Rwanda ashortage of plumbers and sheet metal workershas been identified as a principal factor con-straining the deployment of the simple kindof rain-harvesting technologies that havesucceeded in increasing the supply of sani-tary drinking water in neighboring countries


136

Table 3.9 Relatively high youth literacyrates

Youth literacy rate

1990–2005 2005a

(% of(% point population

Regions change) age 15–24)East Asia and the Pacific 3.43 98Europe and Central Asia 1.67 99Latin America and the Caribbean 2.29 96Middle East and North Africa 12.23 88South Asia 13.72 73Sub-Saharan Africa 5.42 70

Income groupsWorld 4.16 88High-income countries 0.23 99Upper-middle-income countries 1.82 98Lower-middle-income countries 3.68 96Low-income countries 10.41 73

Source: World Bank; World Development Indicators.a. Actual reference year varies by country.

(Watkins and Verma 2007). While nearly 19per-cent of the population completes secondaryschool in high-income countries, secondaryschool attainment rates range between 10 and14 percent in East Asia and the Pacific, Europeand Central Asia, and the Middle East andNorth Africa, but are below 9 percent in LatinAmerica and the Caribbean, South Asia, andSub-Saharan Africa (table 3.8). Between 1990and 2000, secondary completion rates morethan doubled in the Middle East and NorthAfrica, East Asia and the Pacific, and LatinAmerica and the Caribbean, but remain low. Incontrast, improvements have been much lessmarked in South Asia and Sub-Saharan Africa,and no appreciable gain was apparent in Eu-rope and Central Asia. By income grouping, thestrongest gain in the secondary completion rateis reported by the lower-middle-income coun-tries with an increase of 1.4 percentage pointsduring 1990–2000, compared with half asmuch or less of an improvement among otherincome groupings.

This same general pattern is observed fortertiary-level students. In East Asia and thePacific, South Asia, and Sub-Saharan Africa,2.5 percent or less of the population aged15 years or older has completed tertiary edu-cation. Europe and Central Asia and LatinAmerica and the Caribbean have almost doublethe number of tertiary-level graduates, whilethe Middle East and North Africa falls in be-tween these two groups of regions (table 3.8).The share of secondary graduates who go onto tertiary studies is relatively high in LatinAmerica and the Caribbean, 60 percent, com-pared with 70 percent in high-income coun-tries. However, the share of university students following a scientific as opposed to asocial science curriculum is relatively low(Maloney 2006). Elsewhere, the ratio is 42 per-cent or lower, with only 18 percent of sec-ondary graduates going on to the tertiary levelin East Asia and the Pacific.

This performance in educational attainmentacross levels of schooling is largely consistentwith patterns of public expenditures dedicatedto education. For example, public expenditures

on education among lower-middle-incomecountries averaged about 3.3 percent ofGDP during the 1970s and 1980s and roseto a fairly steady 4.3 percent during 1990through 2000 (figure 3.16). Among low-income countries, the share of public outlaysrose by about half a percentage point to anaverage of 3.3 percent of GDP during1990–2000, up from an average of 2.9 percentduring 1980 through 1985. These sharescompare with government expenditures oneducation of close to 5.3 percent of GDP inhigh-income countries, down from a peak ofjust over 5.6 percent during the mid-1970s.34

While educational attainment rates and ex-penditure data are available for a fairly widenumber of countries, the data are silent on thequality of the education (or knowledge) re-ceived, and it is the quality of education thatdetermines the effectiveness with which indi-viduals can absorb and exploit technology.Here significant concerns have been raised,particularly for poorer countries where thequality of educators and limited resources sapthe value of time spent in school. This suggeststhat the education deficits in developing re-gions could be larger than indicated by na-tional indicators, which, in turn, implies thatthe expansion of knowledge attained (and thecapacity to adopt and adapt technological


137

Figure 3.16 Educational expenditures haverisen in some regions

Public expenditures as a percent of GDP

Sources: UNESCO Institute of Statistics; World Bank.

1970 1975 1980 1985 1990 1995 20002.0

3.0

4.0

6.0

5.5

5.0

2.5

3.5

4.5


Low-incomecountries



knowledge) is not necessarily rising in accor-dance with higher educational attainment.The availability of international test scoresacross countries is very limited outside ofhigh-incomes countries, but some data under-score these concerns. For example, in a num-ber of middle-income countries, the majorityof primary school students fail to meet OECDliteracy standards. In Sub-Saharan Africa, de-spite enrollment rates of close to 100 percent,in some countries fewer than half of grade sixstudents are deemed literate (figure 3.17).Although based primarily on data from high-income countries, research suggests thatteacher quality is a key determinant of differ-ences in student outcomes (Hanushek andWoessmann 2007).

Financing innovative firmsSo far, the discussion has described how the pol-icy environment and human capital can pro-mote technology diffusion. At the same time,and as indicated in chapter 2, affordability, bothat the level of the firm and of the consumer, canbe a major impediment to the diffusion of tech-nology within a country. In this regard, the suc-cess with which an economy’s financial systemsucceeds in channeling resources (savings)

toward their most productive use (investments)is an important determinant of its technologicaladaptive capacity.

Limited financial intermediation restrictstechnology diffusion Neither the banking system, nor equity mar-kets, nor private sector bond markets in devel-oping countries have channeled savings intothe private sector to the same extent as theyhave in high-income countries (table 3.10). Asa result, the arm’s-length channels throughwhich private savings can be directed towardinnovative firms are limited. While banks inhigh-income countries play a significant role inrelaying private savings to investors (private-sector debt is equivalent to some 50 percent ofGDP in high-income countries), this kind of in-termediation occurs at about half that level inmiddle-income countries and almost not at allin low-income countries. On a more encourag-ing note, the run-up in international investors’appetite for risk has increased market capital-ization in developing countries by significantmargins since 2000 (with the exception of EastAsia, where valuations declined). Valuationratios are now much closer to those observedin high-income countries.


138

Figure 3.17 Many developing country students fail to meet literacy standards

Source: World Bank 2007d.

0

20

10

100

Percent Percent

50

40

30

80

70

60

90

Turkey Argentina Colombia Morocco

Fourth grade reading performance on OECD test Sixth grade reading performance on South Africanregional test

0

20

10

100

50

40

30

80

70

60

90

South Africa Uganda Namibia Malawi

Percent ever enrolled Percent enrolled and achieved minimal literacy

Barriers to the finance of high-riskactivities severely impede the spread of technology Although weak intermediation is a generalproblem in developing countries, the problemfor innovative firms or companies seeking toemploy an untested new-to-the-market tech-nique or product is more severe. Innovationcan involve high risk, and traditional sourcesof capital—banks, stock exchanges, andbond markets—often lack the technical exper-tise to evaluate innovative investments. Thusin the absence of demonstrated cash flowsor enforceable collateral, innovative firms or

entrepreneurs are less likely to obtain financ-ing than experienced entrepreneurs operatingwith proven techniques. Coupled with thinmarkets, this translates into higher capitalcosts for innovative firms in developing coun-tries than for those in high-income countries, afact that is reflected in lower R&D intensities(Lederman and Maloney 2006) and a reducedlikelihood that their financing needs are met.35

Innovative firms in developing countriesare also less likely to have access to equityfinancing than do their counterparts in high-income countries because of strict listingrequirements imposed by the regulators of


139

Table 3.10 Weak financial intermediation hinders technology in developing countries

% change % change 1990 2000 2005 2005/1990 1990 2000 2005 2005/1990

Financial system deposits Stock market capitalization

(% GDP) (% GDP)RegionsEast Asia and the Pacific 40.9 41.7 45.6 11.5 38.7 44.1 51.3 32.5Europe and Central Asia 23.8 22.8 31.2 31.0 8.5 13.8 19.7 130.8Latin America and

the Caribbean 31.1 39.9 39.8 28.1 11.7 28.2 46.4 298.3Middle East and

North Africa 46.5 47.4 59.8 28.7 16.0 26.1 63.8 298.3South Asia 22.5 33.3 43.7 94.0 6.7 13.4 26.1 290.0Sub-Saharan Africa 18.5 19.9 24.4 31.4 31.0 27.0 34.9 12.4

Income groupsHigh-income countries 76.0 87.0 91.4 20.4 45.1 105.0 112.2 148.7Upper-middle-income countries 34.3 43.6 45.2 31.8 37.7 36.5 50.2 33.4Lower-middle-income countries 36.0 34.3 39.4 9.7 13.4 20.4 34.3 156.7Low-income countries 16.7 17.4 21.8 30.7 7.6 10.8 22.3 194.7

Private-sector credit Private-sector debt

(% GDP) (% GDP)RegionsEast Asia and the Pacific 30.3 36.7 44.9 48.2 — 37.5 31.0 ..Europe and Central Asia 22.3 18.2 27.9 25.3 — 12.7 12.2 .Latin America and

the Caribbean 28.6 40.6 32.1 12.3 — 27.2 23.3 ..Middle East and

North Africa 35.3 40.5 46.1 30.7 — — — ..South Asia 16.2 21.6 34.1 110.4 — 1.9 4.0 ..Sub-Saharan Africa 17.4 16.3 18.2 4.4 — 20.1 30.2 ..

Income groupsHigh-income countries 81.1 94.2 108.2 33.3 — 47.9 50.0 ..Upper-middle-income countries 32.8 42.5 40.5 23.6 — 27.5 26.1 ..Lower-middle-income countries 27.3 29.3 31.9 16.7 — 27.7 23.4 ..Low-income countries 14.9 13.4 16.4 9.7 — 1.9 4.0 ..

Source: Beck, Demirgüç-Kunt, and Levine 2000. Financial Structure Dataset updated March 20, 2007. For Private Sector Debt the source is World Bank, Financial Sector Development Indicators. http://www.financial-indicators.org (February 2007).Note: — � not available; .. � undefined.

emerging-market exchanges36 (Pfeil 2000).Less stringently regulated, so-called new mar-kets, modeled after the NASDAQ in the UnitedStates, have been developed to fill the void.Many of these markets help investors by pro-viding some, albeit less rigorous, due diligenceof listed firms and by offering risk-poolingservices (offerings of bundled shares that reduceinvestors’ exposure to any one firm). Such cap-ital pool companies allow listing firms to accessequity finance in amounts that are too large forangel investors to provide, but are too small forinstitutional investors. However, the fullpromise of such markets has yet to be felt, inpart because many of them have been obliged tomaintain relatively strict listing requirements toattract foreign investors (Yoo 2007).

Increasingly, venture capitalists and “busi-ness angels” are playing a role in financingnew technologically sophisticated firms indeveloping economies.37 These investors tendto have more technological know-how thando traditional lenders and to be better able tojudge the potential profitability of new ven-tures. Often the transfer of business and mar-keting expertise is as important as the infusionof capital in determining the difference be-tween success and failure for young firms(Avnimelech and Teubal 2004; Mayer 2003).Even though empirical evidence is still scarce,this activity appears to be translating intoincreased innovation (Pfeil 2000). Western-based venture capitalists are increasingly be-coming involved in markets in Asia (Chinaand India), Eastern Europe (notably the CzechRepublic, Hungary, and Russia), and SouthAfrica. Notwithstanding this increased activ-ity, Nastas (2007) reports that only 1 in 200small and medium enterprises in emergingmarkets is likely to secure venture capital fi-nancing, and the ratio is undoubtedly lowerfor firms in less-developed countries.

Supporting innovative firms with R&D and outreachWhile the process of technological advanceoccurs fundamentally at the firm level,the government, along with international

and national organizations, includingnongovernmental ones, can play a role in pro-moting the dissemination of knowledgewithin the domestic economy. In addition tothe formal education system, less formalcontinuing education—notably, outreach pro-grams and R&D programs that focus onadapting technologies to local conditions—have a central role to play.

R&D efforts to adapt existing technologyto local conditions are expandingDomestic R&D capacity is critical indetermining an economy’s capacity both togenerate new technologies and to absorb tech-nologies from abroad. Foreign technologiesfrequently need to be modified so that they aresuitable for domestic circumstances. For ex-ample, equipment and processes may need tobe adapted to differences in the quality of in-puts and in the relative abundance of laborand capital, and a stock of researchers is oftennecessary to understand and evaluate ad-vanced technology (Cohen and Levinthal1989). Building up R&D capacity facilitatesthe imitation and adaption of foreign tech-nologies and improves the extent to which pos-itive spillovers from FDI and trade accrue tothe rest of the economy (Fagerberg 1988; Ki-noshita 2000). Moreover, countries tend to ac-quire technology more readily when domesticfirms have R&D programs and when publicresearch laboratories and universities have rel-atively close ties to industry (Maskus 2000).

Available data indicate that most develop-ing regions have been increasing their R&Dexpenditures relative to GDP (table 3.11).38

East Asia and the Pacific has experienced aparticularly rapid rate of increase in R&D ex-penditures and also has the highest level ofsuch expenditures among those regions forwhich data are available. In contrast withother regions, in Latin America and theCaribbean, both the number of researchersand expenditures on R&D have been fallingor stagnant, reflecting both a reorientation ofpolicy away from university-led R&D (Mal-oney 2006) and tighter fiscal policies.


140

While developing countries spend less onR&D than high-income countries, the gap isnot extreme. Relative to GDP, low-incomecountries spend about one-third as much onR&D than high-income countries. One issuewith the data is that the coverage of commer-cial R&D expenditures is poor in manydeveloping countries, so the figures largely re-flect R&D expenditures by the public sectorand universities. As firms in many developingcountries probably focus on adapting foreigntechnology to local conditions, a significantportion of this important activity may there-fore not be captured.

Firm-level R&D is most effective in promoting technological progress All R&D can contribute to an economy’s ca-pacity to create, adapt, and adopt technology.Nevertheless, because of the fundamental rolethat firms play in diffusing technologythrough the economy, the most productiveR&D tends to be that conducted by firms orby public or university laboratories workingactively with the private sector. Across devel-oping countries, the share of R&D conductedby firms (as opposed to government oruniversity laboratories) is highest in East Asiaand the Pacific, where it rivals the share in


141

Table 3.11 R&D intensities have increased

R&D expenditure Researchers in R&D

1997–2002 2002 1997–2002 2002

Regions (% point change) (% of GDP) (% change) (per million people)East Asia and the Pacific 0.45 1.06 4.4 545.3Europe and Central Asia 0.06 0.88 –1.4 2008.7Latin America and the Caribbean 0.01 0.57 — —Middle East and North Africa — — — —South Asia 0.13 0.77 — —Sub-Saharan Africa — — — —

Income groupsWorld 0.12 2.18 — —High-income countries 0.12 2.43 2.6 3750.0Upper-middle-income countries 0.09 0.71 — —Lower-middle-income countries 0.46 1.01 3.9 499.9Low-income countries 0.18 0.80 — —

Source: World Development Indicators.Note:* Interpolation applied where appropriate; — � not available.

Table 3.12 Private-public sector R&D

Sector of performance Sector of funding

Higher HigherR&D spending Business Government education Business Government education

(% GDP) (share of total) (share of total)World 2.28High-income countries 2.45 63 13 27 49 34 2.1Developing countries 0.83 — —

East Asia and the Pacific 1.44 62 22 14 54 35 2.3Europe and Central Asia 0.94 43 29 20 38 54 0.5Latin America and

the Caribbean 0.56 29 27 33 33 37 27Middle East and North Africa — — — — — — —South Asia 0.73 — — — — — —Sub-Saharan Africa — — — — — — —

Source: Gill and Karas 2007.Note: — � not available.

high-income countries (table 3.12). Note,however, that in China, despite rapid techno-logical progress, the efficiency of R&D spend-ing is relatively low and the impact of R&D isimpaired because of poor linkages among gov-ernment R&D institutes, businesses, and uni-versities (Zeng and Wang 2007). In Europeand Central Asia and in Latin America and theCaribbean, academia and the government areresponsible for a much higher share of R&D.Moreover, in the latter region coordinationbetween R&D carried out by government in-stitutions and private firms has been poor,reducing the impact of R&D on productivitygrowth (de Ferranti and others 2003).

Research on OECD countries suggests thatthe more R&D is conducted at the firm level,the higher the rate of return to public and aca-demic R&D, presumably because having R&Dexpertise close to the firm increases the likeli-hood of successful adaptation of a technologycreated in government, academic, or even for-eign laboratories (Guellec and Pottelsberge dela Potterie 2004). Maloney (2006) concludesthat state-funded R&D that is too academicand/or too disconnected from the private sec-tor is less effective at promoting technologicalprogress than firm-conducted R&D or state-supported R&D that has a strong connectionto business needs. Indeed, the relatively highshare of private sector R&D in East Asia andSouth Asia may have contributed to the morerapid technological progress in those regionsthan in Latin America and the Caribbean andEurope and Central Asia.

Outreach plays a critical role in bringingtechnology to the broader populationToo often the overall effectiveness of R&Dundertaken by government and specialized re-search institutes is reduced because such orga-nizations are divorced from their eventualclients and their incentives are poorly alignedwith the ultimate dissemination of their inven-tions and adaptations.39 Especially in poorcountries plagued by illiteracy and weak com-munication networks, technology outreachprograms can play a critical role in increasing

the diffusion of often simple but importanttechnologies. Agricultural outreach programswere instrumental to the green revolution,even though it took much longer than initiallyexpected for those programs to bear fruit(World Bank 1998). Difficulties encounteredin other efforts to disseminate technologyinclude a lack of skilled personnel to staffthe outreach program and the need to earn thetrust of the local population. Here challengesinclude minimizing the risks people run intrying a new technology, listening to their ex-periences, and adapting techniques as a conse-quence (World Bank 2007d). Enhancing therole of farmers in agricultural outreach pro-grams and relying more on cooperation be-tween government and the private sector—inthose areas where private benefits fromtechnology transfer can be substantial—mayimprove both the impact and financial sus-tainability of outreach efforts.

Direct government policies topromote technologyInnovation requires entrepreneurs: peoplewho are willing to take risks to invest in un-certain projects and who have the organiza-tional skills required to bring new products tothe market. Given the high risks involved, thereturns to successful entrepreneurship mustbe high, but the returns to investment in newtechnology in developing countries can be lim-ited, because potential profits may be reducedby imitation, because of a lack of coordinationbetween firms that produce complementaryinputs, or because economies of scale and ag-glomeration generate threshold effects thatprevent firms from breaking into mature mar-kets (box 3.6).

Government policy can play a central rolein helping firms overcome market failuresThe difficulties that these externalities pose forfirms in certain sectors and those seeking toadopt a new-to-market (or even new-to-the-firm) technology imply that specificgovernment interventions may be necessary toencourage investment in technology.


142

Governments in developing countries have un-dertaken a host of direct interventions inproductive activities to provide demonstrationeffects, encourage innovation that otherwisewould not occur because imitators reap thelion’s share of benefits, resolve coordinationfailures, and move an industry toward efficient

technologies that it would not otherwise adoptbecause of capital market imperfections. Thesesteps have included the following:

• Providing support for industry-specificresearch. For example, Malaysiafunded industry-specific R&D, provided


143

The nonpatentability of new-to-the-market prod-ucts and processes. The vast majority of innova-

tion that occurs in developing countries involves theadaptation of already discovered techniques andproducts to the domestic market, and such innova-tion is not patentable. Lack of patent protectionfacilitates imitation, which may speed diffusion, butreduces the returns to the individual or firm intro-ducing the technique or product to the domesticmarket. As a result, private entrepreneurs under-invest in easily reproduced techniques, even thoughthey could have large social benefits.

Coordination failures limit investment in technol-ogy. Some technologies rely on the availability ofcomplementary inputs. Coordination failures canarise when new industries exhibit scale economiesand some of the inputs require geographical proxim-ity. For example, producing cut flowers for exportrequires an adequate electrical grid, irrigation, logis-tics and transport networks, quarantine and otherpublic health measures, and resources devoted tomarketing the country as a dependable supplier(Rodrik 2004). However, these services have highfixed costs and will not be supplied unless demand issufficient, creating a vicious circle where demand isnot forthcoming because of the lack of supply. Themarket for training is another example of potentialcoordination failure, as workers will demand train-ing only if a demand for trained workers exists, butin the absence of training there is no demand(Rodriguez-Clare 2005). Perhaps reflecting suchfactors, in almost all the successful case studies ofinnovation reported in Chandra (2006), governmentplayed an important role by providing infrastructure,marketing, or training support.

Threshold effects caused by economies of scale inmany manufacturing sectors prevent entry by firmsinto global markets or the introduction of a new

Box 3.6 Principal market failures impedingtechnological progress in developing countries

process in a small market dominated by a preexisting,large-scale competitor. Economies generated fromlearning by doing and the productivity boosts gener-ated by the accumulation of small innovations arerelated impediments that imply that start-ups mustendure an initial period of relatively high costs, whichin the absence of adequate intermediation, may pre-vent them from accumulating sufficient experience orscale to attain adequate levels of profitability. For ex-ample, Arrow (1962) cites evidence that productivityin the production of airframes is a decreasing functionof the total number of airframes of the same type pro-duced previously.

Knowledge spillovers tend to be geographicallybounded within a region where the new economicknowledge is created (Audretsch and Feldman 2004).Audretsch and Feldman (1996) find that the propen-sity of innovative activity to cluster geographicallytends to be greater in industries where new economicknowledge plays a more important role as has oc-curred in, for example, Silicon Valley and Bangalore.Studies also find that access to venture capital in theUnited States is heavily skewed by region (Sorensonand Stuart 2001).

Agglomeration effects, whereby firms benefit fromthe knowledge and human resource spillovers arisingfrom the geographical proximity of firms in the samearea of business, may also prevent developing econ-omy firms from breaking into established markets(Glaeser and others 1992). The absence of such effectsrepresents an important barrier to development inSub-Saharan Africa and in more remote areas ofChina and India, where lack of physical proximityboth raises trading costs and minimizes the potentialfor benefits from interactions with more rapidlygrowing areas.

Source: World Bank.

financing, built infrastructure, and of-fered tax incentives to encourage theprocessing of palm oil (Chandra andKolavalli 2006).40 Governments havealso encouraged innovation by improvingnetworking among enterprises, universi-ties, and government research institutes(Goldman and Ergas 1997). For example,government-funded technology parks inTaiwan, China, encouraged research byproviding high-quality facilities and byfacilitating interactions among scientists.Many governments finance agriculturalresearch and support farmers’ efforts toexploit new technologies.

• Providing direct subsidies for specificproducts. The government started thefirst commercial-scale salmon farming

operation in Chile to demonstrate itsfeasibility (Rodrik 2004). Korea andJapan provided fiscal subsidies to create“national champions” in key sectors(Hoekman, Maskus, and Saggi 2005).Similarly, the Brazilian aircraft andbiofuel sectors (box 3.7), the Indianpharmaceuticals sector, and the SouthAfrican automobile industry were devel-oped using tax incentives, regulatorypolicies that encouraged domestic com-petition, science and technology support,and collaboration with foreign firms(UNCTAD 2003).

• Imposing more dirigiste policies. Somecountries, particularly in East Asia, haveguided production decisions throughinitially high import tariffs, export


144

Brazil launched its National Alcohol Program in1975 to reduce its dependence on crude oil im-

ports and to guarantee the profitability of the sugarindustry by allowing excess sugar production to beconverted into alcohol (ethanol) in special distillerieslocated near sugar mills.

Government support for the initiative has been ex-tensive. Initially it was supported by legislation thatmandated that 24 percent of fuel sold for automobilesmust be in the form of ethanol. This requirement wascomplemented by sponsored research into the produc-tion of, and eventually the subsidization of sales of,cars that ran entirely on ethanol. Moreover, govern-ment credit guarantees and low-interest loans to con-struct the refineries amounted to some 29 percent ofthe overall investment cost of these ventures. At thesame time, the state oil company, Petrobras, wasrequired to make infrastructure investments inethanol distribution and to keep the cost of ethanol toconsumers significantly cheaper than the cost of gaso-line. Overall, the government spent $12.3 billion onthe National Alcohol Program during 1975–98.

During periods of high oil prices the program hasbeen relatively successful, with ethanol-powered carsrepresenting 90 percent of sales between 1983 and

Box 3.7 Government sponsored innovation:Brazilian biofuels

1988. However, as oil prices fell, the technology be-came less attractive and the costs of supporting the in-dustry rose. At the same time, world sugar prices roseand sugar growers shifted their cane to the productionof sugar for export instead of ethanol for the domesticmarket. Ethanol shortages developed and ethanol-powered car sales dropped. As a result, the govern-ment gradually rescinded the program’s incentives andsubsidies, although it still mandated that all gasolinecontain roughly 20 percent ethanol, citing environ-mental benefits to justify the mandate.

The flex-fuel car engine, which was able to run onany combination of ethanol and gasoline, was intro-duced in 2002 and, along with the surge in crude oilprices, led to a revival of ethanol in Brazil. With therecent rise in oil prices, ethanol-based cars are onceagain competitive, and ethanol produced from Brazil-ian sugar can be produced for less than the equivalentquantity of gasoline. Other developing countries areincreasingly interested in adopting the Brazilian tech-nology to reduce their energy dependence and also tosupport their sugar sectors.

Source: Coelho and Goldemberg 2004; World Bank 1994; XavierMarcos 2007.

subsidies, government influence over theallocation of production, and directedcredit programs. The extent to which suchpolicies were necessary to these countries’success has generated some controversy(Hernandez 2004), and the forms ofintervention in high-growth East Asianeconomies were by no means identical.For example, Korea (box 3.8) favored tar-iff protection and constraints on FDI tomaximize technology transfer; Singaporeencouraged FDI; and Hong Kong, China,practiced laissez-faire policies. Neverthe-less, some observers doubt that such in-terventions were pervasive in many of themost successful East Asian economies.

But government efforts at promotingtechnological champions have often failedNotwithstanding the wide range of supportpolicies that governments have tried and theexistence of many apparent success stories,such policies have often been spectacular fail-ures. Even among the examples cited, it is notclear that all should be considered successes.For example, the Brazilian aircraft maker

Embraer did not become commercially suc-cessful until it was privatized. More generally,the import-substitution policies followed bymany countries in Latin America and Africaand India’s inward-focused policies severelyhampered economic and technological devel-opment. To take two of the countless exam-ples, first, rather than promoting the develop-ment of a technologically sophisticated exportindustry, the tariffs, price harmonization, andimport licensing programs imposed in Côted’Ivoire diminished incentives for efficiency inits textile industry, making it internationallyuncompetitive. Second, Brazil’s attempts topromote its domestic personal computer sec-tor by banning imports and FDI, awarding li-censes for production, providing fiscal incen-tives, and establishing a public research centerresulted in an inefficient industry, high domes-tic prices, and lagging technology (WorldBank 1998).

Two important issues distinguish industrialpolicies in the successful East Asian countrieswith those in many other countries. First, incontrast to Latin America, where subsidieswere often provided free of performance


145

GDP per capita in Korea increased from $150 in1960 to $16,000 in 2005, with GDP growing

particularly rapidly during the 1980s and early1990s as technological progress accelerated. Althoughan outcome of many factors, Korea’s technologicallyintensive growth spurt had a strong public policyelement. Government-funded research institutes, in-cluding the Korea Science and Engineering Founda-tion in the 1970s, recognized the need to enhance co-operative research between universities and industry(80 percent of the nation’s research capability wasuniversity based in the 1970s). The government cre-ated research centers located within corporations tosupply industry with high-tech research capabilityand industry dictated the focus and area of research.

Box 3.8 A successful government program oftechnological development and innovation financing in the Republic of Korea

With time, industry realized that domestically-developed technology could be internationally compet-itive and began investing heavily. As a result, Koreanindustry transformed itself from a low-tech, labor-intensive exporter to one of the world's leading high-tech producers. Private sector expenditures on R&Dhave increased rapidly since 1990—almost doublingover the past decade—and are now the main source ofR&D financing. Nanotechnology, informationtechnology, and biotechnology are the main axes ofthe government’s focus on R&D at the generationstage.

Source: World Bank.

criteria, East Asian countries conditionedsubsidies on performance (often export per-formance), essentially relying on externalcompetition to discipline the market.41 Sec-ond, East Asian countries maintained high-quality bureaucracies that for the most partavoided capture by industrial interests, therebymaintaining a balance between knowledgeand involvement in productive activities andstate autonomy. In Latin America, industrial-ists often captured bureaucracies, while inmany Sub-Saharan African countries, the in-terests of industrialists or corrupt officialsdominated government interventions.

Imitation opportunities may boosttechnological diffusion, but have costsThe possibility of adopting technologiesalready elaborated in more technologicallyadvanced economies represents a fundamentaladvantage of less-advanced developingeconomies and is the basis for much of theirR&D and outreach activity. Indeed, many de-veloping countries with relatively advancedlevels of technological achievement (Brazil,China, India, the Republic of Korea, Mexico,and Malaysia), as well as Japan, initially usedan explicit policy of copying foreign technolo-gies. While this strategy proved successful to apoint, eventually the successes that theseeconomies had in the markets of their higher-technology competitors meant that thesecompetitor countries became increasinglyunwilling to share technology with them.

A substantial literature attempts to grapplewith the trade-off between the impact ofweaker intellectual property regimes and thepotential for increased technological diffusionin a host country and with the impacts thatsuch regimes might have on foreign partners’willingness to undertake FDI and licensingagreements. Although the theoretical litera-ture emphasizes the importance of intellectualproperty regimes (Lai 1998; Taylor 1994), theempirical evidence is ambiguous overall.

Some studies find no relationship betweenthe level of intellectual property rights andFDI or licensing (Primo Braga and Fink 2000;

Branstetter, Fisman, and Foley 2005; Maskusand Konan 1994). Other studies show a posi-tive effect of strong intellectual propertyregimes on FDI both in influencing locationdecisions by multinational corporations andin inducing foreign firms to invest in produc-tion rather than in distribution activities(Javorcik 2004; Lee and Mansfield 1996;Mansfield 1994; Maskus 1998).42 Some evi-dence suggests that while a stronger intellec-tual property rights regime is associated witha rise in flows of knowledge to affiliates and ininward FDI toward middle-income and largedeveloping countries, this is not the case forpoor countries (Fink 2005; Hoekman,Maskus, and Saggi 2005; Smith 2001).

Overall, the impact of intellectual propertyrights on FDI depends on the nature of thesector. Intellectual property rights appear tohave little impact on investment in lower-technology goods, such as textiles and ap-parel; services sectors, such as distribution andhotels; or in sectors where the sophisticationof the technology itself or the cost of produc-tion already serves as an effective barrier toentry. Indeed, the increased ease with whichsome products such as pharmaceuticals, chemi-cals, food additives, and software are reproduced


146

Figure 3.18 Levels of intellectual propertyprotection

Index

Source: World Bank calculations based on individual countrydata provided by Walter Park, American University.

Note: A higher score on the index indicates strongerintellectual property rights.

1970 1975 1980 1985 1990 1995 2000 20051

2

3

4

5


Low-incomecountries



may explain the rising interest in establishing in-tellectual property rights (Maskus 2000).

Perhaps reflecting such considerations, ageneral trend toward strengthening the legalprotection afforded by intellectual propertyrights has been apparent since the latter half ofthe 1980s (figure 3.18). Among developingcountries, the legal basis for such rights hasprogressed most in upper-middle-incomecountries, where levels of protection now ex-ceed the levels in high-income countries in themid-1990s. Progress in lower-middle-incomeand low-income countries has been lessmarked, reaching about the same level as inhigh-income countries in the 1990s and 1970srespectively (note, however, that the index infigure 3.18 refers to the protection offered instatutes, not in practice).

Governments can also promotetechnological progress in their ownoperations . . .In many developing countries the governmentaccounts for a significant share of productiveactivities. Using technology to increase the pro-ductivity of government operations can helpraise the efficiency of the economy as a wholeby improving health and education services(see chapter 2 and the foregoing discussion ofhuman capital); enhancing the effectivenessand reducing the costs of publicly-providedpower, telecommunications, and water andsanitation; providing approaches to regulationand tax administration that are less burden-some to firms; and demonstrating the feasibil-ity of new technology that firms can copy. Onearea where dramatic efficiency gains are possi-ble is greater use of technology in governmentprocurement. Countries that have imple-mented Internet-based procurement systemsinclude Brazil (including in some local govern-ments), Chile, Mexico, and the Philippines.43

Implementing e-procurement systems may re-quire changes in laws and policies governinggovernment operations (for example, ensuringthat government agencies can contract withforeign firms that can provide such systems.Although procurement is often a focus of

corruption and is frequently biased in favor oflocal products (most developing countries pro-vide preferential treatment to local suppliers ingovernment procurement [Kohr 2007]), theuse of advanced communications and informa-tion technologies can raise the transparencyand efficiency of government procurementand can ensure greater competition, therebycontributing to an overall improvement in thequality of government services.

More broadly, the integration of informa-tion and communications technology toolshas tremendous potential for improving accessto government information, increasing publicparticipation in government decision making,and making government services more readilyavailable to the public (World Bank Informa-tion for Development Program and Center forDemocracy and Technology 2002). In addi-tion to enhancing government efficiency, suchimprovements can help reduce costs and im-prove services to private sector firms, therebyincreasing the potential for technologicalprogress. While many industrial countrieshave used the Internet to improve local accessto information and services, its potential re-mains largely unexploited in many developingcountries. Nevertheless, some developingcountries are implementing e-government sys-tems that are as or more sophisticated thanthose used in some high-income countries(United Nations 2003). Also, the use ofelectronic systems has helped improve the effi-ciency of customs services in many countries.

A survey of case studies in developingcountries outlines some initial steps in use ofthe Internet to improve tax administrationand general services and to enhance the trans-parency and efficiency of government opera-tions (Ndou 2004). The survey underlines theimportance for the success of e-governmentinitiatives of appropriate stocktaking of thecurrent state of telecommunications networks;of raising awareness of the potential for, in-formation and communications technologybeginning with relatively small projects to testfeasibility; of stimulating collaboration amonggovernment departments; and of making


147

substantial investments in equipment and soft-ware, human capital, and appropriate organi-zational changes. Other important issues perti-nent to implementing e-government systemsinclude its coverage (comprehensive, nationalefforts may be appropriate for small countries,but may be too complex and difficult in largecountries); the ability of the enabling environ-ment to support e-government initiatives, forinstance, adequate infrastructure, an appro-priate legal framework, political commitment,and public involvement; and the availability ofstrong project management skills (Bhatnagarand Deane 2002).

. . . and encourage improved technologythrough product standardsGovernments can play a key role in boostingtechnological progress by defining and pro-moting standards for products made by pri-vate firms and by facilitating quality controlto help firms comply with standards. Goodstandards support technological progress byincreasing consistency and ensuring minimumproduct performance; facilitating the connec-tion of components in complex systems bystandardizing the interfaces between differentparts of the system; offering buyers a greaterchoice of suppliers at lower risk and lowercost and the prospect of faster and morereliable system development; and offeringmanufacturers and vendors easier entry tomarkets, economies of scale, and lower prod-uct liability risks (Yokota and Weiland 2004).The transmission of information about stan-dards can be an extremely useful channel fortechnology transfer. Implementing well-definedstandards, including testing and sanctions fornoncompliance, can be critical in maintaininga country’s reputation for quality, which is im-portant for establishing and maintaining ac-cess to global markets.

The value of a country’s reputation, the ben-efits of coordination, and the protection ofhealth and safety underline the government’srole in promoting and enforcing standards,even in competitive product markets, but thegovernment’s specific role will depend on the

product and the market structure involved.Standards may be defined by private firms withdominant market shares or agreed on througha collaborative process negotiated within thecontext of professional organizations. Theymay also be imposed by government regula-tion, or they may derive from some combina-tion of collaboration and imposition. Partici-pation in international organizations can helpdeveloping countries understand and influ-ence international standards. The Interna-tional Organization for Standardization, with132 developing country members, is a forumfor agreement on technical specifications for avariety of products.44

Coherent policies and committedgovernment leadership are critical fortechnological progress No single blueprint for technological progressexists, but most success stories have involvedstrong central leadership to ensure a consis-tent and effective policy framework that sup-ports the development and commercializationof innovations. Technological progress islargely implemented by private firms. How-ever, progress at the firm level requires gov-ernment support, elements of which includethe following: an appropriate incentivesframework, including overall political andeconomic stability and government trans-parency, along with specific technology poli-cies such as protection of intellectual propertyrights; investments in human capital, includ-ing general education and technical trainingwhere firms underinvest in training because ofthe potential mobility of trained staff; supportfor R&D of new-to-the-market technologiesbecause of difficulties in appropriating the fullbenefits from such efforts; and, where appro-priate, government interventions to overcomemarket failures involving coordination,threshold effects, and agglomeration effects(box 3.6). Most technological success stories,including Germany in the 19th century, Japanbefore and after World War II, the EastAsian miracle countries, Chile, Ireland, andIsrael have involved strong national leadership


148

and a coherent strategy for promotingtechnology.

An overall index of technologicalabsorptive capacityFigure 3.19 summarizes countries’ level oftechnological absorptive capacity andchanges over the past 10 years. The overallindex was generated following the samemethodology used to construct the index oftechnological achievement discussed in chapter2 and the index of exposure to external tech-nologies discussed earlier in this chapter. Aswas the case for technological achievement, a

two-step procedure was followed. The firststep was to estimate a separate summaryindex of the quality of the macroeconomicenvironment, financial market intermedia-tion, human capital, and governance. Thetechnical annex to chapter 2 describesthe estimation process used and the results ofthe principal components analysis in moredetail than provided here and table A2.2summarizes the individual indicators thatwent into the index.

The most important determinants of theoverall index are the governance variables(with a 37 percent weight), followed byhuman capital variables (with a 25 percent


149

Figure 3.19 Level of and recent changes in technological absorptive capacity

Technological absorptive capacity

Source: World Bank.

0

0.3

0.2

0.1

0.4

0.8

0.6

0.7

0.5




Low-incomecountries

Change in technology absorptive capacity (2000s versus 1990s)

�20

10

0

�10

20

40

30




Low-incomecountries

Index

% change

weight), and financial intermediation (with a27 percent weight), followed by the macro-economic environment (with an 11 percentweight).

Compared with the technological achieve-ment index, technological absorptive capacityis more clearly correlated with income, withless of an overlap across countries in differentincome groups. This probably results from thecomplex causal relationship that may exist be-tween technological absorptive capacity, tech-nology, and income, in which technology is afunction of technological absorptive capacityand affordability, income is a function of tech-nology, and affordability is a function of in-come, with income being both an indirectcause and an effect of technological absorptivecapacity.

Reflecting the complexity of the institu-tions that generate technological absorptivecapacity and the difficulties of reformingsome of the measures included in the index(see technical annex to chapter 2), progresshas been more limited than was the case fortechnological achievement (where the indexincreased by 160 percent for low-incomecountries). Relatively few countries improvedtheir overall score for absorptive capacity bymore than 10 percent between 1990 and2000 (the strongest negative score in low-income countries was recorded by Zimbabweand reflects mainly the deterioration inmacroeconomic and governance conditionsthere in recent years). Moreover, in contrastto technological achievement, there is littlesign of catch-up. Developing countries areimproving their technological absorptive ca-pacity at about the same rate as high-incomecountries.45

The relatively weak improvements in ab-sorptive capacity notwithstanding, the relativestrength of the technological improvement ob-served to date might be comforting. At thesame time, the relatively weak increases in tech-nological achievement in Latin America andthe Caribbean and in the Middle East andNorth Africa may reflect that technologicalprogress (and TFP growth) in those countries

has reached the limits that can be achievedfrom relatively easy adoption and imitation ofexisting technologies given current levels of ab-sorptive capacity and that further improvementmay require substantial enhancement of ab-sorptive capacity.

Conclusion

Technology diffusion in developing coun-tries depends both on access to foreign

technology (through trade, FDI, internationalmigration, and other networks) and on theability to absorb technology (as determined bythe quality of government policy and institu-tions, the stock of human capital, the efforts atR&D, and the financial system). One implica-tion of the analysis and data presented in thepreceding two chapters is that prospects forfurther technological progress in low- andmiddle-income countries are good. Over thepast 15 years, the main international channelsthrough which technology is transferred haveincreased. Developing countries’ imports ofhigh-tech goods and of capital goods haverisen relative to GDP, and their share in globalhigh-tech export markets has increased. In-flows of FDI have increased sixfold relative todeveloping countries’ output, and opportuni-ties to purchase technology have risen alongwith FDI outflows.

Simultaneously, the absorptive capacity ofdeveloping countries has been increasing, al-beit more slowly. Youth literacy rates are asmuch as 15 percentage points higher than forthe adult population. As a result, the basictechnical literacy of the population has beenincreasing, and it should continue to do so formany decades. The macroeconomic instabilitythat plagued developing countries during the1970s and 1980s has declined, and the busi-ness climate has improved, although not byas much or as uniformly as one might havehoped. Technological achievement shouldcontinue to rise over the medium term as longas these trends continue and assuming thereare no major disruptions to global trade andfinancial systems.


150

Of particular note is the speed with whichcommunications technologies are evolvingand diffusing in the developing world. Only27 years after the introduction of cell phonetechnology, mobile phones are being used invirtually every country, and penetration ratesare rising rapidly. Moreover, the range of eco-nomic activities that were once heavily depen-dent on infrastructure and that are now beingconducted using mobile phone technology isimpressive and growing daily. Already mobilephones are bringing banking, remittances, andarm’s-length financial transactions to regionsof the world that until recently were unserved.Given the pace at which things are changing,most developing countries should continueto see a rise in their ability to communicateand process information over the next fewdecades, which should help speed the diffu-sion of other technologies as well.

For middle-income countries, the relativelyrapid technological progress of the past fewyears and the improvements in both opennessand technological adaptive capability suggestthat their level of technological sophisticationshould continue to converge with that of higher-income countries. However, even the most ad-vanced of the middle-income countries will beunable to benefit fully from the new technolo-gies that are expected to become both techni-cally and economically viable over the nextseveral years because of inadequacies in theirinfrastructure (unreliable power or communi-cations systems), insufficient technical literacy,or the absence of a critical mass of scientists andengineers necessary to exploit the technology(box 3.9). For some countries, the relative slow-ness with which technological absorptivecapacity has been advancing could slow thepace of convergence as missing competenciesbecome an increasingly binding constraint onthe absorption of additional technologies.

For low-income countries, the prospects aremore complex. On average, among the low-income countries for which sufficient data areavailable to calculate recent increases intechnological achievement, convergence isoccurring and is doing so more quickly than in

middle-income countries. However, this find-ing reflects rapid progress in a few countriesand more modest performance in many othersthat are only maintaining their ground relativeto high-income countries.

Notwithstanding strong technologicalprogress in some cities and greater openness totechnological flows, the gap between existingcompetencies and those needed to convergewith technological progress in high-incomecountries is immense, especially in rural areas.Moreover, the pace at which absorptive ca-pacity is rising is disappointing. While somecountries have recorded significant increases,on average, developing countries are notcatching up to high-income countries, suggest-ing that the gap in their technology potentialis not closing. As a result, unless substantialsteps are taken to raise basic competenciesand invest in local networks that successfullydisseminate technologies and technologicalcompetencies, many of these countries are notexpected to be able to master anything morethan the simplest of forthcoming technologies(box 3.9).

One bright spot is the relatively rapid dif-fusion of some new technologies in low-income countries. Declining computing costsand prospects for rapid declines in the costof wireless Internet connections may en-hance the efficiency of ongoing economicactivities in low-income countries and mayenable them to leapfrog into more advancedtechnologies (Primo Braga, Daly, and Sareen2003).46 However, successful exploitation ofthese new technologies will require stepped-up investments in human capital and reformsin policy and regulation to provide an ap-propriate incentives structure for invest-ments in information and communicationstechnology.

A rigorous road map for achieving rapidtechnological progress does not exist. Never-theless, the evidence presented in this reportpoints to a number of conclusions, principles,and policy directions that appear likely to pro-mote technological progress and that may beable to guide policy makers. Exactly how


151


152

Arecent report by the Rand Corporation (Silber-glitt and others 2006) examines some 56 emerg-

ing technologies expected to be commercially avail-able by 2020 and evaluates in detail the 16 judged tobe most important on the basis of technical feasibil-ity, marketability, and societal impact. These applica-tions include improvements in health services (tar-geted drug delivery, improved diagnostic and surgicalmethods), in access to information (rural wirelesscommunications, quantum cryptography), and in theenvironmental sustainability of products and services(improved water purification, green manufacturing,hybrid vehicles). It then examines the technical basea country requires to make effective use of each tech-nology and the likelihood that each of 80 representa-tive economies, including both high-income and de-veloping countries from every region in the world,will be able to exploit these technologies by 2020.

While many countries are expected to be able totake advantage of some of the simpler-to-use tech-nologies, a wide range of countries are not expectedto be able to do so because they lack the requiredtechnological infrastructure, because their populationis not sufficiently technically literate, or because acritical mass of scientists and engineers is not present

Box 3.9 Technology in 2020to exploit the technology (see the table). The reportfinds that most high-income countries will be able toadopt and exploit all the technologies effectively. Asecond group of countries, including China, India,Russia, and the countries of Eastern Europe, are foundto have a considerable level of scientific and techno-logical proficiency in specific applications. However,barriers to technology adaptation are likely to limittheir ability to take advantage of the most sophisti-cated network applications. A third group of middle-income countries, which consists of several LatinAmerican countries, Indonesia, South Africa, andTurkey, lacks more prerequisites and is therefore ex-pected to exploit fewer of these technologies. A finalgroup that comprises most of the world’s poorestcountries, including most of the countries of Africa,the Middle East, and Oceania, is projected to makeuse of only the simplest of the new technologies.

This analysis provides a useful snapshot of theprospects for technological progress based on currentdata. However, it does not incorporate the potential fordynamic improvements in technological progress, forexample, through the rapid dissemination of existingnew technologies, which could rapidly improve devel-oping countries’ ability to absorb new technologies.

Technological adaptive capacity may restrict the diffusion of future technologies

Req

uir

es in

crea

sed

tec

hn

olo

gic

al s

op

his

tica

tio

n

Technology application

Cheap solar energy ¸

¸

¸

¸

¸

Rural wireless communication

Genetically modified crops

Filters and catalysts

Cheap autonomous housing

Rapid bioassays

Green manufacturing

Ubiquitous RFID tagging

Hybrid vehicles

Targeted drug delivery

Improved diagnostic and surgical techniques

Quantum cryptography

Ubiquitous information access

Tissue engineering

Pervasive sensors

Wearable computers

Latin America,South Africa,

Turkey, Indonesia

Technologies likely to be mastered by 2020 (¸)

China, India,Russia,

Eastern Europe Industrial countries

Most of Africa,Middle East,

Oceania

Source: Silberglitt and others 2006.

Note: RFID � radio-frequency identification.

¸

¸

¸

¸

¸

¸

¸

¸

¸

¸

¸

¸

¸

¸

¸

¸

¸

¸

¸

¸

¸

¸

¸

¸

¸

¸

¸

¸

¸

¸

¸

¸

¸

¸

¸

¸

¸

much weight to give to each of these conclu-sions and how they interact depends on spe-cific country circumstances and should be thesubject of future research. These policy direc-tions include the following:

• Openness to external technologiesthrough foreign trade, FDI, diasporas,and other international networks is criti-cal for technological progress for bothlow- and middle-income countries,where most progress occurs through theadoption, adaptation, and assimilationof preexisting but new-to-the-market ornew-to-the-firm technologies.

• The capacity of firms or individuals to usea technology depends critically on thebasic technological literacy of workers andconsumers. The level of technological lit-eracy, in turn, depends on the government’scapacity to deliver a quality education tothe largest number of people possible.

• The preeminent vehicles for the dissemina-tion and diffusion of technology in a mar-ket economy are firms and entrepreneurs.Their success in doing so depends on theirability to undertake and expand new activ-ities. This requires a stable macroeconomicenvironment, together with a regulatoryenvironment that effectively enforcesproperty rights and the rule of law, doesnot excessively restrict firms’ ability to hireand fire, and does not impose excessiveregulatory or financial burdens.

• The capacity of firms or individuals totake advantage of a technology can beconstrained by affordability and by li-quidity, thereby placing a premium on theefficiency with which the financial systemintermediates between savers and bor-rowers both domestically and abroad.

• Given the existence of market failures,the government has a role to play in as-sisting firms to learn how to adapt,adopt, and market new technologies. Inaddition to focusing on R&D in new-to-the-market technologies, applied R&Dagencies need to emphasize outreach,

testing, marketing, and disseminationactivities. The huge rural–urban divide inboth technology and absorptive capacityin many developing countries underlinesthe importance of such activities to in-clusive development.

• The government can also have an impor-tant impact on economic progress by in-tegrating new technology into its ownoperations, including in the provision ofeducation, health, and publicly-providedinfrastructure; in the procurement ofgoods and services; in the provision ofinformation and in fostering public dia-logue; and in the definition of standardsfor commercial products.

• The principal challenge facing many low-income countries is not their access totechnology, but their absorptive capacity,including physical, human, and institu-tional capacity; their limited financialresources; and the extent to which theirsocial and political environments aresupportive of entrepreneurship, invest-ment, and technological progress.

These conclusions highlight the critical roleof the government in establishing the generalconditions that support rapid technologicalprogress, in helping to overcome market fail-ures that constrain innovations by firms, andin providing (and purchasing) high-qualitygoods and services. Countries that haveachieved sustained and rapid technologicalprogress have generally benefited from com-mitted national leadership that follows coher-ent development policies, although the natureof these policies—in particular, the degree ofpublic sector intervention in private markets—has varied enormously.

Notes 1. The econometric evidence is mixed. Harrison

(1994) for Côte d’Ivoire and Haddad, de Melo, andHorton (1996) for Morocco find no statistically signif-icant impact of import penetration on productivity fol-lowing trade liberalization. Nishimizu and Page (1982)for the former Yugoslavia; Tybout, de Melo, andCorbo (1991) for Chile; and Tybout and Westbrook


153

(1995) for Mexico find a positive relationship betweenimport penetration and firm efficiency. Grether (1999)and van Wijnbergen and Venables (1993) for Mexico,Earle and Estrin (2001) for Russia, Falk and Dierking(1995) for Poland, Levinsohn (1993) for Turkey, andRoberts (1996) for Colombia find a positive impact ofimport penetration on either industry markups or mea-sured labor productivity.

2. Keller (1998) casts doubt on these results byshowing that the relationship also holds for randomlygenerated import shares, but Lumenga-Neso, Olarreaga,and Schiff (2005) find that imports from countries thatimport from other R&D centers are positively relatedto productivity and that these indirect spillovers are atleast as important as direct spillovers.

3. Lall (2000, 2001) identifies four categories ofproducts: resource-based products; low-tech products,which include textiles and fashion; medium-tech prod-ucts; and high-tech products. According to this classifi-cation, which is defined using SITC 3-digit rev. 2,technological products do not include agriculturalproducts, moderately processed food products, to-bacco products, minerals, construction materials, andenergy products.

4. Available data provide only a rough indication ofthe sophistication of economic activity, because ascer-taining the level of sophistication of the capital goodsimported is not possible. Also some countries may im-port relatively sophisticated capital goods for use in en-clave production (for example, oil and minerals) withlittle spillover into the rest of the economy.

5. Chandra and Kolavalli (2006) cite importantspillover effects from exporting electronics and software.

6. The contradictory evidence from case studiesand econometric studies may be due to the differentimpacts of exports across industries and countries, aswell as difficulties inherent in classifying firms (somestudies classify exporters on the basis of surveys withyes-no answers rather than measuring the volume ofexports) (Keller 2004). Also the argument for tech-nology transfers through exports refers only to someexports—namely, new products or products that haveevolved over time, and export statistics may not cap-ture such subtleties. Firms that export the same prod-uct that is not subject to significant upgrading may notbenefit from spillovers. If some firms improve theirproductivity through exports and some do not, andavailable data do not permit distinguishing betweenthese firms, measuring the extent of productivity im-provements over time may be difficult. For specific ex-amples, see Tybout and Westbrook (1996), who findthat trade liberalization in Mexico benefited exportersbecause of declines in prices of imported inputs, buthad no effect on productivity. Soderbom and Teal

(2000) find that Ghanaian firms with higher technicalefficiency become exporters. Isgut (2001) finds thatexporting firms in Colombia had higher labor produc-tivity than nonexporters three years before entering theexport market, but that afterward there is no differencein the growth of labor productivity of exporters as com-pared with nonexporters. Fafchamps, Zeufack, andEl Hamine (2002) find that in Morocco, more produc-tive firms move into exports. However, after initiatingexports they do not achieve more rapid reductions inproduction costs than nonexporters, although they dolearn to improve product design to suit foreign markets.

7. Following Lall (2001), the nomenclature used isSITC 3-digit, Rev. 2.

8. Costa Rica has emerged as a high-tech platformfor foreign investors and increased its world marketshare of high-tech products from 0.01 to 0.20 percentbetween the mid-1990s and 2002–04.

9. Between 1992 and 2003, developing countriesmade some 2,563 favorable changes to national lawsand regulations relating to FDI. The most frequentchanges concerned FDI promotion and incentives(855), sectoral restrictions (497), operational condi-tions (406), guarantees (304), and corporate regula-tions (153). During the same period, 113 developingcountries became members of the World Trade Orga-nization, which required the elimination of many re-strictions and impediments to FDI, particularly in theservices sector (World Bank 2004a).

10. The world’s largest R&D investors conductedan average of 28 percent of their R&D outside theirhome territory in 2003 (UNCTAD 2005).

11. This section builds on many studies of FDIspillovers that have identified possible channels fortechnology transfers and knowledge spilloversthrough FDI (Görg and Greenaway 2004; Görg andStrobl 2001; Javorcik 2007; Lipsey 2002; Moran2007; Saggi 2002).

12. Javorcik (2004) finds that the TFP of Lithuan-ian firms is positively correlated with the extent of po-tential contacts with multinational customers in down-stream sectors. Blalock and Gertler (forthcoming) andKugler (2006) find strong evidence that vertical supplychains were a channel for technology transfers inColombian and Indonesian manufacturing sectors.Swinnen and others (2006) show that investments byforeign companies in processing and retailing in East-ern Europe have introduced higher standards, which inturn led to significant efficiency gains by suppliers.

13. Javorcik (2007) documents the increased com-petitive pressures from foreign entry in Czech and Lat-vian firms, and the McKinsey Global Institute (2003)cites case studies where competition is a key factor indiffusing FDI-introduced innovations.


154

14. Ayyagari and Kosova (2006), using Czech FDIdata for 1994 to 2000, show that spillovers vary sub-stantially across industries. Although service industriesbenefited from huge FDI spillover effects through bothhorizontal and vertical channels, manufacturing indus-tries did not show any significant positive spillover ef-fects from FDI.

15. Belderbos, Capannelli, and Fukao (2000) findthat the proportion of inputs sourced locally by Japan-ese multinationals increases with the number of yearsof operation in a given host country.

16. These data are incomplete, as only 90 of 150developing countries (on average across 1999–2006)reported royalty and license fee payments. The datamay also overstate payments for technology transfer, asdeveloping countries with mineral or oil investmentsabroad may report the payment of substantial royaltiesthat represent fees for extraction rights rather than forthe purchase of technology.

17. Between 1990 and 2000, the number of tertiary-educated emigrants from developing countries thatresided in OECD countries rose from 19.1 million to37.8 million (Docquier and Marfouk 2004).

18. Even though a majority of Argentine doctoralgraduates in the United States prefer to remain in thehost country, most respondents in a survey of high-skilled Argentine diaspora members in Europe, theUnited States, and elsewhere expressed their willing-ness and interest in helping develop science, technol-ogy, and education in their home country (Kuznetsov,Nemirovsky, and Yoguel 2006).

19. Of these technologically sophisticated émigrés,56 percent were born in Asia, with Latin America andthe Caribbean accounting for another 15 percent(Kannankutty and Burrelli 2007).

20. Agrawal, Kapur, and McHale (2007), usingpatent data, find evidence of the influence of the dias-pora in technology transfers to home countries.

21. The Mexican Ministry of Science and Technol-ogy views the presence of 1 million tertiary-educatedMexican migrants in the United States, with an esti-mated 400,000 in managerial positions, as a unique, un-explored opportunity for knowledge transfers(Kuznetsov 2006). Emigrants from China and Indiawere running almost 30 percent of Silicon Valley’s(California) technology businesses by the end of the1990s (Saxenian 2000, 2002). In addition, 25 percent ofall engineering and technology companies started in theUnited States during 1995–2005 had a foreign-born per-son as a key founder (Wadhwa and others 2007).

22. Page and Plaza (2006) argue that technologytransfer by migrants takes place through several chan-nels: (a) licensing agreements between diaspora-ownedor managed firms in host countries and firms in

sending countries, (b) knowledge spillovers when re-turning migrants assume managerial positions in theirhome country, (c) networks of diaspora researchersand scientists performing research directed at the needsof their country of origin, (d) “virtual” return throughextended visits and electronic communication in fieldssuch as medicine and engineering, and (e) return topermanent employment in the country of origin aftergaining work experience in the host country.

23. Estimates suggest that a 10 percent increase inskilled migrant stock in the United States is associatedwith a 4 percent increase in the flow of FDI (in currentdollars) to the home country (Mattoo, Özden, andNeagu 2005).

24. This result is supported by work on high-income countries that shows immigrant ties have beenimportant determinants of U.S and Canadian bilateraltrade (Gould 1994; Head and Ries 1998; Wagner,Head, and Ries 2002).

25. Kuznetsov (2007) argues that diasporas can actas global search networks by leveraging their contex-tual knowledge of their home countries’ economy andinstitutions to identify untapped resources and oppor-tunities, such as research capabilities, availability oftechnical manpower, and business-friendly local gov-ernments.

26. Among members of the Philippines Brain GainNetwork, 35 percent have a master’s degree and 23 per-cent hold a doctorate, while 49 percent of the membersof the South African Network of Skills Abroad have amaster’s degree and another 30 percent have a doctor-ate (Brown 2000).

27. The Taiwanese diaspora and returning migrantswere active conduits for technology transfers. For ex-ample, in 2000, 113 out of 289 companies at the Hin-schu Science-Based Industrial Park in Taiwan, China,were started by U.S.-educated Taiwanese (O’Neil 2003).

28. Countries with strong institutions such asChile, the Republic of Korea, and Scotland have beenable take advantage of their high-skilled diasporas,while others such as Argentina, Armenia, andColombia have not succeeded as well despite havingmany programs (Kuznetsov 2006).

29. Finding a relevant, available indicator of thesize of the diaspora to include in the index proved dif-ficult. The data series used included FDI net inflows,royalties and license fee payments, imports of high-tech goods, imports of capital goods, and imports ofintermediate goods—all as a percent of GDP. Importsof intermediate goods and net FDI inflows have thelargest weight in the calculation, accounting for morethan half the total.

30. The productivity benefits from the adoption ofnew technology are best realized in the context of low


155

inflation, stable exchange rates, sustainable governmentfinances, and positive income growth (Pack 2006).

31. Liu and Tybout (1996) and Roberts and Tybout(1997) present data from Chile, Colombia, andMorocco confirming that the entry and exit of firmsmakes an important contribution to productivitygrowth.

32. Data are taken from the World Bank’s DoingBusiness Web site (http://www.doingbusiness.org).

33. See Keller (2004) for a survey of the economicliterature on this topic. Education levels are typicallyimportant in empirical studies of cross-country differ-ences in growth rates and in labor productivity (Chenand Dahlman 2004), but these studies do not deter-mine the channel through which human capital con-tributes to growth.

34. In some countries, the limited rise in public ex-penditures on education may have been balanced by in-creases in private expenditures.

35. Ayyagari, Demirgüç-Kunt, and Maksimovic(2007) find a positive correlation between financialmarket depth (proxied by credit to the private sector asa percent of GDP) and R&D intensities.

36. These requirements are generally imposed toreduce volatility in these often thin markets and tobolster investor confidence in the safety of investingin listed firms. While they enable some well-established firms to access global capital markets byproviding the exchanges with legitimacy, they excludemore speculative firms less well. Similarly, the use ofAmerican deposit receipts, which make investing inemerging-market firms easier for foreign investors, islargely restricted to large, well-established, andmainly export-oriented firms (Claessens andSchumkler 2007).

37. The OECD (2005) defines venture capital “ascapital provided by firms who invest alongside man-agement in young companies that are not quoted onthe stock market. The objective is high return from theinvestment. Value is created by the young company inpartnership with the venture capitalist’s money andprofessional expertise.” The flow of venture capitalfrom the investor to a start-up company and back canbe thought of as a cycle that runs through severalphases. The International Finance Corporation moni-tors about 90 venture capital funds active in develop-ing countries.

38. Internationally comparable data have beenavailable only since 1997.

39. See the discussion from the World BankGlobal Forum on Science, Technology, and Innovation,February 13–15, 2007, in Washington, DC. http://go.worldbank.org/DWODQ7E3E0.

40. Note that an OECD (2003) study found thatfiscal incentives to support R&D in rich countries were

not very effective on average, with success being heav-ily dependent on the design of the tax measures.

41. Such programs are more difficult to implementtoday in light of World Trade Organization restrictionson export subsidies (Rodrik 2004).

42. The intellectual property regime is only oneconsideration among many, including various localmarket and sector characteristics, that enter into multi-national corporations’ decisions on how to deploytechnology internationally (Mansfield 1994, 1995).

43. See World Resources Institute Digital Dividend(http://www.digitaldividend.org) and the WorkingGroup on E-Government in the Developing World(http://www.pacificcouncil.org/pdfs/e-gov.paper.f.pdf).

44. See http://www.iso.org.45. There is a slightly inverted U shape to the dis-

tribution of improvements in technological absorptivecapacity, with high-income countries recording a9.1 percent improvement, compared with 9.4 percentfor upper-middle-income countries, 9.8 percent forlower-middle-income countries, and 8.6 percentamong those low-income countries for which data areavailable.

46. The development of simple, low-cost computersand the spread of open-source technology has alreadyenhanced the affordability of new technologies forlow-income countries.

ReferencesAcemoglu, Daron, and Fabrizio Zilibotti. 2001. “Pro-

ductivity Differences.” Quarterly Journal of Eco-nomics 116 (2): 563–606.

Agrawal, Ajay K., Devesh Kapur, and John McHale.2004. “Defying Distance: Examining the Influenceof the Diaspora on Scientific Knowledge Flows.”Unpublished paper. Queen’s School of Business,Queen’s University, Ontario. Presented at theFourth Annual Roundtable for Engineering Entre-preneurship Research, Georgia Institute of Tech-nology, College of Management, December 4.

______. 2007. “Brain Drain or Brain Bank? The Impactof Skilled Emigration on Poor-Country Innova-tion.” Unpublished paper. University of Toronto,Rotman School of Management, Ontario.

Aitken, Brian J., and Ann E. Harrison. 1999. “DoDomestic Firms Benefit from Foreign DirectInvestment? Evidence from Venezuela.” AmericanEconomic Review 89 (3): 605–18.

Antras, Pol, Mihir A. Desai, and C. Fritz Foley. 2007.“Multinational Firms, FDI Flows, and ImperfectCapital Markets.” Working Paper 12855. NationalBureau of Economic Research, Cambridge, MA.

Arrow, Kenneth. 1962. “The Economic Implications ofLearning by Doing.” Review of Economic Studies29 (3): 155–73.


156

Audretsch, David B., and Maryann P. Feldman. 1996.“R&D Spillovers and the Geography of Innova-tion and Production. American Economic Review86 (4): 253–73.

_____. 2004. “Knowledge Spillovers and the Geogra-phy of Innovation.” In Handbook of Urban andRegional Economics, vol. 4, ed. J. V. Hendersonand J. F. Thisse, 2,713–39. New York: NorthHolland.

Avnimelech, G., and M. Teubal, 2004: “Venture Capi-tal Policy in Israel: A Comparative Analysis andLessons for Other Countries.” In Financial Sys-tem, Corporate Investment in Innovation andVenture Capital, ed. M. Sunil A. and Bartzokas.Cheltenham, U.K.: Edward Elgar.

Ayyagari, M., and R. Kosova. 2006. Does FDI Facili-tate Domestic Entrepreneurship? Evidence fromthe Czech Republic. Washington, DC: GeorgeWashington University, School of Business.

Ayyagari, Meghana, Asli Demirgüç-Kunt, and VojislavMaksimovic. 2007. “Firm Innovation in Emerg-ing Markets: The Roles of Governance and Fi-nance.” Policy Research Working Paper Series4157. World Bank, Washington, DC.

Barro, Robert J., and Jong Wha Lee. 2000. “Interna-tional Data on Educational Attainment: Updatesand Implications.” Center for International De-velopment Working Paper No. 42, April.

Beamish, P. W. 1988. Multinational Joint Ventures inDeveloping Countries. London: Routledge.

Beck, Thorsten, Asli Demirgüç, and Ross Levine. 2000.“A New Database on Financial Development andStructure.” World Bank Economic Review 14,597–605.

Behrman, Jere R., Andrew D. Foster, Mark R. Rosen-zweig, and Prem Vashishtha. 1999. “Women’sSchooling, Home Teaching, and EconomicGrowth.” Journal of Political Economy 107 (4):632–714.

Belderbos, Rene, Giovanni Capannelli, and KyojiFukao. 2000. “The Local Content of JapaneseElectronics Manufacturing Operations in Asia.”In The Role of Foreign Direct Investment in EastAsian Economic Development, ed. Takatoshi Itoand Anne O. Krueger, 9–41. Chicago: Universityof Chicago Press.

Bell, C., R. Bruhns, and H. Gersbach. 2006. “Eco-nomic Growth, Education, and AIDS in Kenya:A Long-Run Analysis.” Policy Research WorkingPaper 4025. World Bank, Washington, DC.

Bell, C., S. Devarajan, and H. Gersbach. 2004. “Think-ing about the Long-Run Economic Costs ofAIDS.” In The Macroeconomics of HIV/AIDS,ed. M. Haacker, 96–133. Washington, DC: Inter-national Monetary Fund.

Bhatnagar, Subhash, and Arsala Deane. 2002. “Build-ing Blocks of E-Government: Lessons fromDeveloping Countries.” Poverty Reduction andEconomic Management Note 91. World Bank,Washington, DC.

Blalock, Garrick, and Paul J. Gertler. Forthcoming.“Welfare Gains from Foreign Direct Investmentthrough Technology Transfer to Local Suppliers.”Journal of International Economics.

Blomstrom, M., A. Kokko, and M. Zejan. 2000. For-eign Direct Investment. Firm and Host CountryStrategies. London: Macmillan.

Borensztein, E., J. de Gregorio, and J.-W. Lee. 1998.“How Does Foreign Direct Investment AffectEconomic Growth?” Journal of InternationalEconomics 45 (1): 115–35.

Branstetter, Lee G., Raymond Fisman, and C. FritzFoley. 2005. “Do Stronger Intellectual PropertyRights Increase International Technology Trans-fer? Empirical Evidence from U.S. Firm-LevelPanel Data.” Working Paper 11516. NationalBureau of Economic Research, Cambridge, MA.

Brinkerhoff, Jennifer M. 2006a. “Diasporas, Mobiliza-tion Factors, and Policy Options.” In ConvertingMigration Drains into Gains: Harnessing theResources of Overseas Professionals, ed. ClayWescott and Jennifer Brinkerhoff, 127–53.Manila: Asian Development Bank.

______. 2006b. “Diasporas, Skills Transfer, and Re-mittances: Evolving Perceptions and Potential.”In Converting Migration Drains into Gains: Har-nessing the Resources of Overseas Professionals,ed. Clay Wescott and Jennifer Brinkerhoff, 1–32.Manila: Asian Development Bank.

Broadman, Harry G., and Francesca Recanatini.2005. “Is Russia Restructuring? New Evidenceon Job Creation and Destruction.” PolicyResearch Working Paper 2641. World Bank,Washington, DC.

Brown, Mercy. 2000. “Using the Intellectual Diasporato Reverse the Brain Drain: Some Useful Exam-ples.” In Brain Drain and Capacity Buildingin Africa, ed. S. J. M. Tapsoba, S. Kassoum, P. V. Houenou, B. One, M. Sethi, and J. Ngu. Eco-nomic Commission for Africa, International De-velopment Research Centre, and InternationalOrganization for Migration.

Bureau of Economic Analysis. 2007. “Operations ofMultinational Companies.” U.S. Department ofCommerce. http://www.bea.gov. Accessed Sep-tember 1, 2007.

Caselli, Francesco, and Wilbur John Coleman II. 2001.“Cross-Country Technology Diffusion: The Caseof Computers.” American Economic Review Pa-pers and Proceedings 91 (2): 328–35.


157

Chandra, Vandana, ed. 2006. Technology, Adaptationand Exports: How Some Developing CountriesGot It Right. Washington, DC: World Bank.

Chandra, Vandana, and Shashi Kolavalli. 2006. “Tech-nology, Adaptation and Exports: How SomeDeveloping Countries Got It Right.” In Technol-ogy, Adaptation and Exports: How Some Devel-oping Countries Got It Right, ed. Vandana Chan-dra, 1–41. Washington, DC: World Bank.

Chaparro, Fernando, Hernan Jaramillo, and VladimirQuintero. 2006. “Promise and Frustration ofDiaspora Networks: Lessons from the Networkof Colombian Researchers Abroad.” In DiasporaNetworks and the International Migration ofSkills, ed. Yevgeny Kuznetsov, 187–98. Washing-ton, DC: World Bank Institute.

Chen, Derek H. C., and Carl J. Dahlman. 2004.“Knowledge and Development: A Cross-SectionApproach.” Policy Research Working Paper3366. World Bank, Washington, DC.

Claessens, Stijn, and Sergio L. Schumkler. 2007.“International Financial Integration throughEquity Markets: Which Firms from Which Coun-tries Go Global?” Policy Research Working Paper4146. World Bank, Washington, DC.

Coe, David T., Elhanan Helpman, and Alexander W.Hoffmaister. 1997. “North-South R&DSpillovers.” Economic Journal 107 (January):134–49.

Coelho, S.T., and J. Goldemberg. 2004. “AlternativeTransportation Fuels: Contemporary Case Stud-ies.” In Encyclopedia of Energy, Vol. 1. Elsevier.

Cohen, Wesley M., and Daniel A. Levinthal. 1989. “In-novation and Learning: The Two Faces of R&D.”Economic Journal 99 (September): 569–96.

Correa, Carlos M. 2003. “Can the TRIPS AgreementFoster Technology Transfer to Developing Coun-tries?” Unpublished manuscript. Duke University,Durham, NC.

Dahlman, Carl J., Bruce Ross-Larson, and Larry E.Westphal. 1987. “Managing Technological Devel-opment: Lessons from the Newly IndustrialisingCountries.” World Development 15 (6): 759–75.

de Ferranti, David, Guillermo Perry, Indermit Gill,J. Luis Guasch, William F. Maloney, CarolinaSanchez-Paramo, and Norbert Schady. 2003.Closing the Gap in Education and Technology.Latin American and Caribbean Studies.Washington, DC: World Bank.

Docquier, Frederic, and Abdeslam Marfouk. 2004.“Measuring the International Mobility of SkilledWorkers (1990–2000).” Policy Research Work-ing Paper Series 3381. World Bank, Washington,DC.

Earle, John S., and Estrin, Saul. 2001. “Privatizationand the Structure of Enterprise Ownership.” InRussia’s Post-Communist Economy, ed. BrigitteGranville and Peter Oppenheimer, 149–68.Oxford, U.K.: Oxford University Press.

Easterly, William, and Yaw Nyarko. 2005. “Is theBrain Drain Good for Africa?” Unpublishedpaper. New York University, New York.

Eaton, Jonathan, and Samuel Kortum. 1996. “Trade inIdeas: Patenting and Productivity in the OECD.”Journal of International Economics 40: 251–78.

______. 2001. “Trade in Capital Goods.” EuropeanEconomic Review 45 (7): 1195–235.

Enos, John, and W. H. Park. 1987. The Adoption andDiffusion of Imported Technology. London:Croom Helm.

Fafchamps, Marcel, Albert Zeufack, and Said ElHamine. 2002. “Learning to Export: Evidencefrom Moroccan Manufacturing.” Policy ResearchWorking Paper 2827. World Bank,Washington, DC.

Fagerberg, Jan. 1988. “Why Growth Rates Differ.”In Technical Change and Economic Theory, ed. Giovanni Dosi, Christopher Freeman, RichardNelson, Gerald Silverber, and L. Soete. London:Pinter Publishers.

Fajnzylber, Pablo, and J. Humberto López. 2007. Closeto Home: The Development Impact of Remit-tances in Latin America. Washington, DC: WorldBank.

Falk, J., and L. Dierking, eds. 1995. Public Institutionsfor Personal Learning: Establishing a ResearchAgenda. Washington, DC: American Associationof Museums.

Fink, Carsten. 2005. “Intellectual Property Rights andU.S. and German International Transactions inManufacturing Industries.” In Intellectual Prop-erty Rights and Development: Lessons from Re-cent Economic Research, ed. Carsten Fink andKeith Maskus, 75–110. Washington, DC: WorldBank.

Finn, Michael G. 2003. “Stay Rates of Foreign Doc-torate Recipients from U.S. Universities, 2001.”Unpublished paper. Oak Ridge Institute forScience and Education, Science and EngineeringEducation Program, Oak Ridge, TN.

Fosfuri, A., M. Motta, and T. Ronde. 2001. “ForeignDirect Investment and Spillovers through Work-ers’ Mobility.” Journal of International Econom-ics 53: 205–22.

Gereffi, Gary. 1999. “International Trade and Indus-trial Upgrading in the Apparel CommodityChain.” Journal of International Economics 48:37–70.


158

Gill, Indermit, and Homi Kharas. 2007. An East AsianRenaissance. Washington, DC: World Bank.

Girma, Sourafel, and Katherine Wakelin. 2001. “AreThere Regional Spillovers from FDI in the UK?”Research Paper 2000/16. University of Notting-ham, School of Economics, Centre for Researchon Globalization and Labour Markets, Notting-ham, U.K.

Glaeser, Edward L., Hedi D. Kallal, Jose A.Scheinkman, and Andrei Shleifer. 1992. “Growthof Cities.” Journal of Political Economy 100 (6):1126–52.

Glass A. J., and K. Saggi. 2002. “Multinational Firmsand Technology Transfer.” Scandinavian Journalof Economics 104 (4): 495–513.

Goldman, Melvin, and Henry Ergas. 1997. “Technol-ogy Institutions and Policies: Their Role in Devel-oping Technological Capability in Industry.”Technical Paper 383. World Bank, Washington,DC.

Görg, Holger, and David Greenaway. 2004. “MuchAdo about Nothing? Do Domestic Firms ReallyBenefit from Foreign Direct Investment?” WorldBank Research Observer 19 (2): 171–97.

Görg, Holger, and Eric Strobl. 2001. “MultinationalCompanies and Productivity Spillovers: A Meta-Analysis.” Economic Journal 111: 723–39.

Gould, David. 1994. “Immigrant Links to the HomeCountry: Empirical Implications for US BilateralTrade Flows.” Review of Economics and Statis-tics 76 (2): 302–16.

Grether, J. M. 1999. “Determinants of TechnologicalDiffusion in Mexican Manufacturing: A Plant-Level Analysis.” World Development 27 (7):1287–98.

Guellec, Domininque, and Bruno van Pottelsberge dela Potterie. 2004. “From R&D to ProductivityGrowth: Do the Institutional Settings and theSource of Funds of R&D Matter?” Oxford Bul-letin of Economics and Statistics (July): 353–78.

Guerrieri, P., and C. Milana. 1998. “High-TechnologyIndustries and International Competition.” InTrade Growth and Technical Change, ed. D.Archibugi and J. Michie, 188–207. Cambridge,U.K.: Cambridge University Press.

Gupta, Sanjeev, Catherine Pattillo, and Smita Waugh.2007. “Impact of Remittances on Poverty and Fi-nancial Development in Sub-Saharan Africa.”Working Paper 07/38. International MonetaryFund, Washington, DC.

Haddad, Mona, Jaime de Melo, and Brendan Horton.1996. “Morocco 1984–1989: Trade Liberaliza-tion, Exports and Industrial Performance. In In-dustrial Evolution in Developing Countries, ed.

Mark J. Roberts and James R. Tybout. NewYork: Oxford University Press.

Handoussa, Heba, Mieko Nishimizu, and John Page.1986. “Productive Change in Egyptian PublicSector Industries after the ‘Opening,’ 1973–79.”Journal of Development Economics 20 (1): 53–73.

Hanushek, Eric A., and Ludger Woessmann. 2007.“The Role of Education Quality in EconomicGrowth.” Working Paper 4122. World Bank,Washington, DC.

Harrison, Anne E. 1994. “Productivity, ImperfectCompetition, and Trade Reform: Theory and Ev-idence.” Journal of International Economics 36(1–2): 53–73.

Head, Keith, and John Ries. 1998. “Immigration andTrade Creation: Econometric Evidence fromCanada.” Canadian Journal of Economics 31 (1):47–62.

Helms, Brigit. 2006. Access for All: Building InclusiveFinancial Systems. Washington, DC: ConsultativeGroup to Assist the Poor.

Hernandez, Zenaida. 2004. “Industrial Policy in EastAsia: In Search of Lessons.” Background Paperfor the World Development Report 2005. WorldBank, Washington, DC.

Hobday, Michael. 1995. Innovation in East Asia: TheChallenge to Japan. Aldershot, U.K.: EdwardElgar.

Hoekman, Bernard, and Beata Smarzynska Javorcik.2006. “Lessons from Empirical Research on In-ternational Technology Diffusion through Tradeand Foreign Direct Investment.” In Global Inte-gration and Technology Transfer, ed. BernardHoekman and Beata Smarzynska Javorcik, 1–28.New York: Palgrave MacMillan.

Hoekman, Bernard, Keith E. Maskus, and KamalSaggi. 2005. “Transfer of Technology to Devel-oping Countries: Unilateral and MultilateralPolicy Options.” World Development 33 (10):1587–1602.

Hou, Chi-Ming, and San Gee. 1993. “National SystemsSupporting Technical Advance in Industry: TheCase of Taiwan.” In National Innovation Systems:A Comparative Analysis, ed. Richard R. Nelson,384–413. New York: Oxford University Press.

Isern, Jennifer, William Donges, and Jeremy Smith.2006. “Making Money Transfers Work for Micro-finance Institutions: A Technical Guide to De-veloping and Delivering Money Transfers.” Draftfor public review. Consultative Group to Assistthe Poor, Washington, DC.

Isgut, Albert E. 2001. “What’s Different About Ex-porters? Evidence from Colombian Manufactur-ing.” JournalofDevelopmentStudies37(5):57–82.


159

Jaffe, A. B. 1989. “Real Effects of Academic Research.”American Economic Review 79 (5): 957–70.

Javorcik, Beata Smarzynska. 2004. “Does ForeignDirect Investment Increase the Productivity ofDomestic Firms? In Search of Spillovers throughBackward Linkages.” American EconomicReview 94 (3): 605–27.

______. 2006. “Technological Leadership and theChoice of Entry Mode by Foreign Investors.” InGlobal Integration and Technology Transfer, ed.B. Hoekman and B. Javorcik, 179–206. NewYork: Palgrave Macmillan.

______. 2007. “Foreign Direct Investment and Interna-tional Technology Transfer: What Do We Know?What Would We Like to Know?” Unpublishedpaper. World Bank, Washington, DC.

Javorcik, Beata Smarzynska, Wolfgang Keller, andJames Tybout. 2006. “Openness and IndustrialResponses in a Wal-Mart World: A Case Study ofMexican Soaps, Detergents, and Surfactant Pro-ducers.” Working Paper 12457. National Bureauof Economic Research, Cambridge, MA.

Javorcik, Beata Smarzynska, and Mariana Spatareanu.2005. “Disentangling FDI Spillover Effects: WhatDo Firm Perceptions Tell Us?” In Does ForeignDirect Investment Promote Development?, ed.Theodore Moran, Edward Graham, and MagnusBlomstrom, 45–72. Washington, DC: Institute forInternational Economics.

Jordan, Mary. 2006. “New Conductors Speed GlobalFlows of Money: Cellphones Make TransfersFaster, Cheaper.” Washington Post, October 3,p. A01.

Kannankutty, Nirmala, and Joan Burrelli. 2007. “WhyDid They Come to the United States? A Profile ofImmigrant Scientists and Engineers.” InfoBriefNSF07-324, Science Resources Statistics, Direc-torate for Social, Behavioral, and Economic Sci-ences, United States National Science Foundation,Arlington, VA. June.

Kapur, Devesh. 2001. “Diasporas and TechnologyTransfer.” Journal of Human Development 2 (2):265–86.

Kapur, Devesh, and John McHale. 2005a. Give UsYour Best and Brightest: The Global Hunt forTalent and Its Impact on the Developing World.Washington, DC: Center for Global Development.

______. 2005b. “Sojourns and Software: Internation-ally Mobile Human Capital and High-Tech In-dustry Development in India, Ireland and Israel.”In Underdogs to Tigers: The Rise and Growth ofthe Software Industry in Some EmergingEconomies, ed. Ashish Arora and AlfonsoGambardella. Oxford University Press.

Kaufmann D., A. Kraay, and M. Mastruzzi. 2007.Governance Matters VI: Governance Indicators-for 1996–2006. Washington, DC: World Bank.

Keller, Wolfgang. 1998. “Are International R&DSpillovers Trade Related? Analyzing Spilloversamong Randomly Matched Trade Partners.”European Economic Review 42: 1469–81.

______. 2004. “International Technology Diffusion.”Journal of Economic Literature 42 (September):752–82.

Khadria, Binod. 2004. “Human Resources in Scienceand Technology in India and the InternationalMobility of Highly Skilled Indians.” Science,Technology and Industry Working Paper 2004/7.Organisation for Economic Co-operation and De-velopment, Paris. http://www.olis.oecd.org/olis/2004doc.nsf/43bb6130e5e86e5fc12569fa005d004c/c1912c8b0517eba3c1256ea1003a038d/$FILE/JT00164994.PDF.

Kinoshita, Yuko. 2000. “R&D and TechnologySpillovers via FDI: Innovation and AbsorptiveCapacity.” William Davidson Institute WorkingPaper No. 349. Available at SSRN: http://ssrn.com/abstract=258194.

Kohr, Martin. 2007. “Government Procurement inFTAs: An Outline of the Issues.” Third WorldNetwork. http://www.twnside.org.sg. AccessedNovember 15, 2007.

Kugler, Maurice. 2006. “Spillovers from Foreign DirectInvestment: Within or between Industries?” Jour-nal of Development Economics 80 (2): 444–77.

Kui-Yin, Cheung, and Ping Lin. 2007. “Spillover Ef-fects of FDI on Innovation in China: Evidencefrom the Provincial Data.” http://ssrn.com/abstract=419020. Accessed November 15, 2007.

Kuznetsov, Yevgeny. 2006. “Leveraging Diasporas ofTalent: Towards a New Policy Agenda.” In Dias-pora Networks and the International Migrationof Skills, ed. Yevgeny Kuznetsov, 221–37. Wash-ington, DC: World Bank Institute.

______. 2007. “The New Argonauts: Global Searchand Local Institution-Building.” Unpublishedpaper. World Bank Institute, Washington, DC.

Kuznetsov, Yevgeny, Adolfo Nemirovsky, and GabrielYoguel. 2006. “Argentina: Burgeoning Networksof Talent Abroad, Weak Institutions at Home.” InDiaspora Networks and the International Migra-tion of Skills, ed. Yevgeny Kuznetsov, 153–70.Washington, DC: World Bank Institute.

Lai, Edwin L. C. 1998. “International Property RightsProtection and Rate of Product Innovation.”Journal of Development Economics (February):133–53.


160

Lall, Sanjaya. 2000. “The Technological Structure andPerformance of Developing Country Manufactur-ing Exports, 1985–98.” Oxford DevelopmentStudies 28 (Part 3): 337–69.

______. 2001. “Competitiveness Indices and Develop-ing Countries: An Economic Evaluation of theGlobal Competitiveness Report.” World Devel-opment 29 (9): 1501–25.

______. 2003. “Indicators of the Relative Importanceof IPRs in Developing Countries.” Issue Paper 3.UNCTAD-ICTSD Project on IPRs and SustainableDevelopment.

Lederman, Daniel, and William F. Maloney. 2006.“Innovation in Mexico: NAFTA Is Not Enough.”In Global Integration and Technology Transfer, ed.Bernard Hoekman and Beata Smarzynska Javorcik,305–339. New York: Palgrave MacMillan.

Lee, Jeong-Yeon, and Edwin Mansfield. 1996. “Intel-lectual Property Protection and U.S. Foreign In-vestment.” Review of Economics and Statistics.78 (2): 181–86.

Lemoine, Françoise, and Deniz Ünal-Kesenci. 2003.“Trade and Technology Transfers: A Compara-tive Study of Turkey, India, and China.” WorkingPaper 2003-16. Centre d’Etudes Prospectives etd’Informations Internationales.

Levinsohn, James. 1993. “Testing the Imports-as-Market-Discipline Hypothesis.” Journal of Inter-national Economics 35 (1): 1–22.

Lipsey, Robert. 2002. “Home and Host Country Effectsof FDI.” Working Paper 9293. National Bureau ofEconomic Research, Cambridge, MA.

Lipsey, Robert E., and Fredrik Sjoholm. 2005. “TheImpact of Inward FDI on Host Countries: WhySuch Different Answers?” In Does Foreign In-vestment Promote Development?, ed. TheodoreH. Moran, Edward M. Graham, and MagnusBlomstrom. Washington, DC: Institute for Inter-national Economics.

Liu, Lili, and James Tybout. 1996. “ProductivityGrowth in Colombia and Chile: Panel-Based Evi-dence on the Role of Entry, Exit and Learning.”In Industrial Evolution in Developing Countries,ed. Mark J. Roberts and James R. Tybout. NewYork: Oxford University Press.

Lumenga-Neso, A., M. Olarreaga, and M. Schiff.2005. “On ‘Indirect’ Trade-Related R&DSpillovers and Growth.” European EconomicReview 49 (7): 1785–98.

Maloney, William. 2006. “Missed Opportunities: In-novation and Resource-Based Growth in LatinAmerica.” World Bank, Washington, DC.

Mansfield, Edwin. 1994. “Intellectual Property Protec-tion, Foreign Direct Investment, and Technology

Transfer.” International Finance Discussion Paper19. Washington, DC, World Bank.

______. 1995. “Intellectual Property Protection, DirectInvestment, and Technology Transfer: Germany,Japan, and the United States.” Discussion Paper.World Bank and International Finance Corpora-tion, Washington, DC.

Mansfield, E., and A. Romeo. 1980. “TechnologyTransfer to Overseas Subsidiaries by U.S.-BasedFirms.” Quarterly Journal of Economics 95 (4):737–50.

Maskus, Keith. 1998. “The International Regulationof Intellectual Property.” WeltwirtschaftlichesArchiv 134 (2): 186–208.

______. 2000. “Intellectual Property Rights andForeign Direct Investment.: Policy DiscussionPaper 22. University of Adelaide, Centre forInternational Economic Studies, Adelaide,Australia.

______. 2002. “Intellectual Property Issues for theUnited States and Japan: Disputes and CommonInterests.” In Issues and Options for U.S.–JapanTrade Policies, ed. R.M. Stern, 165–90. AnnArbor, MI: University of Michigan Press.

Maskus, Keith, and D. Konan. 1994. “Trade-RelatedIntellectual Property Rights: Issues and Ex-ploratory Results. In Analytical and NegotiatingIssues in the Global Trading System, ed. A. Dear-dorff and R. Stern, 401–46. Ann Arbor, MI: Uni-versity of Michigan Press.

Mattoo, Aaditya, Caglar Özden, and Cristina Neagu.2005. “Brain Waste? Educated Immigrants in theU.S. Labor Market.” Policy Research WorkingPaper 3581, World Bank, Washington, DC.

Mattoo, Aaditya, and Lucy Payton, eds. 2007. ServicesTrade and Development: The Experience of Zam-bia. Washington, DC: World Bank.

Mayer, School Yafeh. 2003. Sources of Funds and In-vestments Activities of Venture Capital Funds.Oxford University Press.

McCormick, Barry, and Jackline Wahba. 2003. “Re-turn Migration and Geographical Inequality: TheCase of Egypt.” Journal of African Economies 12(4): 500–32.

McKinsey Global Institute. 2003. New Horizons:Multinational Investment in Emerging Economies.San Francisco: McKinsey & Company, Inc.http://www.mckinsey.com/mgi/publications/newhorizons/.

Moran, Theodore. 2007. “How to Investigate the Im-pact of Foreign Direct Investment on Develop-ment and Use the Results to Guide Policy.” Un-published report. Georgetown University,Washington, DC.


161

Nastas, Tomas. 2007. “Scaling up Innovation: The Go-Forward Plan to Prosperity.” Development Out-reach, World Bank Institute. Washington, DC.

Ndou, Valentina. 2004. E-Government for DevelopingCountries: Opportunities and Challenges. Alba-nia: University of Shkoder, Department of Busi-ness Administration.

Nishimizu, Mieko, and John M. Page Jr. 1982. “TotalFactor Productivity Growth, TechnologicalProgress and Technical Efficiency Change:Dimensions of Productivity Change inYugoslavia, 1965–78.” Economic Journal 92(368): 920–36.

OECD (Organisation for Economic Co-operation andDevelopment). 2003. Tax Incentives for Researchand Development: Trends and Issues. Paris: OECD.

———. 2005. “Financing ICTs for Development Ef-forts of DAC Members.” Development Assis-tance Committee, OECD.

______. 2007. Reviews of Innovation Policy. China,Synthesis Report. Paris: OECD.

O’Neil, Kevin. 2003. Brain Drain and Gain: The Caseof Taiwan. Migration Information Source.Washington, DC: Migration Policy Institute.http://www.migrationinformation.org/Feature/display.cfm?ID=155).

Pack, Howard. 2006. “Econometric Versus Case StudyApproaches to Technology Transfer.” In GlobalIntegration and Technology Transfer, ed. BernardHoekman and Beata Smarzynska Javorcik,29–50. New York: Palgrave MacMillan.

Pack, Howard, and Kamal Saggi. 1997. “Inflows ofForeign Technology and Indigenous Technologi-cal Development.” Review of Development Eco-nomics 1 (1): 81–98.

Page, John, and Sonia Plaza. 2006. “Migration, Re-mittances and Development: A Review of GlobalEvidence.” Journal of African Economics 15(Supplement 2): 245–336.

Pandey, Abhishek, Alok Aggarwal, Richard Devane,and Yevgeny Kuznetsov. 2006. “The IndianDiaspora: A Unique Case?” In DiasporaNetworks and the International Migration ofSkills, ed. Yevgeny Kuznetsov, 71–97. Washing-ton, DC: World Bank Institute.

Parente, Stephen L., and Edward C. Prescott. 1994.“Barriers to Technology Adoption and Develop-ment.” Journal of Political Economy 102 (2):298–321.

Pfeil, Andreas. 2000. Venture Capital: New Ways ofFinancing Technology Innovation. Human Devel-opment Report Office.

Primo Braga, Carlos A., John Daly, and Bimal Sareen.2003. “The Future of Information and Commu-nication Technologies for Development.” Paper

presented at the Information and Communica-tions Technology Development Forum, Peters-burg, Germany, May 21.

Primo Braga, Carlos A., and Carsten Fink. 2000.“International Transactions in IntellectualProperty and Developing Countries.” Interna-tional Journal of Technology Management 19:35–56.

Puri, Shivani, and Tineke Ritzema. 1999. “MigrantWorker Remittances, Micro-Finance, and the In-formal Economy: Prospects and Issues.” WorkingPaper 21. International Labour Organisation,Social Finance Unit, Geneva.

Rauch, James, and Vitor Trindade. 2002. “Ethnic Chi-nese Networks in International Trade.” Review ofEconomics and Statistics 84 (1): 116–130.

Reserve Bank of India. 2007. “Balance of PaymentData.” Reserve Bank of India. http://www.rbi.org.in/scripts/Statistics.aspx. Accessed Sep-tember 1, 2007.

Rhee, Yung Whee, Katharina Katterback, and JannetteWhite. 1990. Free Trade Zones in Export Strate-gies. Washington, DC: World Bank.

Rhee, Yung Whee, Bruce Ross-Larson, and GaryPursell. 1984. Korea’s Competitive Edge: Manag-ing the Entry into World Markets. Washington,DC: World Bank.

Robbins, C. 2006. “Measuring Payments for theSupply and Use of Intellectual Property.” U.S.Department of Commerce, Bureau of EconomicAnalysis, Washington, DC.

Roberts, Mark. 1996. “Colombia 1977–1985: Pro-ducer Turnover, Margins, and Trade Exposure.”In Industrial Evolution in Developing Countries:Micro Patterns of Turnover, Productivity, andMarket Structure. New York: Oxford UniversityPress.

Roberts, Mark J., and James R. Tybout. 1997.“Producer Turnover and Productivity Growth inDeveloping Countries.” World Bank ResearchObserver 12 (1): 1–18.

Rodriguez-Clare, Andres. 2005. Coordination Fail-ures, Clusters and Microeconomic Interventions.Washington, DC: Inter-American DevelopmentBank.

Rodrik, Dani. 2004. Industrial Policy for the Twenty-First Century. Cambridge, MA: HarvardUniversity.

Rosenberg, Nathan. 1976. Perspectives on Technology.Cambridge, U.K.: Cambridge University Press.

Rosenzweig, Mark R., and Kenneth I. Wolpin. 1994.“Are There Increasing Returns to the Intergener-ational Production of Human Capital? MaternalSchooling and Child Intellectual Achievement.”Journal of Human Resources 29 (2): 670–93.


162

Saggi, Kamal. 2002. “Trade, Foreign Direct Investmentand International Technology Transfer: A Sur-vey.” World Bank Research Observer 17 (2):191–235.

Saxenian, AnnaLee. 2000. “Silicon Valley’s NewImmigrant Entrepreneurs.” Working Paper 15.University of California, Center for ComparativeImmigration Studies, San Diego.

______. 2002. “Brain Circulation: How High-SkilledImmigration Makes Everyone Better Off.”Brookings Review 20 (1): 28–31.

Schiff, Maurice, and Yanling Wang. 2006a.“North–South and South–South Trade-RelatedTechnology Diffusion: An Industry-Level Analy-sis of Direct and Indirect Effects.” CanadianJournal of Economics 39 (3): 831–44.

______. 2006b. “On the Quantity and Quality ofKnowledge Diffusion: The Impact of Opennessand Foreign R&D on North–North and North–South R&D Spillovers.” In Global Inte-gration and Technology Transfer, ed. BernardHoekman and Beata S. Javorcik. New York: Pal-grave McMillan.

Silberglitt, Richard, Philip S. Anton, David R. Howell,and Anny Wong. 2006. The Global TechnologyRevolution 2020. Santa Monica, CA: RANDCorporation, National Security ResearchDivision.

Sjoholm, F. 1998. “Joint Ventures, Technology Trans-fer, and Spillovers: Evidence from Indonesian Es-tablishment Data.” In Foreign Direct Investmentin Asia, ed. Economic Planning Agency, 587–616.Tokyo: Economic Planning Agency.

Smith, Pamela J. 2001. “How Do Foreign PatentRights Affect U.S. Exports, Affiliate Sales, and Li-censes?” Journal of International Economics 55(2): 411–40.

Soderbom, Mans, and Francis Teal. 2000. “Skills, In-vestment and Exports from Manufacturing Firmsin Africa.” Journal of Development Studies 37(2): 13–43.

Sorenson, Olav, and Toby E. Stuart. 2001. “SyndicatedNetworks and the Spatial Distribution of VentureCapital Investments.” American Journal of Soci-ology 106 (6): 1546–88.

Soubbotina, Tatyana P. 2006. “Generic Models ofTechnological Learning by Developing Countries.”Draft. World Bank, Science and Technology Pro-gram, Washington, DC.

Sullivan, Nicholas P. 2007. You Can Hear Me Now:How Microloans and Cell Phones are Connectingthe World’s Poor to the Global Economy. Jossey-Bass.

Swinnen, J. F. M., L. Dries, N. Noev, and E. Germenji.2006. “Foreign Investment, Supermarkets, and

the Restructuring of Supply Chains: Evidencefrom Eastern European Diary Sectors.” WorkingPaper Series WP 165/2006, 2006. LICOS Centrefor Transition Economics.

Taylor, M. S. 1994. “TRIPS, Trade and Growth.” In-ternational Economic Review 35: 361–81.

Tybout, James R., Jaime de Melo, and Vittorio Corbo.1991. “The Effects of Trade Technical Efficiency:New Evidence from Chile.” Journal of Interna-tional Economics 31 (3–4): 231–50.

Tybout, James R., and M. Daniel Westbrook. 1995.“Trade Liberalization and the Dimensions ofEfficiency Change in Mexican Manufacturing In-dustries.” Journal of International Economics 39(1–2): 53–78.

———. 1996. “Trade Liberalization and the Dimen-sions of Efficiency Change in Mexican Manufac-turing Industries.” Journal of International Eco-nomics 39 (1–2): 53–78.

UNCTAD (United Nations Conference on Trade andDevelopment). 2001. World Investment Report2001: Promoting Linkages. New York andGeneva: United Nations.

______. 2003. Transfer of Technology for SuccessfulIntegration into the Global Economy. UNC-TAD/ITE/IPC/2003/6. Geneva: UNCTAD.

______. 2005. World Investment Report 2005:Transnational Corporation and Internationaliza-tion of R&D. New York and Geneva: UnitedNations.

______. 2007. The Least Developed Countries 2007:Knowledge, Technological Learning and Innova-tion for Development. Geneva: UNCTAD.

UNDP (United Nations Development Programme).2003. Millennium Development Goals: A Com-pact among Nations to End Human Poverty.New York: Oxford University Press.

United Nations. 2003. United Nations World PublicSector Report 2003: E-Government at the Cross-roads. New York: United Nations.

van Wijnbergen, Sweder, and Anthony Venables. 1993.“Trade Liberalization, Productivity, and Compe-tition: The Mexican Experience.” DiscussionPaper 345. London School of Economics.

Wadhwa, Vivek, AnnaLee Saxenian, Ben Rissing, andGary Gerefi. 2007. America’s New ImmigrantEntrepreneurs. Report prepared for the Universityof California–Berkeley, Master of EngineeringManagement Program.

Wagner, Don, Keith Head, and John Ries. 2002. “Im-migration and the Trade of Provinces.” ScottishJournal of Political Economy 49 (5): 507–25.

Wahba, Jackline. 2007. “Returns to Overseas WorkExperience: The Case of Egypt.” In InternationalMigration, Economic Development, and Policy,


163

ed. Çaglar Özden and Maurice Schiff. Washing-ton, DC: World Bank.

Watkins, Alfred, and Anubha Verma. 2007. “Govern-ment of Rwanda-World Bank, Science, Technol-ogy and Innovation Capacity Building AssistanceProgram: Developing Practical Solutions to Prac-tical Problems.” Internal report. World Bank,Washington, DC.

Wei, Yingqi. 2004. “Foreign Direct Investment inChina.” In Foreign Direct Investment: Six Coun-try Case Studies, ed. Yingqi Wei and V. N.Balasubramanyam, 9–37. Cheltenham, U.K.:Edward-Elgar.

Woodruff, Christopher, and Rene Zenteno. 2001. “Re-mittances and Microenterprises in Mexico.”Working Paper. University of California, SanDiego, Graduate School of International Rela-tions and Pacific Studies.

______. 2007. “Migration Networks and Microenter-prises in Mexico.” Journal of Development Eco-nomics 82 (2): 509–28. March.

World Bank. 1994. “Sugar Ethanol Policy.” In Brazil:The Management of Agriculture, Rural Develop-ment, and Natural Resources. Vol. 2. ReportNo. 11783–BR. July 31.

———. 1998. World Development Report: Knowledgefor Development. New York: Oxford UniversityPress.

______. 2004a. Global Development Finance 2004.Washington, DC: World Bank.

______. 2004b. India: Investment Climate and Manu-facturing Industry. Washington, DC: World Bank.

______. 2005. Brazil Investment Climate Assessment.Washington, DC: World Bank.

______. 2006a. Global Economic Prospects 2006:Economic Implications of Remittances and Mi-gration. Washington, DC: World Bank.

———. 2006b. World Development Report. Washing-ton, DC: World Bank.

______. 2007a. Global Development Finance: TheGlobalization of Corporate Finance in Develop-ing Countries. Washington, DC: World Bank.

______. 2007b. Our Commitment: The World Bank’sAfrica Region HIV/AIDS Agenda for Action2007–2011. Report No. 40513-AFR. Washing-ton, DC: World Bank.

______. 2007c. World Development Indicators. Wash-ington, DC: World Bank.

———. 2007d. Global Monitoring Report. 2007.Washington, DC: World Bank.

______. 2008. World Development Report. Washing-ton, DC: World Bank.

______. Forthcoming. “Export Diversification inEgypt, Jordan, Morocco, and Tunisia.” WorldBank, Washington, DC.

World Bank Information for Development Programand Center for Democracy and Technology. 2002.The E-Government Handbook for DevelopingCountries. Washington, DC: World Bank Infor-mation for Development Program and Center forDemocracy and Technology.

Xavier, Marcos Renato. 2007. “The Brazilian Sugar-cane Ethanol Experience.” Competitive Enter-prise Institute, February 15.

Xu, Bin. 2000. “Multinational Enterprises, TechnologyDiffusion, and Host Country ProductivityGrowth.” Journal of Development Economics62: 477–93.

Yang, Dean. 2006. “International Migration, Remit-tances, and Household Investment: Evidencefrom Philippine Migrants’ Exchange RateShocks.” Working Paper 12325. National Bureauof Economic Research, Cambridge, MA.

Yokota, Toshiyuki, and Richard J. Weiland. 2004. “ITSStandards for Developing Countries.” ITS Techni-cal Note 4. Washington, DC: World Bank.

Yoo, JaeHoon. 2007. “Financing Innovation: How toBuild an Efficient Exchange for Small Firms.”Public Policy Journal. No. 315. World BankWashington, DC.

Zeng, Douglas Zhihua, and Shilin Wang. 2007.“China and the Knowledge Economy: Challengesand Opportunities.” Staff Working Paper 4223.World Bank, Washington, DC.


164

165

AppendixRegional Economic Prospects

East Asia and the Pacific Recent developmentsGrowth in the developing countries of EastAsia and the Pacific strengthened in 2007,with gross domestic product (GDP) advancinga full 10 percent in the year, up from 9.7 per-cent in 2006. The expansion was powered byChina’s 11.3 percent gain, with other coun-tries in the region growing at a 5.9 percentpace (figure A1). Domestic demand was a keydriving force for many economies, as a down-turn in the global high-tech cycle for most of2006–07 served to blunt the momentum of

exports for technology-producing countries inthe region. The pickup in regional growth wasall the more notable because it occurreddespite a slowdown in the U.S. economy: totalU.S. imports fell from 5.8 percent growth in2006 to 2 percent in 2007. Regional exportsnonetheless advanced 17.8 percent, a modestpickup from 2006 outturns.

East Asia appears to have absorbed the ef-fects of the financial turmoil in the high-income markets well. Stock markets in themain East Asian economies dropped a median14 percent during July and early August, withequity prices increasing a median 22 percentto mid-October. Similar developments werewitnessed in foreign exchange markets and insovereign bonds. Since the beginning of theyear, several East Asian currencies have ap-preciated sharply against the dollar, with thePhilippine peso up 12 percent and the Thaibaht up 11.3 percent. The Chinese yuan con-tinued its gradual rise against the dollar, a4.7 percent gain since the beginning of theyear, but at the same time, the yuan depreci-ated against many other currencies.

On the policy front, East Asian centralbanks generally tightened monetary condi-tions from mid-2004 to the early part of 2006to curb rising inflation. As a result, inflationstabilized in 2007 (although headline inflationrates, which include fuels and food, haveturned up recently in some countries becauseof higher food price inflation), allowing cen-tral banks to keep policy rates stable, andeven to begin easing in Indonesia and Thailand

Figure A1 East Asian growth moves up in2007



2

East A

sia a

nd

the

Pacific

20052006

2007

4

6

10

12

8

China

East A

sia a

nd th

e Pac

ific

exclu

ding

China

South

Eas

t Asia

Vietna

m

Small

coun

tries

(figure A2). Fiscal balances have improved andgovernment debt has declined over the courseof the decade in most of the larger East Asianeconomies thanks to fiscal consolidation

efforts, relatively low interest rates, and sus-tained economic growth since 2001.

External conditions during 2007 remainedsufficiently positive for surplus positions towiden across countries. Central banks inChina, Indonesia, Malaysia, the Philippines,and Thailand continued to build up reserves astheir current accounts remained in surplus.East Asia’s aggregate current account sur-plus as a share of GDP increased to 10.1 per-cent in 2007, up from 8.4 percent in 2006.Gross capital flows, including bond and equityissuance and net bank borrowing, amountedto a remarkable $170 billion over the yearthrough October. This contrasts favorablywith inflows of $153 billion for all of 2006and $107 billion for 2005, indicating that mar-ket access has remained largely unencumbered.At the same time, the contribution of net ex-ports to GDP growth increased 3 percentagepoints, as exports expanded by 17.8 percent,significantly outpacing the 15.3 percentgrowth in imports (table A1).


166

Philippines

Figure A2 Except for China, inflation isnow stabilizing across East Asia

Consumer price inflation (percent change year over year)

0Jan.2006

May2006

Sept.2006

Jan.2007

Sept.2007

May2007

2

4

6

China8

0

9

6

3

12

15

18

Source: World Bank.

Indonesia (right axis)

ThailandMalaysia

Table A1 East Asia and Pacific forecast summary(annual percent change unless indicated otherwise)

Forecast

1991–2000a 2004 2005 2006 2007 2008 2009

GDP at market pricesb 8.4 9.0 9.1 9.7 10.0 9.7 9.6GDP per capitac 7.1 8.1 8.2 8.8 9.1 8.8 8.7Purchasing Power Parity GDPd — 9.2 9.3 9.9 10.2 9.9 9.7

Private consumption 7.3 6.8 7.5 7.4 7.6 7.6 7.6Public consumption 9.0 6.7 10.9 8.5 9.0 8.5 8.6Fixed investment 10.3 11.5 12.7 10.9 11.3 9.9 9.6Exports, GNFSe 11.7 22.6 17.8 17.7 17.8 15.2 18.5Imports, GNFSe 11.3 20.6 10.5 14.8 15.3 14.2 19.4

Net exports, contribution to growth 0.3 1.8 3.9 2.8 3.0 2.3 2.1Current account balance/GDP (%) 0.1 3.4 5.7 8.4 10.1 8.6 7.6GDP deflator (median, LCU) 6.5 6.1 3.8 4.3 4.6 2.9 3.8Fiscal balance/GDP (%) �0.7 �1.5 �1.4 �0.5 �0.9 �1.1 �1.2

Memo items: GDPEast Asia, excluding China 4.8 6.1 5.4 5.7 5.9 5.9 6.2China 10.4 10.1 10.4 11.1 11.3 10.8 10.5Indonesia 4.2 5.1 5.7 5.5 6.3 6.3 6.5Thailand 4.5 6.2 4.5 5.0 4.3 4.6 5.2

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. Measured in U.S. dollars.d. GDP measured at purchasing power parity exchange rates.e. Exports and imports of goods and nonfactor services.

In China, growth continued at a robustpace in 2007, underpinned by strong contri-butions to GDP from net exports and bybuoyant domestic demand, led by investment.Growth achieved a 11.5 percent run-up in thefirst half of the year, including an exceptional11.9 percent advance during the second quar-ter. The soaring current account surplus ofabout $380 billion in 2007, some 12 percentof GDP, is adding to domestic liquidity andcontributing to asset price increases, whilesupporting, along with other factors such assharp gains in food prices, an upward drift inconsumer price inflation.

Elsewhere investment growth picked upin most economies, as capacity utilizationreached high levels, corporate profits rose,and the health of balance sheets improved. InIndonesia, fixed investment surged 11.3 per-cent (seasonally adjusted annual rate) in thesecond quarter, and GDP growth increased to6.3 percent, up from 5.5 percent in 2006.Against this background, inflation is now agrowing concern, reaching 6.9 percent inSeptember (year-on-year), at the upper end ofthe central bank’s target range. GDP growthshould register a solid 6.3 percent for the year.

Malaysia suffered subpar export perfor-mance during the first half of 2007, tied inpart to sluggish demand for semiconductorsand other high-tech inputs, slow hydrocarbonshipments, and difficulties in several ex-portable food and raw material commodities.Export growth declined to 2.5 percent in thefirst half of 2007, down from 6 percent in thesecond half of 2006, contributing to a slow-down in GDP growth to 5.6 percent in thefirst half of the year. Equity markets were de-pressed for a short time during the period ofglobal financial stress, but have bounced backsharply since mid-August. Domestic demandis expected to sustain growth, offsetting weak-ness in trade and allowing Malaysia to register5.7 percent GDP growth for 2007.

In the Philippines, GDP growth ramped upto 7.3 percent during the first half of 2007based on strong investment outlays and apickup in services; growth for the year is

expected to register 6.7 percent. The countryis beginning to enjoy the benefits stemmingfrom the substantial fiscal adjustment, publicdebt reductions, and balance-of-payments surpluses of recent years. The current accountsurplus rose sharply as a result of large remit-tance inflows and a diminishing trade deficit.Foreign direct investment (FDI) inflows in-creased 70 percent from 2006 to $1.6 billionin the first half of the year, while internationalreserves have increased to some $30.7 billion,enough to cover 5.5 months of imports.

In Thailand, where continuing political andpolicy uncertainties have significantly damp-ened business and consumer confidence, GDPis anticipated to grow by 4.3 percent in 2007,down from 5 percent in 2006. Growth hasrelied on conditions in the external environ-ment, where the news is somewhat discourag-ing, given an 11 percent appreciation of thebaht against the dollar over 2007 to date andsluggish conditions in the U.S. market.

Growth has also continued to run at strong7 to 10 percent rates in several low-incomeeconomies of the region, including Cambodia,the Lao People’s Democratic Republic,Mongolia, and Vietnam, powered by across-the-board strength in exports and domesticdemand. Growth is also above historical ratesin some of the small island states of the region,pushed up by high commodity prices, and insome cases by improved economic manage-ment. At the same time, political instabilityand social tensions continue to undermineperformance in some of the Pacific islands,including Fiji, where GDP is expected tocontract this year.

Medium-term outlookGrowth in East Asia and the Pacific is pro-jected to remain strong, with GDP easing byjust 0.3 percentage points to 9.7 percent in2008 and retaining strength in 2009 with anadvance of 9.6 percent. Growth in China isexpected to slow modestly, dropping less thana percentage point over the period to 10.5 per-cent by 2009, as authorities’ long-standing at-tempts to rein in certain investment projects

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

167

and to avoid overheating in several sectorscome to fruition (table A2 and figure A3).Substantial reform efforts in several countriesof the Association of Southeast Asian Nationsshould yield acceleration in activity throughthe forecast period. Growth among East Asianeconomies other than China is expected toregister 6.2 percent by 2009.

A projected slowdown in export growthfrom 17.8 percent in 2007 to 15.2 percent in2008 echoes a softening in demand by coun-tries of the Organisation for Economic Co-operation and Development (OECD)—with U.S. imports increasing by a meager1.3 percent—as well as second-round effectson intraregional trade. However, the decline


168

Table A2 East Asia and Pacific country forecasts(annual percent change unless indicated otherwise)

Forecast

1991–2000a 2004 2005 2006 2007 2008 2009

CambodiaGDP at market pricesb — 10.0 13.5 10.8 9.5 8.0 9.0Current account balance/GDP (%) — �3.7 �7.1 �5.8 �7.8 �11.2 �9.0

ChinaGDP at market pricesb 10.4 10.1 10.4 11.1 11.3 10.8 10.5Current account balance/GDP (%) 1.5 3.6 6.9 9.4 11.9 10.3 9.1

FijiGDP at market pricesb 2.1 5.3 0.7 3.6 �3.1 1.9 2.8Current account balance/GDP (%) �3.1 �16.8 �22.7 �22.2 �20.1 �24.0 �26.2

IndonesiaGDP at market pricesb 4.2 5.1 5.7 5.5 6.3 6.3 6.5Current account balance/GDP (%) �0.4 0.6 0.4 3.1 2.7 1.5 0.9

Lao People’s Democratic RepublicGDP at market pricesb — 6.4 7.1 7.6 7.1 7.9 7.5Current account balance/GDP (%) — �6.3 �26.4 �18.7 �16.8 �19.0 �19.0

MalaysiaGDP at market pricesb 7.1 7.2 5.0 5.9 5.7 5.9 6.0Current account balance/GDP (%) �0.4 12.6 15.3 17.1 13.8 12.2 10.2

Papua New GuineaGDP at market pricesb 4.8 2.7 3.4 2.6 5.2 4.0 4.2Current account balance/GDP (%) 2.2 �1.5 2.6 4.4 2.7 3.6 2.8

PhilippinesGDP at market pricesb 3.0 6.2 4.9 5.4 6.7 6.2 6.5Current account balance/GDP (%) �3.1 1.9 2.0 5.3 4.3 2.4 1.8

ThailandGDP at market pricesb 4.5 6.2 4.5 5.0 4.3 4.6 5.2Current account balance/GDP (%) �1.2 1.7 �4.6 1.6 2.4 1.4 1.5

VanuatuGDP at market pricesb 4.1 4.0 6.5 7.2 5.0 5.0 5.0Current account balance/GDP (%) �8.2 �19.7 �20.2 �22.1 �19.9 �20.8 �20.4

VietnamGDP at market pricesb 7.6 7.7 8.4 8.2 8.3 8.2 8.3Current account balance/GDP (%) �5.1 �1.0 �0.3 1.1 �1.1 �0.6 �1.8

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. American Samoa, Dem. Rep., the Federated States of Micronesia, Kiribati, Korea, Northern Mari-ana Islands, Marshall Islands, Mongolia, Myanmar, Palau, Solomon Islands, Timor-Leste, and Tonga are not forecast because ofdata limitations.— � not availablea. 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.

should not make a serious dent in regionalGDP, and the contribution of net trade togrowth is projected to fall only moderately,from 3 percentage points in 2007 to 2.3 in2008. Largely reflecting developments inChina, the momentum underlying fixed in-vestment begins to dissipate during 2008, eas-ing from 11.3 percent in 2007 to 9.9 percent.By 2009, the external environment is expectedto feature a revival in U.S. GDP growth com-plemented by recovery in Europe and Japan.OECD import demand is forecast to increasefrom 5 percent in 2008 to 7.8 percent in 2009,and conditions in financial markets are ex-pected to stabilize.

RisksThe year 2008 will likely be challenging forpolicy makers, with a large number of interre-lated downside risks. These include the possi-bility of a full-fledged recession in the UnitedStates, higher oil prices, and further escalationof turbulence in financial markets linked tothe U.S. subprime debacle.

Among principal concerns is the extent towhich both financial and real side effects ofthe U.S. subprime crisis might increase in the

coming year. China should be well positionedto weather the continuing turmoil in financialmarkets. The impact on Chinese financial in-stitutions holding overseas collateralized debtobligations and other U.S. mortgage-backedsecurities appears likely to be small in relationto the size of China’s economy and its huge in-ternational reserves ($1.4 trillion), but othercountries may be more vulnerable to effectsflowing through both direct and indirectchannels.

Should losses by large international insti-tutions mount to substantial levels, other in-vestments, including those in East Asia, couldbe called in an effort to rebalance portfoliosand mitigate the effects on trading profits.Several countries in East Asia are exposed tothis risk, particularly those that have been re-cipients of large capital inflows intermediatedthrough the yen carry trade. Policy makerswill need to keep a close eye on the volume,direction, and volatility of short-term flows,including those into local equity markets.Nevertheless the large holdings of foreign ex-change reserves and the current account sur-plus positions of most East Asian economiesshould provide a significant buffer and reducemacroeconomic vulnerability to a reversal incapital flows.

The obverse of this risk is that interest ratereductions in high-income economies mayboost liquidity to the point of touching off an-other upward cycle in equities, including inemerging markets, setting the stage for aneven more pronounced adjustment later. Evenin the absence of such a scenario, manyeconomies in the region have been strugglingto curb liquidity growth caused by burgeoningcurrent account surpluses and large-scalebuildup of reserves. If not managed properly,excess liquidity could jeopardize price stabil-ity, form asset price bubbles, and expose acountry to serious financial and macroeco-nomic vulnerabilities. Finally, a slowdown inthe high-income countries that is more severethan projected would subject many East Asianeconomies to a substantial downdraft in ex-port growth.


169

Figure A3 Performance improves forEast Asian countries other than China

42005 2006 2007 2008 2009

12

10

8

6

GDP growth (percent change)

Source: World Bank.

ChinaEast Asia

Excluding China

Europe and Central AsiaRecent developmentsGDP growth in the Europe and Central Asiaregion eased slightly, from 6.9 percent in 2006to 6.7 percent in 2007, reflecting a modest soft-ening of both external and domestic demand(table A3). With a stable population in the re-gion, this means that per capita productioncontinued to increase at remarkable rates ofmore than 6 percent. High productivity gainshave been made possible by technologydiffusion, double-digit growth in investmentsupported by rapid credit expansion throughlending by domestic and foreign banks, highenergy prices for hydrocarbon exporters, andlarge remittance inflows from workers over-seas. These same factors have boosted privateconsumption and consistently raised importgrowth 3 or more percentage points above thealready robust expansion of exports. These de-velopments, which have helped to rapidly

increase standards of living, are not withoutshadow costs. Capital inflows have createdchallenges for macroeconomic management;inflation remains high relative to that in theEuro Area, making it more difficult for severalcountries to maintain effective exchange ratepegs; and current account deficits in many oil-importing countries have become unsustain-ably high.

At the subregional level, growth in Centraland Eastern Europe (CEE) moderated to a stillrobust 6.0 percent in 2007 from 6.5 percent in2006 (figure A4), buoyed by rapid growthin credit and in real wages, strong capitalinflows, and high remittance inflows. Thefalloff in growth in CEE is attributable in largemeasure to a slowdown in Hungary, wherea program of fiscal consolidation pushedgrowth down 3.2 percentage points to 2.2 per-cent in 2007. The decline in growth in CEEalso stems from continued moderation in


170

Table A3 Europe and Central Asia forecast summary(annual percent change unless indicated otherwise)

Forecast

1991–2000a 2004 2005 2006 2007 2008 2009

GDP at market pricesb �1.0 7.4 6.1 6.9 6.7 6.1 5.7GDP per capitac �1.2 7.4 6.2 6.9 6.7 6.0 5.7Purchasing Power Parity GDPd �0.9 7.6 6.1 7.1 7.0 6.2 5.8

Private consumption 0.5 8.6 7.8 7.4 7.3 6.9 6.8Public consumption 0.1 2.3 3.4 5.0 4.5 5.4 3.4Fixed investment �6.6 14.1 11.5 16.5 14.9 13.0 10.3Exports, GNFSe 0.9 12.4 7.0 10.3 9.2 8.5 8.7Imports, GNFSe �1.6 17.4 10.2 14.0 12.8 12.2 11.2

Net exports, contribution to growth 0.9 �1.7 �1.4 �1.8 �2.0 �2.3 �1.9Current account balance/GDP (%) — 0.8 1.5 0.6 �1.3 �1.9 �2.6GDP deflator (median, LCU) 118.5 6.6 5.8 7.3 6.9 6.4 5.5Fiscal balance/GDP (%) �6.1 �0.7 2.0 2.9 1.6 1.6 1.5

Memo items: GDPTransition countries 2.0 6.9 5.7 6.3 5.7 5.5 5.3Central and Eastern Europe 1.2 5.7 4.7 6.5 6.0 5.5 5.2Commonwealth of Independent States �4.2 8.0 6.8 7.8 8.2 6.8 6.2

Russian Federation �3.9 7.1 6.4 6.7 7.5 6.5 6.0Turkey 3.6 8.9 7.4 6.1 5.1 5.4 5.7Poland 3.8 5.3 3.6 6.1 6.5 5.7 5.1

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. Measured in U.S. dollars. d. GDP measured at purchasing power parity exchange rates.e. Exports and imports of goods and nonfactor services.

Turkish GDP, which slowed by 1 percentagepoint to 5.1 percent in 2007 from an unsus-tainable 8.9 percent pace posted in 2004.

In contrast with CEE, growth in the Com-monwealth of Independent States (CIS) accel-erated sharply from 7.8 percent in 2006 to8.2 percent in 2007 (figure A4). A strong in-crease in public consumption and investment,combined with a slightly improved contribu-tion from net exports, formed the foundationfor this pickup in growth. Revenue gains forthe oil-exporting economies, notably Azerbai-jan, Kazakhstan, and the Russian Federation,continue at robust rates, given increasing oilprices, and are providing ongoing support todemand growth through fiscal linkage. A con-struction boom in residential, commercial,and civil engineering (infrastructure) projectsis contributing to a rise in the non-oil sectors.Among the smaller CIS economies, highworker remittance inflows, FDI, and vibrantdemand from regional oil exporters (notablyRussia) and from Asia (especially China) areunderpinning growth. Gross worker remit-tances represented a substantial share of GDPfor several of the region’s countries in 2006,

ranging from the equivalent of 4 percent ofGDP to as much as 38 percent: Albania (15 percent), Armenia (19 percent), Azerbaijan(4 percent), Georgia (7 percent), the KyrgyzRepublic (12 percent), Moldova (38 percent),and Tajikistan (20 percent). Remittances areanticipated to maintain this strong level in2007.

Europe and Central Asia’s regional currentaccount position shifted to a modest deficitequivalent to 1.3 percent of GDP in 2007 afterposting a surplus of 0.6 percent in 2006 (fig-ure A5) and an average surplus of nearly 1percent of GDP over 2000–05. Strong domes-tic demand is driving import volume growthof 12.8 percent, well in excess of exports at9.2 percent. Sizable current account deficitsamong countries in CEE have been financed toa large extent by FDI, although for the Balticstates, foreign borrowing by banks has cometo represent a substantial share of external fi-nance, leading to higher external debt-to-GDPratios. In the case of Latvia, that ratio reached112 percent in 2006 and is projected to re-main above 100 percent during 2007–09.Moreover, short-term debt as a share of total


171

Figure A4 Mixed growth outturns acrossEurope and Central Asia


9.0

3.0

4.5

6.0

7.5

GDP growth (% year-on-year)

Centra

l and

Easte

rn E

urop

e

Europ

e an

d

Centra

l Asia

Comm

onwea

lth o

f

Inde

pend

ent S

tate

sPola

nd

Turk

ey

Russia

20052006

2007

Figure A5 External positions vary widelyacross Europe and Central Asia

�25

0

�5

5

15

20

Uzbek

istan

Russia

n Fed

erat

ion

Ukrain

e

Poland

Hunga

ry

Roman

ia

Latvi

a

Current account balance as a percent of GDP (%)

10

�20

�15

�10


Centra

l and

Easte

rn E

urop

e

Europ

e an

d

Centra

l Asia

Comm

onwea

lth o

f

Inde

pend

ent S

tate

s

20052006

2007

external debt is high in a number of countries,reaching more than 40 percent in Latvia,Lithuania, and the Slovak Republic. These in-dicators point to potential future problemswith currency and maturity mismatches.

In the CIS, worker remittances havehelped finance significant external deficits ina number of smaller countries. In Kazakhstanand the Kyrgyz Republic, external debt as ashare of gross national income stood at 83and 86 percent, respectively, in 2005, and inthe case of Kazakhstan, this share has in-creased by 12 percentage points since 2001.Indeed, the rise in the international indebted-ness of Kazakh banks has recently focusedattention on the country, given the turbulencein international financial markets. In the KyrgyzRepublic, debt burdens remain at high levels,but they have been reduced sharply since2000—by more than 50 percentage points asa share of gross national income—largely be-cause of debt rescheduling by the Paris Clubin 2002 and an improvement in debt man-agement strategy.

Fiscal positions in the region generallydeteriorated in 2007, with the largest shiftsposted in the CIS. The most notable declinehas been in Tajikistan, where the fiscal bal-ance shifted from a surplus of 1.6 percent ofGDP in 2006 to a deficit of 10.3 percent. Thisreflects, in part, an effort to offset weakeningexports to sustain domestic consumption andinvestment. Marked deteriorations in fiscalpositions of 2 percentage points or moreduring 2007 have been recorded in Belarus(2.0 points), Bosnia and Herzegovina(3.4 points), Kazakhstan (3.0 points), Russia(2.2 points), and Turkey (3.4 points). In con-trast, notable consolidation has been achievedin Hungary, where the austerity program re-duced the deficit from 9.2 percent of GDP in2006 to 6.4 percent. Firming government rev-enues underpinned by stronger than expectedGDP growth helped manage a reduction ofPoland’s deficit from 3.9 percent of GDP in2006 to 3.0 percent in 2007. In Azerbaijan,increasing oil revenues and new productivecapacity have led to a considerable rise in the

fiscal surplus, which is equivalent to 5.0 per-cent of GDP, up from 0.1 percent in 2006.

Monetary policy across the region hasbecome more restrictive to counter rising infla-tionary pressures. Price increases are stemmingfrom sustained high domestic demand growthand rising fuel and grain prices, the latteraggravated locally by drought conditions inBulgaria and Romania and globally by thesurge in the use of cereals for biofuels.Moldova posted the largest escalation in prices.In Hungary and Latvia, prices were up 3 per-centage points; inflation in Hungary is beingdriven by increases in indirect taxes and ad-ministered prices and in Latvia by rapid creditexpansion tied to vibrant capital inflows. InAzerbaijan, inflation is expected to rise to16 percent in 2007, double the rate of 2006; inUkraine and Uzbekistan inflation is expected toaverage 17.5 and 17.0 percent, respectively.

In several countries, however, inflationarypressure has eased. In Romania, consumerprices fell from 6.6 percent during 2006to 4.6 percent in 2007, thanks to currencyappreciation and a delay in regulated priceadjustments. The Slovak Republic is also pro-jected to see an easing of inflation of some2 percentage points. Tighter domestic condi-tions and exchange rate appreciation helpedmoderate inflationary pressures in Croatia,Kazakhstan, and Turkey during the year.Nonetheless, inflation pressures are expectedto rekindle, in part reflecting higher foodprices and energy costs, which are affecting awide spectrum of countries.

The impact of market turbulence tied to theU.S. subprime mortgage market has been fairlylimited in the region, and initial downside ad-justments in currency and asset prices havelargely been recouped. Bond spreads increased,but not as much as in other markets. Never-theless, concerns about potential spillovers re-main for a number of countries in the region,particularly those that have experienced rapidcredit growth and private sector borrowingfrom abroad, the proportions of which maybe underestimated. Signs of overheating areclearly evident in Bulgaria and the Baltic states,


172

where already worrisome external positionshave deteriorated even further during 2007.Given that foreign inflows are financing muchof the credit expansion in these economies, in-creased market volatility points to heightenedconcerns in relation to currency mismatches,sudden stops, and contagion. A potential ex-change rate risk is present in a number of coun-tries where loans denominated in foreign cur-rencies make up a large share of total loans bydomestic banks. In the Baltic states, Hungary,Kazakhstan, Romania, and Ukraine, this ratiowas 40 percent or more in 2006, and in thecase of Latvia, the share increased by morethan 15 percentage points since 2001 to nearly80 percent in 2006.

Medium-term outlookFrom GDP gains of 6.7 percent in 2007,growth is projected to continue easing, fallingoff to 6.1 percent in 2008 and to 5.7 percentin 2009 (tables A3 and A4, figure A6). Theslowdown in 2008 is expected to be wide-spread across countries in the region, givenheightened risk aversion and volatility on in-ternational financial markets, which could

spill over to the region both directly and alsoindirectly through a faltering of externaldemand. Slower growth in the OECD coun-tries, especially in Germany and the EuroArea, may dampen export growth for CEEduring the year. Difficulties among Europeanfinancial institutions would also have reper-cussions throughout Europe and Central Asia.Domestic demand growth is anticipated tomoderate from recent highs, with the contri-bution to growth from both private consump-tion and investment projected to fall by0.2 percentage points during 2008. The con-tribution of trade to growth—reflecting weak-ened external demand, and despite a degreeof softening import growth—is expected tobecome still more negative in 2008.

Three notable exceptions to the projectedgrowth slowdown in 2008 are Albania, Hun-gary, and Turkey. In Albania, continued strongdomestic demand is expected to help firm upgrowth. A key component of that demand isincreased public investment to mitigate thepower shortages that have created a bottle-neck to growth. In Hungary and Turkey, im-provements in domestic conditions shouldpermit additional easing of monetary policy,bolstering demand sufficiently to bring abouta pickup in GDP growth.

By 2009, external demand is projected tostrengthen in concert with GDP growth in theOECD, leading to an improvement in contri-butions to growth from net exports, equiva-lent to 1.9 percentage points in 2009 (follow-ing a drop by 2.3 percentage points in 2008).A further falloff in domestic demand growth;particularly the investment in the CIS coun-tries, is projected to offset this improvementsomewhat, resulting in modest deceleration inregional growth to 5.7 percent in 2009. Inlarge measure, the projected slowdown in theCIS is driven by the near completion of majorhydrocarbon investment projects that led tothe expansion of production and exportcapacity in recent years.

Despite a rise in external demand and con-tinued moderation in domestic demand, theregional current account is anticipated to


173

Figure A6 Growth in Europe and CentralAsia eases into 2009

0

4

2

8

10Forecast

2000

2001

2002

2003

2004

2005

2006

2007

2008

2009

Real GDP (annual percent change)

6

�4 �7.5

�2

Source: World Bank.

Central and Eastern Europe

Commonwealth of Independent States

Turkey

Europe andCentral Asia


174

Table A4 Europe and Central Asia country forecasts(annual percent change unless indicated otherwise)

Forecast

1991–2000a 2004 2005 2006 2007 2008 2009

AlbaniaGDP at market pricesb 1.4 5.9 5.5 5.0 5.5 6.0 6.2Current account balance/GDP (%) �5.6 �4.8 �7.8 �7.2 �8.4 �8.4 �8.0

ArmeniaGDP at market pricesb �3.8 10.5 13.9 13.3 11.0 8.5 7.5Current account balance/GDP (%) �12.0 �4.5 �4.4 �4.7 �5.7 �5.6 �4.8

AzerbaijanGDP at market pricesb �5.2 10.2 26.4 34.5 33.5 19.4 14.9Current account balance/GDP (%) �15.8 �29.8 1.3 18.2 24.2 31.6 33.3

BelarusGDP at market pricesb �1.2 11.4 9.4 9.9 7.8 6.4 5.7Current account balance/GDP (%) — �5.2 1.8 �4.1 �8.0 �8.4 �8.3

BulgariaGDP at market pricesb �1.7 5.7 5.5 6.3 6.1 6.0 5.2Current account balance/GDP (%) �2.3 �6.9 �12.2 �15.8 �19.2 �18.1 �17.3

CroatiaGDP at market pricesb �1.5 3.8 4.3 4.8 5.8 4.9 4.5Current account balance/GDP (%) 1.1 �5.2 �6.7 �7.7 �8.4 �8.0 �7.9

GeorgiaGDP at market pricesb �9.3 5.9 9.6 9.8 10.0 9.0 8.0Current account balance/GDP (%) — �8.3 �9.8 �13.8 �15.0 �14.1 �12.0

HungaryGDP at market pricesb 0.8 5.2 6.0 5.4 2.2 3.1 3.8Current account balance/GDP (%) �5.4 �8.5 �6.8 �5.7 �4.2 �5.0 �5.9

KazakhstanGDP at market pricesb �3.6 9.6 9.7 10.7 8.5 7.1 6.4Current account balance/GDP (%) �1.8 0.8 �1.9 �2.2 �2.0 �4.6 �7.7

Kyrgyz RepublicGDP at market pricesb �4.0 7.0 �0.6 2.7 7.5 7.0 6.7Current account balance/GDP (%) �10.6 �4.6 �9.3 �6.6 �17.9 �15.1 �12.2

LithuaniaGDP at market pricesb �3.3 7.0 7.5 7.4 7.8 6.8 6.0Current account balance/GDP (%) �5.9 �7.7 �7.1 �10.8 �13.3 �13.6 �12.3

LatviaGDP at market pricesb �2.8 8.6 10.2 11.9 9.7 7.4 6.4Current account balance/GDP (%) �1.6 �12.9 �12.7 �21.4 �22.7 �19.5 �15.3

MoldovaGDP at market pricesb �9.8 7.4 7.5 4.0 6.0 6.8 7.0Current account balance/GDP (%) — �2.2 �9.0 �9.3 �8.0 �14.9 �12.7

Macedonia, FYRGDP at market pricesb �0.9 4.1 4.1 3.0 5.0 5.0 5.5Current account balance/GDP (%) — �8.0 �1.5 �0.4 �2.9 �5.0 �6.1

PolandGDP at market pricesb 3.8 5.3 3.6 6.1 6.5 5.7 5.1Current account balance/GDP (%) �3.5 �4.2 �1.9 �2.4 �4.3 �5.3 �5.7

RomaniaGDP at market pricesb �1.7 8.4 4.1 7.7 6.1 5.9 5.5Current account balance/GDP (%) �4.8 �8.5 �8.7 �10.5 �13.9 �15.3 �14.9

Russian FederationGDP at market pricesb �3.9 7.1 6.4 6.7 7.5 6.5 6.0Current account balance/GDP (%) — 10.0 11.1 9.7 5.7 4.3 2.2

continue to deteriorate through 2009, largelybecause of declines in the terms of trade forhydrocarbon-exporting countries as oil pricesbegin to soften. Inflationary pressures arelikely to ease over the medium term, withmedian GDP inflation coming down from6.9 percent in 2007 to 6.4 percent in 2008 and5.5 percent in 2009. This decline is tied togenerally tighter credit conditions in both in-ternational and domestic markets. However,this somewhat sanguine picture masks an ex-pected rise in inflation pressures in Belarusand Georgia and slower progress toward sta-bilization of consumer price inflation amongoil exporters (as well as in Moldova andUkraine) because of strong demand pressures.Among the new EU member countries, onlythe Slovak Republic is anticipated to join theEuro Area in the coming years followingSlovenia’s entry in 2007.

RisksDownside risks to regional growth are tied topotential overheating and a sudden unwindingof large external imbalances. The presence oflarge foreign banks in several countries in theregion, and the fact that many of these countries

share common creditors and investors, appearsto expose them to higher contagion risk in theevent of a nondiscriminatory pullout, similar towhat occurred in East Asia in 1997. In particu-lar, large current account deficits (equivalent toabout 12 percent or more as a share of GDP)in Bulgaria, Georgia, Latvia, Lithuania, andRomania remain a concern.

Risks to growth are also associated withthe slowing of reform momentum in the newEU member states and other countries in CEE.In the CIS, slow progress with economic re-forms and only gradual diversification fromcommodity market dependence remain a con-cern, pointing to slower medium-term growthprospects. And a more protracted workout ofthe housing situation in the United States andassociated financial distortions there and inother OECD markets present a substantialdownside risk.

Higher than anticipated oil prices also pre-sent risk for energy-importing countries in theform of higher import bills and increased in-flationary pressures. With respect to the latter,while countries with free-floating currencyregimes in CEE have direct policy levers tomanage inflationary pressures, those with


175

Slovak RepublicGDP at market pricesb 0.3 5.2 6.6 8.8 8.7 7.1 6.8Current account balance/GDP (%) — �3.1 �8.4 �8.0 �4.3 �3.5 �3.1

TurkeyGDP at market pricesb 3.6 8.9 7.4 6.1 5.1 5.4 5.7Current account balance/GDP (%) �1.1 �5.2 �6.2 �8.1 �7.5 �7.7 �7.6

UkraineGDP at market pricesb �8.0 12.1 2.7 7.1 6.3 5.5 5.0Current account balance/GDP (%) — 10.7 2.9 �1.5 �3.6 �6.5 �7.4

UzbekistanGDP at market pricesb �0.2 7.7 7.0 7.3 7.7 5.0 5.0Current account balance/GDP (%) — 10.1 14.9 18.7 14.3 11.9 8.7

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. Bosnia and Herzegovina, Serbia and Montenegro, Tajikistan, and Turkmenistan are not forecastbecause of data limitations.— = 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.

Table A4 (continued )(annual percent change unless indicated otherwise)

Forecast

1991–2000a 2004 2005 2006 2007 2008 2009

currencies pegged to the euro, such as theBaltic states and Bulgaria, have more limitedoptions. Higher than projected grain pricescould also lead to increased inflationary pres-sures, particularly among the low-incomecountries, where food expenditures representa large share of consumption.

Despite gains in a number of MillenniumDevelopment Goal indicators, some countriesin the region, particularly in Central Asia andthe Caucasus, have shown regressing trendsand slower progress toward achieving thegoals. The Caucasus have met the goalsrelated to carbon emissions and primary edu-cation, but concerns remain about the nonin-come poverty indicators, such as malnutrition,access to tertiary education, HIV/AIDS, theenvironment, and soil and water management.

Latin America and the Caribbean Recent developmentsMarking the fourth consecutive year of sus-tained advances, GDP growth in Latin Amer-ica and the Caribbean registered 5.1 percent in2007, following a 5.6 percent gain in 2006.The average yearly rate of output growth since2004 has been 5.3 percent, twice the 2.7 per-cent registered during the previous 15 years.Recent growth has been more broadly based,with positive results shared by all subregions:the Caribbean, Central America, and SouthAmerica. A favorable external environmenttogether with improved domestic macro-economic conditions helped strengthenfundamentals and enhanced growth and stabil-ity (figure A7).

During 2006, the region recorded large cur-rent account surpluses, which have diminishedto a degree in 2007. Growing foreign exchangerevenues made up of export earnings linked tohigh commodity prices for food, metals, andenergy and continued large FDI; portfolio in-vestment; and remittance flows have all con-tributed to the maintenance of high levels offoreign reserves and helped support equity

markets. This positive external situation hasunderpinned government finances by boostingrevenues that, despite a significant increase inpublic spending, limited the region’s primarydeficit to 0.3 percent of GDP in 2007, from0.5 percent the previous year.

Monetary authorities, helped by strongerfiscal positions and supportive exchange rates,have been able to achieve inflation targets inmost countries (figure A8). Excluding Argentinaand the República Bolivariana de Venezuela,average consumer price inflation was stable in2007 at 5.7 percent after declining by almost1.0 percentage point in 2006. Only one coun-try in the region has experienced inflationabove 10 percent in each of the last five years.

The recent turmoil in financial markets thatoriginated in the U.S. subprime mortgage marketappears to have had limited effects on the regionto date. Spreads have increased, though capitalflows have continued (figure A9). Indeed, growthin 2007 continued to be strong, and although anysharp slowdown in the United States would even-tually affect Latin American and the Caribbeanprospects, the region seems better prepared forexogenous shocks than it was during earlierperiods of crisis or financial dislocation.


176

Figure A7 Growth outturns were mixedacross Latin America in 2007



2

Latin

Am

erica

and

the

Caribb

ean

Mex

ico

4

6

Brazil

Argen

tina

Caribb

ean

10

8

Centra

l Am

erica

2005

2006

2007

This broadly positive picture for the regionis qualified by substantial variations fromcountry to country. Between 2005 and 2007,Brazil, which accounts for about one-third ofthe region’s GDP, stepped up growth by al-most 1 percentage point a year. Significantmonetary policy easing has been a key factorbehind increasing private demand, which to-gether with higher public spending, hasboosted GDP growth. The policy interest ratewas reduced further in the first half of 2007 in

line with expectations of additional easing ininflation. Following the credit crisis in theUnited States, the exchange rate of the Brazil-ian real had depreciated from R/$1.86 in mid-July to R/$2.06 in mid-August. The currencyhad fully recovered and appreciated furtheragainst the U.S. dollar to R/$1.78 by the end ofNovember, a move of 9.5 percent. Ample inter-national reserves, a continuing current accountsurplus, and other strong macroeconomicfundamentals suggest increased resilience. In2007, Chile resumed rapid 5.7 percent growthat the same level as in 2005 despite a dip to4 percent in 2006 caused by the delayedeffects of monetary tightening, countercyclicalfiscal policy, mining stoppages, and energyconstraints.

In Mexico, despite uncertainties surround-ing the presidential election during the firsthalf of 2007, investment demand rose 10.3percent. High oil prices supported export rev-enues, which offset increased spending on im-ports and helped contain the trade deficit to$13 billion, up from $6 billion in 2006. De-spite the positive performance of investment,private consumption, and manufacturingoutput in the first half of 2007, concernsregarding a weakening U.S. economy and itsrepercussions for Mexico have mounted be-cause of the trade and financial links, includ-ing migrants’ remittance flows, between Mex-ico and its northern neighbor. Mexico’s GDP isanticipated to grow by 2.9 percent for the year.

Growth in Colombia and Peru has beenabove the regional average thanks to sustainedstrength in investment, which was up 18 per-cent in Colombia and 19 percent in Peru dur-ing 2006. Investment has been led by the pri-vate sector, with foreign companies playing animportant role. The improved security situa-tion in Colombia and several massive projectsin the mining and energy sectors in Peru haveattracted investors, and FDI has risen consid-erably. These factors will continue to supportthe pace of economic activity in 2007 at ratesof 6.5 for Colombia and 7.5 percent for Peru.

Central American countries have also per-formed exceptionally well in recent years.


177

Figure A8 Latin American inflationeases over the last 15 years

Median consumer price inflation in the seven largest LatinAmerican countries (percent change)

Source: World Bank.

0

1990

1992

1994

1996

1998

2000

2002

2004

2006

5

35

10

15

20

25

30

45

40Median

Average 2003–07

Average 1990–98

Average 1999–2002

Figure A9 Latin America and theCaribbean sovereign bond spreadsdecline, then increase again

Emerging markets bond index global

Brazil

Mexico


32530027525022520017515012510075

Jan.

2, 2

007

Feb.

11, 2

007

Mar

. 23,

200

7

May

2, 2

007

Jun.

11, 2

007

Jul. 2

1, 2

007

Aug. 3

0, 2

007

Oct. 9

, 200

7

Nov. 1

8, 2

007

Basis points

Source: JPMorgan-Chase.

Average growth for the aggregate of CostaRica, El Salvador, Guatemala, Honduras,Nicaragua, and Panama has amounted to 5.5percent over 2004–07, the strongest since theearly 1990s. Strong remittance inflows andthe recent implementation of the DominicanRepublic–Central American Free-TradeAgreement (a bilateral free trade agreementbetween the United States, Central America,and the Dominican Republic) have under-pinned increasing consumer and business con-fidence and domestic spending. By increasingexports of manufactured goods to the UnitedStates, this agreement is helping CentralAmerica reduce export market losses in thegarment sector resulting from the expirationof the Multi-Fiber Agreement and increasedcompetition from China.

Growth in Argentina has been easinggradually from extremely high rates. GDPincreased 7.8 percent in 2007, down from8.5 percent in 2006. The prolonged expansionin Argentina can be explained in part by an un-dervalued real exchange rate, an expansionaryfiscal policy, and an accommodative monetarystance. Contrasted with the experience ofBrazil, where the currency appreciated andfiscal deficits contracted, in Argentina, inter-vention in the foreign exchange market hasprevented nominal appreciation, while steril-ization operations have contained expansionof the monetary base. Argentina has also puta series of supplementary measures in placeto suppress inflation (actual levels of infla-tion remain unclear given the lack of trans-parency in official consumer price inflationcalculations): the government raised exporttaxes on food and fuel and imposed directcontrols on basic consumer prices, formalwages, and the tariffs on most energy prod-ucts and public services. Despite massive rev-enue growth, the government’s fiscal policyhas been procyclical (increasingly so in thefirst half of 2007). Were external conditionsto deteriorate, Argentina would have littlemargin to devalue the real exchange rate fur-ther or to contract its fiscal surplus in ameaningful way.

In the República Bolivariana de Venezuela,massive oil earnings have continued to financelarge and procyclical government spending.These outlays supported growth of 8.3 per-cent in 2007, down 2 percentage points fromthe previous year. Apart from a clear accelera-tion in inflation, substantial export revenueshave masked the adverse effects of increasedstate intervention on the economy. Significantdeclines in oil prices or oil production couldlead to large liquidity problems in the future,especially in light of the increased spending,including spending on nationalized enterprisesand numerous social programs.

The outcomes of seven presidential elec-tions held in the region during 2006 did notresult in major shifts in macroeconomic policyfor the region as a whole. However, policy inArgentina, Bolivia, the República Bolivarianade Venezuela, and more recently Ecuador andNicaragua is now more oriented toward an in-creasing role for the state in the economy.

Medium-term outlook Regional GDP is expected to slow further inthe years ahead, coming in at 4.5 percent in2008 and at 4.3 percent in 2009 (figure A10).This measured slowdown is supported by con-tinued strong growth in Brazil and a reboundfrom a weak 2007 for Mexico. Growth inother countries—notably Argentina and theRepública Bolivariana de Venezuela—is likelyto slow. Excluding those two countries, re-gional GDP growth is expected to moderateonly marginally from 4.4 percent in 2007 to4.2 percent in 2008—because of weakness inthe United States—before picking up to 4.3 per-cent in 2009.

Should these outturns be realized, theywould represent the longest positive growthspell for Latin America since the 1960s. De-spite a gradual worsening of current accountbalances due to stabilizing commodity pricesand slower growth in global demand, thisstronger growth is likely to persist, supportedby continued expansion in consumptionand investment and buoyed by an environ-ment of low inflation (excluding Argentina


178

and the República Bolivariana de Venezuela);improved fiscal policy (particularly in Mexico);and continued strong capital inflows (espe-cially to Brazil).

Among various groupings of Latin Ameri-can economies, a number of themes emerges.Growth among agricultural exporters is ex-pected to slow from 7.1 percent in 2007 to4.6 by 2009. However, if Argentina—wheregrowth is expected to slow to more sustain-able rates—is excluded from the group, thedeceleration is less marked, from 5.0 percentin 2007 to 4.5 in 2009. Growth among metalsexporters is projected to remain buoyant, eas-ing from 5.1 percent in 2007 to 4.7 percent by2009, in large part because of expansionarypolicies in Brazil and Chile. Growth amongenergy exporters is expected to slow gradually,from 5.2 percent in 2007 to 4.4 percent in2008, easing further to 4.2 percent in 2009as oil prices begin to soften. A reduced pace


179

Table A5 Latin America and the Caribbean forecast summary(annual percent change unless indicated otherwise)

Forecast

1991–2000a 2004 2005 2006 2007 2008 2009

GDP at market pricesb 3.4 5.9 4.6 5.6 5.1 4.5 4.3GDP per capitac 1.7 4.6 3.3 4.3 3.8 3.2 3.0Purchasing Power Parity GDPd 4.3 5.6 4.5 5.4 5.1 4.6 4.4


Net exports, contribution to growth �0.3 �0.4 �0.7 �1.4 �1.3 �1.2 �0.9Current account balance/GDP (%) �2.8 1.0 1.4 1.6 0.5 0.1 �0.2GDP deflator (median, LCU) 10.9 7.2 6.6 10.0 8.7 5.1 4.2Fiscal balance/GDP (%) — 0.0 �0.7 �0.5 �0.3 �0.9 �1.1

Memo items: GDPLAC excluding Argentina 3.2 5.5 3.9 5.1 4.7 4.3 4.3Central America 3.6 4.1 3.0 4.9 3.2 3.4 3.8Caribbean 3.6 2.6 6.7 8.8 5.5 5.2 5.0

Brazil 2.7 4.9 2.9 3.7 4.8 4.5 4.5Mexico 3.5 4.1 2.8 4.8 2.9 3.2 3.6Argentina 4.5 9.0 9.2 8.5 7.8 5.7 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. Measured in U.S. dollars.d. GDP measured at purchasing power parity exchange rates.e. Exports and imports of goods and nonfactor services.

Figure A10 Growth in Latin America andthe Caribbean eases into 2009

Real GDP (annual percent change)

Source: World Bank.

0

�2

2000

2001

2002

2003

2004

2005

2006

2007

2008

2009

4

6

2

10

8

Forecast


North and Central America South America

Caribbean


180

Table A6 Latin America and the Caribbean country forecasts(annual percent change unless indicated otherwise)

Forecast

1991–2000a 2004 2005 2006 2007 2008 2009

ArgentinaGDP at market pricesb 4.5 9.0 9.2 8.5 7.8 5.7 4.7Current account balance/GDP (%) �2.9 1.9 2.8 3.4 2.2 2.0 1.3

Antigua and BarbudaGDP at market pricesb 3.3 4.3 5.3 11.5 5.0 4.4 4.2Current account balance/GDP (%) �6.0 �11.9 �8.7 �16.0 �16.6 �17.6 �16.3

BelizeGDP at market pricesb 5.9 4.6 3.1 5.6 3.5 3.3 3.4Current account balance/GDP (%) �7.3 �14.4 �14.6 �1.9 �4.4 �6.5 �5.7

BoliviaGDP at market pricesb 3.8 3.9 4.1 4.6 4.1 4.4 4.2Current account balance/GDP (%) �6.1 3.9 5.4 11.8 9.2 8.4 7.9

BrazilGDP at market pricesb 2.7 4.9 2.9 3.7 4.8 4.5 4.5Current account balance/GDP (%) �2.1 1.9 1.7 1.4 0.7 �0.1 �0.2

ChileGDP at market pricesb 6.4 6.2 5.7 4.0 5.7 5.1 5.0Current account balance/GDP (%) �2.7 2.2 1.1 4.0 4.0 1.6 0.8

ColombiaGDP at market pricesb 2.5 4.8 4.7 6.8 6.5 5.3 4.8Current account balance/GDP (%) �1.9 �0.9 �1.6 �1.5 �1.6 �1.1 �1.0

Costa RicaGDP at market pricesb 5.2 4.1 5.9 8.2 6.1 5.0 4.9Current account balance/GDP (%) �3.6 �4.3 �4.9 �4.9 �4.8 �5.4 �5.0

DominicaGDP at market pricesb 1.8 3.2 3.4 4.1 3.2 3.1 3.0Current account balance/GDP (%) �16.6 �19.5 �28.8 �18.4 �19.2 �20.6 �20.7

Dominican RepublicGDP at market pricesb 6.0 2.0 9.3 10.7 7.2 5.4 4.8Current account balance/GDP (%) �3.2 5.7 �2.0 �2.6 �2.9 �3.8 �4.2

EcuadorGDP at market pricesb 1.8 7.9 4.7 4.1 2.4 2.5 2.7Current account balance/GDP (%) �2.3 �1.7 0.8 3.5 1.6 2.9 2.3

El SalvadorGDP at market pricesb 4.6 1.8 3.1 4.2 4.2 3.8 4.0Current account balance/GDP (%) �2.0 �4.0 �5.4 �4.7 �5.2 �6.1 �5.7

GuatemalaGDP at market pricesb 4.1 2.7 3.2 4.6 5.0 4.6 5.0Current account balance/GDP (%) �4.6 �4.4 �4.4 �4.3 �4.1 �5.3 �5.0

GuyanaGDP at market pricesb 4.9 3.3 �1.9 4.7 4.5 3.7 3.5Current account balance/GDP (%) �15.1 �2.5 �12.0 �26.6 �21.2 �22.8 �16.4

HondurasGDP at market pricesb 3.3 5.0 4.1 6.0 6.0 5.5 4.7Current account balance/GDP (%) �7.7 �5.7 �1.4 �1.9 �4.9 �5.0 �5.2

HaitiGDP at market pricesb �1.3 �2.2 2.0 2.3 3.5 3.8 4.0Current account balance/GDP (%) �1.8 �1.7 1.4 �0.3 �2.0 �3.8 �3.8

JamaicaGDP at market pricesb 1.9 1.1 1.8 2.5 1.1 3.0 3.1Current account balance/GDP (%) �2.7 �5.7 �11.1 �10.7 �11.5 �13.7 �13.6

MexicoGDP at market pricesb 3.5 4.1 2.8 4.8 2.9 3.2 3.6Current account balance/GDP (%) �3.7 �1.0 �0.6 �0.2 �1.0 �0.9 �1.0


181

NicaraguaGDP at market pricesb 3.4 5.1 4.4 3.7 4.2 4.7 4.5Current account balance/GDP (%) �28.7 �14.6 �13.0 �13.9 �11.4 �12.1 �10.2

PanamaGDP at market pricesb 5.1 7.6 6.4 7.5 8.3 7.2 7.1Current account balance/GDP (%) �4.8 �7.5 �5.1 �0.2 �5.6 �6.8 �7.5

PeruGDP at market pricesb 4.0 5.2 6.7 7.6 7.5 6.4 6.1Current account balance/GDP (%) �5.5 0.0 1.5 2.9 1.8 0.8 0.4

ParaguayGDP at market pricesb 1.8 4.1 2.7 4.0 4.0 3.9 3.8Current account balance/GDP (%) �2.2 2.0 0.0 �2.6 �2.6 �3.8 �3.8

St. LuciaGDP at market pricesb 3.1 3.9 5.8 5.4 5.2 5.1 5.0Current account balance/GDP (%) �11.4 �16.6 �22.5 �18.9 �16.8 �16.9 �15.0

St. Vincent and the GrenadinesGDP at market pricesb 2.1 5.4 2.2 4.5 5.5 6.3 5.9Current account balance/GDP (%) �19.8 �29.0 �29.1 �24.2 �19.8 �17.7 �13.9

UruguayGDP at market pricesb 3.0 11.8 6.8 7.0 5.5 4.2 3.8Current account balance/GDP (%) �1.5 0.0 0.0 �2.7 �2.7 �3.3 �2.9

Venezuela, R.B. deGDP at market pricesb 2.1 17.9 10.3 10.3 8.3 5.8 4.2Current account balance/GDP (%) 2.6 14.1 18.5 14.4 7.4 6.9 4.7

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.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.


Forecast

1991–2000a 2004 2005 2006 2007 2008 2009

of economic activity in Argentina and theRepública Bolivariana de Venezuela couldspeed up the regional slowdown for energyexporters, but a pickup in growth in Mexicowould offset that to some extent.

Despite the gradual reduction in oil prices,growth among small energy importers islikely to moderate from 6 percent in 2007 to5.1 percent by 2009. This outturn reflects thecombined effects of weaker import demand inthe United States, reduced remittance inflows(linked to a slowdown of construction activ-ity in the United States), and increasing com-petition from China following the phaseoutof textile and apparel quotas under the Multi-Fiber Agreement. At the same time, GDPshould find support through continued strongFDI inflows following implementation of the

Dominican Republic–Central American Free-Trade Agreement.

Growth in Brazil is expected to moderatefrom 4.8 percent in 2007 to 4.5 percent inboth 2008 and 2009. The positive fundamen-tals observed in 2006 and 2007—low interestrates, a strong currency, and falling unem-ployment—should continue to underpin a ro-bust pace of domestic activity. At the sametime, a further near-term acceleration ingrowth is unlikely. The Central Bank hastaken a pause in interest rate reductions,while quarter-on-quarter GDP figures show aslight moderation from the first quarter to thesecond quarter of 2007.

Argentina’s growth is expected to decel-erate sharply from 7.8 percent in 2007 to5.7 percent in 2008 and to 4.7 percent in

2009. Despite rising fiscal revenues, persistentand accelerating inflation and increased publicintervention in private sector activity appearlikely to discourage growth prospects. Recentproduction and goods trade numbers under-score a weakening in the pace of activity,caused in large part by the imposition ofenergy rationing. Industrial output gained2.3 percent in July (year-on-year), the weakestshowing in more than five years, while steelproduction contracted by 26 percent.

Growth in Mexico is likely to reboundfrom 2.9 percent in 2007 to 3.2 percent in2008 and to 3.6 percent in 2009. Although re-mittances will contribute less to domestic ac-tivity as growth in the United States slows, in-creased domestic credit to consumers andbusinesses should support spending. Recentfiscal reforms also point toward better growthprospects, as improved revenue collectionshould allow the government to boost neededinfrastructure outlays. With a rebound of U.S.activity in the second half of 2008 and 2009,Mexican exports will make a stronger contri-bution to growth.

Finally, activity in the República Bo-livariana de Venezuela is anticipated to slowfairly rapidly, with GDP growth decliningfrom 8.3 percent in 2007 to 5.8 percent in2008 and 4.2 percent in 2009. Activity in late2006 and early 2007 was supported by large-scale run-ups in fiscal outlays associated withthe presidential elections, but falling oil rev-enues over the forecast period make continuedexpansion in government spending unlikely.Private activity is also expected to slow sub-stantially. Higher inflation will likely result indeclines in real wages, a sharp falloff in FDIwill reduce the contribution of investment togrowth, and an uncertain regulatory environ-ment (notably with regard to property andcontract rights) is anticipated to hamper pri-vate sector startups and job creation.

RisksA stabilization of growth in Latin Americaand the Caribbean at high rates in a historicalcontext and broad improvements in the region’s

fundamentals have been supported by an ex-ceptionally favorable external environment.Should that environment deteriorate substan-tially, growth is likely to slow. Moreover, despiteimproved fundamentals, some countries may bevulnerable to a sharp reversal in conditions.

Overall, the region’s resilience to shockshas improved, and better economic policieshave supported this improvement. In particu-lar, the structure of external debt—in terms ofmaturity and foreign currency exposure—hasimproved; the accumulation of reserves has, inmost cases, surpassed external borrowing re-quirements; the region’s fiscal and current ac-counts have been in surplus during the currentgrowth cycle; and independent central banksare imposing countercyclical monetary poli-cies, especially in those countries where infla-tion targeting is the adopted policy regime.Notwithstanding these advances, some riskslinked to a worsening of the external environ-ment may remain a concern for the region.

In a potentially difficult phase of rising riskaversion in global financial markets, the regionhas, up to now, demonstrated that it has de-veloped a stronger “immune system.” Asshown in figure A11, the responsiveness of thespreads on sovereign bonds (noninvestmentgrade) for Latin American and the Caribbeancountries with respect to changes in risk aver-sion, though still elevated, is much lower whencontrasting the July 2006 to end-2007 periodwith the August 1997 to August 1999 period(the period of the financial crises in East Asiaand Russia). Contagion effects to Latin Americaand the Caribbean through the financial chan-nel seem less likely than during past challengingphases. However, there are still some country-specific issues: Argentina (because of theINDEC incident—the national statisticsagency producing questionable infection data)and Ecuador (because of the outright discus-sion of default) have undermined the percep-tion of their willingness to pay, regardless oftheir reserves cushion, and therefore were hitharder and earlier than the other countries.The República Bolivariana de Venezuela wasalso affected because of political concerns.


182


183

EMBI Latin index (noninvestment grade)

EMBI Latin index(noninvestment grade)

High yield index

y � 0.9717x � 0.6581

y � 0.6955x � 1.5904

Figure A11 Financial test: Credit

Lower elasticity of LAC EMBI with respect to riskaversion (HY bonds)…

4.85.8 5.9 6.0 6.1 6.2 6.3 6.4 6.5 6.6 6.7

5.0

5.1

5.2

5.3in

5.8

6.0

5.9

6.1

6.2

6.3

4.9

Source: Bloomberg weekly data; Credit Suisse high-yieldindex; JP Morgan EMBI.

Note: EMBI � Emerging markets bond index.

August 1997–August 1999(left axis)

July 2006-current(right axis)

The region could still suffer adverse conse-quences from a worsening and spreading ofthe U.S. subprime crisis through real channels.Lower remittances may hurt some of thepoorest among Latin America and theCaribbean’s recipient countries, while lowerexternal demand may be exacerbated byweaker competitiveness resulting from appre-ciating currencies, as shown in figure A12.

Finally, terms-of-trade deterioration mayalso exert negative pressure on aggregateoutput. Figures A13 and A14 summarize theregion’s vulnerability to commodity priceshocks. In recent years, export concentrationfor the region, as captured by the Herfindahlindex of exported goods, did not decline. In-deed, it increased in many countries, possiblybecause of an increase in specific commodityprices (as the index is measured in nominalterms). Increases in this index signal a potentialworsening of the risk associated with volatilityof global commodity prices. Counterbalancingthis tendency is the geographic diversificationof Latin American and the Caribbean’s exportmarkets. The concentration of Latin exports tothe U.S. market is quite strong, especially for

Real effective exchange rate indexes (January 2000 � 100)

Figure A12 Exchange rate policydilemmas?

301993 1995 1997 1999 2001 2003 2005 2007

150

170

190

210

130

110

90

70

50

Source: International Monetary Fund.

Argentina Brazil Colombia

Peru

Mexico

countries such as Mexico, but the overall trendis toward diversification.

The emergence of China and India playeda role in explaining the trade patterns dis-played in figures A13 and A14, and on apositive note, may have permanently and fa-vorably changed the terms of trade faced bythe region. In this rosier situation, the regionmay be drawn into intensive activities in areas

Export products concentration, Herfindahl index

Figure A13 Export product (value)concentration is increasing

0

Argen

tina

Brazil

Chile

Colom

bia

Ecuad

or

Mex

icoPer

u

Urugu

ay

Austra

lia

Canad

a

New Z

eland

Latin

Am

erica

and

the

Caribb

ean

0.30

0.35

0.25

0.20

0.15

0.10

0.05

Sources: World Integrated Trade System.

2005

1996

of comparative advantage, such as naturalresources and skills and technology, but addi-tional progress on some structural factors isneeded to take full advantage of these oppor-tunities. In particular, the region has to im-prove its investment climate and the skill com-position of its labor force.

Compared with other regions, LatinAmerican and Caribbean fundamentals mayappear to be unimpressive, and the region doesnot seem to be catching up in a significant way.Even for the most recent period, aggregateGDP growth, gross fixed capital formation (asa share of GDP), and TFP growth remain wellbelow averages for other developing regions.The Latin America and the Caribbean regiondoes not appear, at least not yet, to have beencapitalizing on the favorable external environ-ment to raise its growth path.

Middle East and North Africa Recent developmentsGDP for the developing countries of the Mid-dle East and North Africa region grew nearly5 percent in 2007, matching the decade-highpace achieved the preceding year (table A7


184

Figure A15 Growth in Middle East andNorth Africa picks up

02000

8

6

4

5

2


7

1

3

Source: World Bank.

2001 2002 2003 2004 2005 2006 2007 2008 2009


Oil exporters Diversified exporters

and figure A15).1 Continued growth in hy-drocarbon receipts among the region’s devel-oping (and high-income) oil exporters was akey factor supporting growth, as world oilprices breached records at close to $100 abarrel toward the end of the year. Notwith-standing the severe drought that afflictedcountries in the Maghreb, notably Morocco,a revival in European demand helped under-pin exports for the resource-poor, labor-abundant countries in the region, especiallythe Arab Republic of Egypt, where a depreci-ation of the pound against the euro also as-sisted, and Jordan.2 Moreover, record inflowsof FDI, ample liquidity, and strong domesticdemand all bolstered growth across the di-verse countries of the region.

Oil exporters registered growth of 4.5 per-cent for the year, up from 4.0 percent in 2006,as GDP in Algeria and the Islamic Republic ofIran moved higher. The group of diversifiedexporters witnessed an easing of growth from6.2 percent in 2006 to 5.4 percent, largely asoutput gains in Morocco plummeted from8 percent to 2 percent. However, GDP in allother countries picked up or repeated itsstrong performance of 2006.

Oil-related revenues for the year increased9 percent for developing exporters in the

Export market concentration, Herfindahl index

Figure A14 Export market (value)concentration is falling

0

Argen

tina

Brazil

Chile

Colom

bia

Ecuad

or

Mex

icoPer

u

Urugu

ay

Austra

lia

Canad

a

New Z

eland

Latin

Am

erica

and

the

Caribb

ean

0.6

0.7

0.5

0.4

0.3

0.2

0.1

Sources: World Integrated Trade System.

2005

1996

volume constraints on oil production andshipment, caused the current account surplusto narrow from $34.5 billion in 2006 to $31billion in 2007, or from 13.3 to 11.1 percentof GDP.

The dollar’s swift decline—tied in part toreductions in U.S. interest rates to shore upthe country’s interbank market—has had dif-fering effects in the region. Oil exporters findtheir dollar-based receipts falling sharply inrelation to the euro, the currency in which amajor proportion of imports are denomi-nated, and are consequently suffering a formof terms-of-trade loss. In contrast, diversifiedexporters may find their competitivenessenhanced in the EU market, with local cur-rencies depreciating moderately against theeuro.


185

Table A7 Middle East and North Africa forecast summary(annual percent change unless indicated otherwise)

Forecast

1991–2000a 2004 2005 2006 2007 2008 2009

GDP at market pricesb 3.8 4.8 4.3 5.0 4.9 5.4 5.3GDP per capitac 1.6 3.0 2.5 3.1 3.1 3.6 3.5Purchasing power parity GDPd 4.7 4.8 4.3 5.0 4.9 5.4 5.1


Net exports, contribution to growth 0.7 �1.9 �1.2 �1.6 �2.3 �1.5 �1.5Current account balance/GDP (%) �0.5 7.4 9.6 9.6 8.2 9.5 6.2GDP deflator (median, LCU) 7.4 9.1 11.7 8.7 4.4 6.1 4.3Fiscal balance/GDP (%) �2.8 �2.7 3.4 2.2 �0.9 �1.8 �1.8

Memo items: GDPMENA Geographic regionf 3.4 5.0 5.2 4.8 4.6 5.1 4.9Resource poor-labor abundantg 4.2 4.8 3.8 6.2 5.4 6.3 6.1Resource rich-labor abundanth 3.3 4.9 4.6 3.8 4.4 4.6 4.4Resource rich-labor importingi 3.0 5.2 6.8 4.7 4.2 4.6 4.5

Egypt, Arab Rep. of 4.6 4.1 4.5 6.8 7.1 7.0 6.8Iran, Islamic Rep. of 3.7 5.1 4.3 4.6 5.0 5.0 4.7Algeria 1.7 5.2 5.1 1.8 3.4 4.0 3.8

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. Measured in U.S. dollars.d. GDP measured at purchasing power parity exchange rates.e. Exports and imports of goods and nonfactor services.f. Geographic region includes high-income countries: Bahrain, Kuwait and Saudi Arabia.g Egypt, Jordan, Lebanon, Morocco and Tunisia.h. Algeria, Iran, Syria and Yemen.i. Bahrain, Kuwait, Oman and Saudi Arabia.

region, rising to $160 billion. Among high-income oil exporters, oil-related revenues rose13 percent to $382 billion, sufficient to fundongoing infrastructure and social programswhile also adding to massive reserve levels.3

Among developing country oil exporters,which are dominated in size by Algeria andthe Islamic Republic of Iran, but also includeOman, Syria, and the Republic of Yemen, theupturn in oil prices had varying effects (figureA16). For example, in Algeria, hydrocarbonreceipts increased moderately to $54 billion,widening Algeria’s current account surplusslightly to $29 billion, or 23 percent of GDP.In the Islamic Republic of Iran, hydrocarbonrevenues grew 12 percent to $65 billion, butrapid growth in consumer and governmentspending (as well as imports), together with

A fall in hydrocarbon output has con-strained growth in Algeria, with GDPadvancing just 1.8 percent in 2006 and 3.4percent in 2007. Oil and gas output growthdeclined 2.6 percent in 2006, but activity un-related to hydrocarbons expanded by a robust6 percent in 2007. A major government in-vestment initiative got under way and is slatedto expend more than $22 billion over the com-ing years on housing, transport, and agricul-ture. This is boosting job growth in construc-tion and related sectors and underpinningstrong household spending.

In the Islamic Republic of Iran, despitemajor fiscal expansion in 2007—seen in thewidening of the budget deficit from 0.2 per-cent of GDP in 2006 to 2.4 percent—growthis likely to step up just 0.4 percentage points,to 5 percent given the extent of leakage intoimports that occurred over the year. Exportsadvanced a modest 1.3 percent against a13.5 percent gain in imports.

For the diversified exporters, or resource-poor, labor-abundant economies, 2007marked a watershed for several countries inthe area of finance. The Fitch Agency raised

Egypt’s Issuer Default Rating to a positive out-look, with growth supported by double-digitgains in investment. Morocco was awardedinvestment grade status for its sovereignbonds and quickly raised $675 million at alow 55 basis-point spread. GDP growth forthe group—excluding Morocco, where theeffects of drought bias figures for 2007—increased from 5.8 percent in 2006 to 6.2 per-cent in 2007 on the strength of 7.1 percentgrowth in Egypt, 6.0 percent in Jordan andTunisia, and recovery for Lebanon from thedownturn of 2006.

Growth in Egypt is broadly based, withnon-oil manufacturing and retail trade ac-counting for half of overall output gains. Thefastest-growing sectors include construction,Suez Canal traffic, communications, andtourism. Exports have boomed by 23 percentover 2007 to date, especially non-oil goods,but at 29 percent growth, import demand iseven stronger and has kept the contribution oftrade to growth negative while widening thecountry’s deficit on trade. But for Egypt andmany of the diversified exporters, services re-ceipts and burgeoning remittances outweighshortfalls on trade, allowing countries tomaintain current account surplus positions(figure A17).

In Egypt, Morocco, and Tunisia, reformsare improving the business climate and in-creasing the competitiveness of the export sec-tor. Egypt, Jordan, Morocco, and Tunisiasigned a free trade agreement (the AgadirAgreement) to help promote intraregionaltrade while addressing rules-of-originquestions that often are part of broaderframeworks, such as the European Union–Mediterranean agreements. FDI is becomingan important driver for private investmentand growth in this group of countries, and asreforms proceed, the potential for attractingadditional FDI grows in step.

Indeed, large-scale FDI has been flowinginto the region for the last three years on theback of emerging and increasingly sustainedeconomic growth, diversification, and ongo-ing reforms. Flows to the geographic region


186

Figure A16 Hydrocarbon exports continueto rise on higher prices, modest volumegains

02000

175

125

75

100

25

Oil and gas exports ($ billions)

150

50

Source: United Nations Comtrade Database; World Bank.

2001 2002 2003 2004 2005 2006 2007

Iran, Islamic Republic of Algeria Oman

Yemen, Republic of Syrian Arab Republic

(including the high-income oil exporters) in-creased 38 percent, to $64 billion in 2006,after doubling to $46 billion in 2005. Amongthe diversified exporters, the share of FDI inGDP doubled from 3.0 percent, on average,over 2000–04 to 5.6 percent in 2006. FDIflows are tending to focus on the services sec-tor, including finance, telecommunications,and real estate, as well as hydrocarbons andrelated industries. Direct investment flows toEgypt, in particular, are booming, reaching$11 billion during the country’s fiscal 2007,up from $6 billion the previous year. High-income countries in the region have emergedas major investors in the Middle East andNorth Africa. The United Arab Emirates, forexample, invested $8 billion in Egypt in trans-port and tourism infrastructure during 2006.Evidence suggests that FDI continues to flowinto the region’s developing and high-incomecountries, setting a strong foundation for fu-ture growth and eventual alleviation of the re-gion’s unemployment problem.

Median inflation in the region currentlystands at 5.0 percent, up from 3.2 percent in2006. Inflation has become an important issuefor several countries, including the Islamic

Republic of Iran (18 percent year-on-year atlatest readings) and Egypt (10.9 percent),where consumer price inflation has trendedhigher because of domestic supply shocks, forexample, the partial elimination of subsidiesin September 2006. More recent accelerationin Algeria and other oil exporters has accom-panied rapid growth in domestic demand.Moreover, the ratcheting up of food and en-ergy costs—the former notably in grains, witha 40 percent jump over 2007 related, in, partto increased global production dedicated tobiofuels—presents difficulties for a number ofeconomies, causing price pressures throughimports as well as strains on fiscal positionsfrom widespread subsidies covering fuels andcereals.

Medium-term outlookProspects for developing countries in the re-gion appear relatively bright, with aggregateGDP gains projected to top 5 percent in 2008and 2009. Despite uncertainties looming inthe external environment for 2008 related tothe financial and real implications of the U.S.subprime mortgage crisis, growth amongdeveloping oil exporters is anticipated to pickup, in large part tied to domestic develop-ments, although oil prices will likely remain athigh levels ($84 per barrel), thereby helping tosustain revenues supporting large, project-related expenditures. GDP gains for the oil ex-porters are projected to reach 4.7 percent in2008 before receding to 4.6 percent in 2009 asglobal oil prices begin to recede.

For the diversified exporters, growthshould pick up sharply in 2008, to 6.3 percentfrom 5.4 percent in 2007. A rebound in Mo-rocco to 5.5 percent growth from the depths ofdrought is a factor in this outlook. Investment-led growth appears to be increasingly wellestablished in Egypt, where activity is expectedto remain within a higher 7 percent range overthe forecast period. Sustained growth ofaround 6 percent in Jordan and Tunisia islikely, grounded in services exports and in-creasingly in investment and constructionfunded by FDI. In Lebanon, stronger growth


187

Figure A17 Tourism and remittancesoffset widening trade deficits forMaghreb and Mashreq countries

02000

30

20

5

10

Balance of nonfactor services and net transfers ($ billions)

25

15

Source: International Monetary Fund balance of paymentsstatistics.

2001 2002 2003 2004 2005 2006

Net transfers Balance of services

on a sustained basis is unlikely until the politi-cal environment has improved (table A8).

RisksThe region appears to have suffered little inthe way of direct effects from the financialturbulence of mid-2007. Yet the increasing so-phistication of reserves management amonghigh-income oil exporters and the movementtoward establishing large-scale sovereignwealth funds may increase the exposure ofsuch portfolios to innovative, yet complexand hard to price, securities. Regional equity

markets slumped only briefly during the boutof global volatility, and indeed, stock ex-changes in the Gulf states have outperformedthe Morgan-Stanley emerging market aver-age through the final months of 2007 (fig-ure A18).

Markets for non-oil commodities, manu-factures, and tourism services may suffer amore pronounced slowdown linked to the rip-ple effects of financial difficulties already pre-sent in the United States and the Euro Area.Should a significant credit crunch occur, slow-ing growth across the OECD as well as across


188

Table A8 Middle East and North Africa country forecasts(annual percent change unless indicated otherwise)

Estimate Forecast

1991–2000a 2004 2005 2006 2007 2008 2009

AlgeriaGDP at market pricesb 1.7 5.2 5.1 1.8 3.4 4.0 3.8Current account balance/GDP (%) 3.2 13.1 20.7 24.2 23.0 24.4 21.2

Egypt, Arab Rep. ofGDP at market pricesb 4.6 4.1 4.5 6.8 7.1 7.0 6.8Current account balance/GDP (%) 0.0 4.3 3.3 1.6 2.1 2.4 1.9

Iran, Islamic Rep. ofGDP at market pricesb 3.7 5.1 4.3 4.6 5.0 5.0 4.7Current account balance/GDP (%) 1.2 15.9 16.2 13.3 11.1 12.7 6.8

JordanGDP at market pricesb 5.1 8.4 7.3 6.3 6.0 5.8 6.0Current account balance/GDP (%) �4.3 �0.2 �18.7 �15.0 �13.5 �14.6 �12.6

LebanonGDP at market pricesb 7.2 6.3 1.0 0.0 1.0 3.5 4.5Current account balance/GDP (%) — �22.5 �23.2 �16.7 �19.6 �19.0 �15.9

MoroccoGDP at market pricesb 2.2 5.2 2.4 8.0 2.0 5.5 4.5Current account balance/GDP (%) �1.4 1.7 1.9 2.8 2.3 0.6 0.5

OmanGDP at market pricesb 4.6 3.1 5.8 6.0 6.0 6.3 5.8Current account balance/GDP (%) �3.7 2.3 13.7 12.4 12.2 17.2 11.3

Syrian Arab RepublicGDP at market pricesb 5.1 3.9 4.5 5.1 3.9 3.7 4.8Current account balance/GDP (%) 1.0 2.4 1.1 2.8 2.5 3.2 0.8

TunisiaGDP at market pricesb 4.7 6.0 4.0 5.4 6.0 6.2 6.0Current account balance/GDP (%) �4.3 �2.0 �1.1 �2.1 �2.5 �2.6 �2.1

Yemen, Rep. ofGDP at market pricesb 5.5 2.5 4.6 4.0 3.8 4.3 4.0Current account balance/GDP (%) �4.3 1.7 2.3 0.4 �2.2 �2.3 �5.1

Source: World Bank.Notes: Growth and current account figures presented here are World Bank projections and may differ from targets contained inother World Bank documents. Djibouti, Gaza, Iraq, Libya, and West Bank are not forecast because of data limitations.— = 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.

developing countries, demand for crude oiland refined petroleum products could declineand lead to a sharp fall in prices with the at-tendant effects ensuing for revenues andgrowth.

Notes1. For the purposes of this report, coverage of the

region is restricted (for the sake of consistency acrossall regions) to include only low- and-middle-incomecountries. These countries include Algeria, the ArabRepublic of Egypt, the Islamic Republic of Iran, Jor-dan, Lebanon, Morocco, Oman, the Syrian Arab Re-public, Tunisia, and the Republic of Yemen. Because ofdata limitations, the middle-income economies of Dji-bouti, Iraq, Libya, and the West Bank and Gaza are notincluded in the aggregates. High-income countries inthe region are Bahrain, Kuwait, Qatar, Saudi Arabiaand the United Arab Emirates, to which references aremade. Qatar and the United Arab Emirates are not in-cluded in the high-income aggregate because of datalimitations.

2. For the developing countries in the region, the re-source-poor, labor-abundant countries are Egypt, Jor-dan, Lebanon, Morocco, and Tunisia. The resource-rich, labor-abundant countries are Algeria, the IslamicRepublic of Iran, Syria, and the Republic of Yemen.Oman is the developing, resource-rich, labor-importingcountry. However, memo items in table A7 cover thefull geographic region, including Bahrain, Kuwait, and

Saudi Arabia. For reference, the GDP of the developingcountries of the Middle East and North Africa regionaccounts for just 62 percent of the geographic region.

3. Changes in the volume of oil and gas productionhave been modest in recent years, ranging from 0.5 to1.0 percent annual gains in Algeria to a decline in theIslamic Republic of Iran and the Republic of Yemen.Hence the buildup in export revenues is largely due tothe large-scale increase in global oil prices over 2005through 2007.

South Asia Recent developmentsSouth Asia’s regional GDP was vibrant at 8.4percent growth in 2007, easing only moder-ately from the 8.8 percent outturn of 2006.The region continues to build on the momen-tum of recent fiscal and business-oriented re-forms. Private consumption and investmentaccelerated in 2007 despite more restrictivemonetary policies. Large capital inflows, ris-ing incomes, and strong worker remittanceshave supported private spending. Improve-ment in business sentiment—both foreign (to-ward India) and domestic—and rising corpo-rate profits have provided a strong foundationfor investment. Growth in government spend-ing has come down from 12.5 percent in 2006and is now in line with overall economicactivity. The decline in government spendingis primarily attributable to a sharp falloff inPakistan’s outlays following a spike in 2006linked to recovery and reconstruction effortsafter the December 2005 earthquake.

Among the larger economies, a modest eas-ing of GDP growth in India, from 9.4 percentin 2006 to 9.0 percent in 2007 reflects a firm-ing in Indian import demand that yielded anegative net export position, further under-pinned by strong appreciation of the rupee(figure A19). The falloff in growth in Pakistanfrom 6.9 percent in 2006 to 6.4 percent in2007 stems from a decline in private sectorcredit growth, a slowdown in FDI inflows be-ginning in July 2007, and weaker exports. InBangladesh, tighter domestic credit conditionsinduced a softening in investment growth,while net exports turned negative, explaining


189

Index, Jan. 1, 2007 � 100

Figure A18 Middle East and North Africaequities rebound from the mid-2007 slump

Jan.

2, 2

007

Feb.1

5, 2

007

Mar

. 31,

200

7

May

14,

200

7

Jun.

27,

200

7

Aug. 1

0, 2

007

Sep. 2

3, 2

007

Nov. 6

, 200

7

113

125

138

150

100

88

Source: Morgan-Stanley.

Egypt, ArabRep. of

Morgan-StanleyComposite Index total

GCC

Middle Eastand North Africa

the slight moderation in growth from 6.6 in2006 to 6.5 percent in 2007. In Sri Lanka, anescalation of the conflict between govern-ment-led forces and separatists dampenedprospects in the tourism industry and con-tributed to a 1.1 percentage point falloff ingrowth to 6.3 percent in 2007. In contrast,growth in Nepal is on course to pick up mod-estly from 2.3 percent in 2006 to 2.5 percentin 2007 as the peace process moves forward,bringing increases in foreign assistance and agradual recovery in the tourism sector.

At the regional level, robust investment de-mand drove rapid import growth—which av-eraged 11 percent a year and exceeded exportgrowth by 2.5 percentage points—while cur-rency appreciation also increased demand forimports. Exports remained healthy at 8.5 per-cent, but were slowed by softening demandin high-income countries and real apprecia-tions that have made export markets morecompetitive.

The regional current account deficit in-creased as a share of GDP, moving from 1.3percent in 2006 to 2.4 percent. Current ac-count positions worsened in a number ofcountries over the course of 2007, withdeficits reaching close to 5 percent of GDP in

Pakistan and about 2 percent in India.Pakistan’s current account is of some concern,having deteriorated by the equivalent of morethan 5 percentage points of GDP in the lastfour years. Strong flows of worker remittancesinto the region have helped contain the extentof deterioration in current account positions.Remittances amount to substantial propor-tions of GDP in many South Asian countries,helping to smooth private consumption whileoffering a buffer for external trade shortfalls.Remittance receipts are the equivalent of alarge 16 percent of GDP in Nepal, 9 percent inSri Lanka, 7 percent in Bangladesh, and 4 per-cent in Pakistan. Strong foreign capital flowsto India and Pakistan, in particular, have com-plemented remittances to provide generallycomfortable levels of external finance.

Inflation pressures in India and Pakistaneased over the first three quarters of 2007 (fig-ure A20). In India, the appreciating rupee con-tributed to an easing of wholesale price infla-tion to 3 percent as of early November(year-on-year), breaching a five-year low.However, risks of a regional revival ininflation remain, stoked by an incomplete


190

Figure A20 Monetary policy is tightenedin response to a buildup in inflation

3

Jan.

200

6

Apr. 2

006

Jul. 2

006

Oct. 2

006

Jan.

200

7

Apr. 2

007

Jul. 2

007

Oct. 2

007

10

8

6

7

4

Consumer price inflation (percent change year-on-year)and policy interest rates (percent)

9

5

Source: World Bank.

India policy rate

Pakistan policy rate

India Consumer Price Inflation

Pakistan Consumer Price Inflation

Figure A19 South Asian economies easeinto 2007


10.0

20052006

7.55.0


2.50

Nepal

Pakistan

India

Sri Lanka

Bangladesh

South Asia

2007

pass-through of higher energy costs to thefinal consumer in several countries, increasingand widespread upward pressure on foodprices, strong credit growth, and sharp gainsin equity markets that have fueled liquidity.In Sri Lanka, inflation reignited in 2007, with12-month moving average inflation increasingto 18 percent in September, up from 6.8 a yearearlier. Inflation has also picked up in Mal-dives, from 3.7 percent in 2006 to 7.0 percentin 2007, reflecting expansionary governmentspending and a budget deficit anticipated toreach 24 percent of GDP in 2007, caused inpart by continuing post-tsunami reconstruc-tion and large tourism construction projects.Inflation also appears to be stepping up inBangladesh in the wake of the recent cyclone.

In India, monetary tightening and largecapital inflows led to significant currencyappreciation during 2007, with the rupeereaching a near decadal high of Rs 39.2against the dollar by the end of November.This development has contributed to an easingof inflationary pressures through prices ofimports, while at the same time it has reducedthe price competitiveness of India’s exporters.The strong currency gains prompted theReserve Bank of India to sell rupees to helpcontain the rate of appreciation and to intro-duce selected capital control measures. Pak-istan’s currency had been appreciating slightlyagainst the dollar over much of the year, but itbegan to depreciate in response to heighteneduncertainty on international capital marketsand the president’s imposition of a state ofemergency in early November.

The increased volatility of internationalcredit markets during mid-2007 tied to thehousing market downturn in the United Statesinitially led to a falloff in equity prices in Indiaand Pakistan, but markets in both countrieshave since recouped their losses. In October,India’s Sensex index rose above 19,000 for thefirst time, driven by overseas fund purchases.According to the Securities and ExchangeBoard of India, foreign investors bought anaverage of nearly $500 million more than theysold in Indian equities during the first half of

October, compared with daily net purchasesof $232 million in September. In relation tolocal currencies, equity markets were up 36percent in India (50 percent in dollar terms)and 20 percent in Pakistan (19 percent indollar terms) from January 1 to October 31,2007.

Medium-term outlookTighter credit conditions, volatility in interna-tional financial markets, a risk of recession inthe United States, and slowing growth in theEuropean Union should yield a fairly pro-nounced slowing of external demand forSouth Asia’s exports during 2008. This islikely to worsen the region’s overall currentaccount deficit, as will increases in global oilprices into 2008. Nonetheless, contributionsto growth from domestic demand—both pri-vate consumption and investment—are ex-pected to remain relatively high, despite facingheadwinds from tighter monetary conditionsand further consolidation in fiscal positions.

These factors are expected to lead to aneasing of regional GDP growth from 8.4 per-cent in 2007 to 7.9 percent in 2008, with mostcountries experiencing a modest deceleration(tables A9 and A10). In Bangladesh, however,heightened political tensions and severe flood-ing were curbing demand in the second half of2007 and will contribute to a full percentagepoint reduction in growth to 5.5 percent for2008. Regional growth is expected to pick upto 8.1 percent by 2009, as recovering growthin the OECD firms up external demand and asreceding oil prices ease pressures on the im-port bill. Domestic absorption is projected toregain momentum in 2009, assuming thatprice conditions permit an easing of monetarypolicy during the second half of 2008.

Tighter credit conditions in internationalmarkets and a decreased appetite for riskamong investors could result in a falloff in re-gional capital inflows, which have contributedto recent strong growth outturns, particularlyin India and Pakistan. Although trade open-ness has generally increased (except inAfghanistan, Nepal, and Sri Lanka, which are


191


192

Table A10 South Asia country forecasts(annual percent change unless indicated otherwise)

Forecast

1991–2000a 2004 2005 2006 2007 2008 2009

BangladeshGDP at market pricesb 4.8 6.3 6.0 6.6 6.5 5.5 6.5Current account balance/GDP (%) �0.4 �0.5 �0.3 1.8 1.3 0.9 0.2

IndiaGDP at market pricesb 5.5 8.3 9.2 9.4 9.0 8.4 8.5Current account balance/GDP (%) �1.1 0.1 �1.0 �1.1 �2.2 �2.8 �2.5

NepalGDP at market pricesb 5.0 3.7 2.7 2.3 2.5 4.0 4.5Current account balance/GDP (%) �6.4 �0.7 0.0 �0.1 �1.7 �2.4 �2.9

PakistanGDP at market pricesb 3.9 6.4 7.7 6.9 6.4 6.5 6.7Current account balance/GDP (%) �3.7 �0.8 �3.3 �4.3 �4.9 �5.8 �5.3

Sri LankaGDP at market pricesb 5.2 5.4 6.0 7.4 6.3 6.2 6.5Current account balance/GDP (%) �4.6 �3.4 �3.1 �4.9 �4.8 �5.1 �4.5

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 averages; growth contributions, ratios, and the GDP deflator are averages.b. GDP measured in constant 2000 U.S. dollars.

Table A9 South Asia forecast summary(annual percent change unless indicated otherwise)

Forecast

1991–2000a 2004 2005 2006 2007 2008 2009

GDP at market pricesb 5.2 7.8 8.7 8.8 8.4 7.9 8.1GDP per capitac 3.2 6.1 7.0 7.2 6.9 6.4 6.6Purchasing power parity GDPd 6.4 7.9 8.7 8.9 8.5 8.0 8.2


Net exports, contribution to growth �0.1 �3.1 �1.3 �0.9 �0.9 �1.0 �0.9Current account balance/GDP (%) �1.5 �0.1 �1.2 �1.3 �2.4 �3.0 �2.7GDP deflator (median, LCU) 8.0 4.9 5.7 7.5 7.0 7.0 6.4Fiscal balance/GDP (%) �7.8 �6.4 �6.4 �6.3 �6.1 �5.5 �5.1

Memo items: GDPSouth Asia, excluding India 4.4 6.1 6.7 6.7 6.3 6.0 6.5India 5.5 8.3 9.2 9.4 9.0 8.4 8.5Pakistan 3.9 6.4 7.7 6.9 6.4 6.5 6.7Bangladesh 4.8 6.3 6.0 6.6 6.5 5.5 6.5

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. Measured in U.S. dollars.d. GDP measured at purchasing power parity exchange rates.e. Exports and imports of goods and nonfactor services.

(17 percent), Pakistan (11 percent), and SriLanka (12 percent). India’s food imports ac-count for just 3 percent of its total merchan-dise imports. Aside from putting increasedpressure on external positions, higher interna-tional food prices carry potentially serious im-plications for the poorest members of thesesocieties and could strain government coffersand generate increased inflationary pressuresgiven widespread food subsidies. Similarly,further increases in energy prices remain a riskfor the region, which is highly dependent onoil imports. In India and Pakistan, for exam-ple, fuel imports represent more than 30 per-cent and 20 percent of merchandise imports,respectively.

Internal political and military tensions con-tinue to represent downside risks to the projec-tions, as demonstrated, for example, by in-creased fighting in the civil war in Sri Lankafrom mid-2006 through 2007. Correspond-ingly, improving relations represent potentialupside opportunities, such as the ongoingprogress in reestablishing ties between India andPakistan with the opening on October 1, 2007,of the first overland truck route across the bor-der after several decades of closed borders.

Sub-Saharan Africa Recent developmentsEconomic growth in Sub-Saharan Africa ac-celerated from 5.7 percent in 2006 to 6.1 per-cent in 2007, the region’s fastest pace ofgrowth in more than three decades (figureA21). Robust 8.1 percent GDP growth amongthe region’s oil exporters and 5.3 percent gainsfor oil-importing countries, not includingSouth Africa, powered the growth. High oilprices and new oil production, notably inAngola and Sudan, helped propel growth inthe oil-exporting countries to the highest ratein almost 35 years. At the same time, theboom in non-oil commodity prices, in conjunc-tion with increased openness and improvedmacroeconomic stability, boosted growth inoil-importing economies to a 10-year high(table A11).


193

all suffering from civil strife), openness re-mains relatively low in most countries. As aconsequence, the region is likely to be bufferedto a degree from a falloff in external demandtied to the downturn in the credit cycle.

RisksRelatively large current account deficits inMaldives, Pakistan, and Sri Lanka remain ofconcern, particularly in an environment ofincreased volatility in international financialmarkets. In Maldives and Sri Lanka, fiscaldeficits exceed 7.3 percent of GDP, and for-eign reserve positions were fairly tight as ofmid-2007, equivalent to 2.5 months of importcover. Although import cover in Pakistan isstill a relatively comfortable four months, it ison a declining trend, suggesting that furtheradjustments are required. Accordingly, a sud-den and marked slowdown in capital inflowsor a discrete adjustment in global financialmarkets could have noticeable adverse eco-nomic effects on these countries.

European and U.S. restrictions on somecategories of Chinese textile and clothing ex-ports will be lifted at the end of 2008, and in-creased competition in 2009 could hurt re-gional exporters. Potential effects may bediscerned by examining developments inCanada, which has not imposed safeguard re-strictions on China: Bangladesh’s share ofCanada’s textile and clothing market declinedfrom 7.4 percent in 2005–06 to 6.9 percentover 2007 to date. Among regional exporters,Sri Lanka appears to be most at risk: nearlyone-third of its total merchandise exports tothe EU and nearly one-fifth of its shipments tothe United States are in categories in whichChinese trade will be liberalized (see box 1.1in chapter 1).

High, and in some cases increasing, com-modity prices also present a risk for the re-gion’s economies. Sharp gains in internationalfood prices are a growing threat in a regionwhere food imports represent 11–20 percentof total merchandise imports: Bangladesh(19 percent), Maldives (16 percent), Nepal

Output expansion in South Africa, the re-gion’s largest economy, appears slated to easeto 5.0 percent in 2007 from 5.4 percent in2006. Quarterly output readings point to

robust expansion, with third-quarter growthaccelerating to 4.7 percent (seasonally ad-justed annual rate) from 4.4 percent in the sec-ond quarter. There are indications, however,that higher borrowing costs have dampenedthe pace of consumption growth, while higherinterest rates and a stronger rand are under-mining growth in the manufacturing sector.Meanwhile gross fixed capital formation con-tinues to expand at a robust pace, and willlikely do so in coming quarters as infrastruc-ture spending is under way for the 2010World Soccer Cup. Softer private consump-tion growth and a modest improvement in theterms of trade reduced South Africa’s currentaccount deficit to 6.5 percent of GDP in thesecond quarter of 2007, from a 6.9 percentgap in the previous quarter. Despite softer pri-vate consumption, increased capital and con-struction materials imports and deteriorationin the terms of trade will cause the currentaccount balance to widen (table A12).


194

Table A11 Sub-Saharan Africa forecast summary(annual percent change unless indicated otherwise)

Forecast

1991–2000a 2004 2005 2006 2007 2008 2009

GDP at market pricesb 2.3 5.3 5.8 5.7 6.1 6.4 5.8GDP per capitac �0.4 2.9 3.4 3.7 4.0 4.4 3.8Purchasing Power Parity GDPd 3.4 5.5 6.1 6.1 6.5 6.7 6.1


Net exports, contribution to growth 0.2 �0.9 �1.1 �3.0 �1.2 �0.9 �1.4Current account balance/GDP (%) �2.1 1.0 �0.3 �0.2 �1.0 �0.9 �2.1GDP deflator (median, LCU) 10.1 7.5 6.0 6.5 5.8 4.8 4.5Fiscal balance/GDP (%) �4.2 �0.7 0.6 1.6 �0.2 0.6 0.1

Memo items: GDPSSA excluding South Africa 2.6 5.5 6.2 5.9 6.7 7.2 6.2Oil exporters 2.2 6.0 7.3 6.7 8.1 9.0 6.9CFA countries 2.6 4.2 4.0 2.8 3.7 4.5 4.1

South Africa 1.8 4.8 5.0 5.4 5.0 5.1 5.3Nigeria 2.8 6.0 6.6 5.6 5.9 7.4 6.1Kenya 1.9 4.9 5.8 6.1 6.3 5.3 5.1

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. Measured in U.S. dollars.d. GDP measured at purchasing power parity exchange rates.e. Exports and imports of goods and nonfactor services.


Figure A21 Growth in Sub-Saharan Africahas accelerated markedly…

�2

1970

1973

1976

1979

1982

1985

1988

1991

1994

1997

2000

2003

2006

2009

5

6

7

8

4

3

2

1

0

�1

Source: World Bank.


195

Table A12 Sub-Saharan Africa country forecasts(annual percent change unless indicated otherwise)

Forecast

1991–2000a 2004 2005 2006 2007 2008 2009

AngolaGDP at market pricesb 0.8 11.2 20.6 16.9 24.3 21.7 13.4Current account balance/GDP (%) �6.1 5.9 15.1 13.6 12.4 9.6 3.7

BeninGDP at market pricesb 4.8 3.1 2.9 3.6 4.4 5.1 5.3Current account balance/GDP (%) �6.8 �7.8 �6.3 �5.4 �5.7 �7.0 �7.4

BotswanaGDP at market pricesb 6.2 5.9 3.8 2.6 4.8 4.4 3.9Current account balance/GDP (%) 8.1 2.9 16.6 18.0 17.6 14.7 12.8

Burkina FasoGDP at market pricesb 4.0 3.9 7.1 6.1 5.3 5.5 5.9Current account balance/GDP (%) �5.6 �11.0 �10.5 �10.1 �10.2 �11.0 �11.3

BurundiGDP at market pricesb �1.7 4.8 0.9 5.4 4.2 5.3 5.1Current account balance/GDP (%) �3.4 �25.0 �17.9 �17.0 �15.5 �14.9 �13.7

Cape VerdeGDP at market pricesb 5.8 4.5 5.9 6.2 6.3 6.6 6.4Current account balance/GDP (%) �8.3 �16.3 �5.5 �12.0 �14.6 �15.8 �13.7

CameroonGDP at market pricesb 1.4 3.7 2.0 3.5 3.8 4.4 4.2Current account balance/GDP (%) �3.0 �2.6 �4.2 �0.7 �0.6 �0.9 �2.4

Central African RepublicGDP at market pricesb 1.6 1.3 2.2 3.3 3.5 3.8 4.1Current account balance/GDP (%) �4.3 �4.6 �4.9 �2.9 �2.9 �3.9 �4.6

ChadGDP at market pricesb 2.3 29.5 7.9 1.4 �1.4 4.2 2.1Current account balance/GDP (%) �5.5 �18.7 �6.6 �7.3 �4.8 �2.6 �1.8

ComorosGDP at market pricesb 1.1 �0.2 4.2 1.3 1.8 2.5 2.7Current account balance/GDP (%) �6.8 �3.4 �4.9 �5.9 �5.1 �5.2 �5.5

Congo, Dem. Rep. ofGDP at market pricesb �5.6 6.6 6.5 5.1 6.1 7.3 6.8Current account balance/GDP (%) 2.0 �8.8 �10.0 �8.9 �9.3 �9.5 �9.9

Congo, Rep. ofGDP at market pricesb 1.5 3.6 7.8 6.6 3.2 6.3 5.9Current account balance/GDP (%) �16.5 15.5 16.2 11.0 6.1 11.0 8.0

Côte d’IvoireGDP at market pricesb 2.3 1.8 1.8 0.9 1.7 2.8 3.1Current account balance/GDP (%) �4.0 1.6 0.4 3.3 2.5 1.0 �0.5

Equatorial GuineaGDP at market pricesb 18.4 10.0 6.5 �5.2 8.8 9.0 �1.4Current account balance/GDP (%) �40.6 7.3 6.8 6.3 4.3 7.7 2.8

EritreaGDP at market pricesb — 1.9 4.8 1.7 1.9 2.0 2.2Current account balance/GDP (%) — �13.6 �28.0 �31.6 �30.9 �30.1 �28.5

EthiopiaGDP at market pricesb 2.3 12.3 10.5 9.6 9.3 7.7 7.4Current account balance/GDP (%) �0.8 �6.9 �7.7 �9.6 �7.3 �7.4 �5.6

GabonGDP at market pricesb 2.4 1.4 3.0 1.2 4.8 3.9 3.7Current account balance/GDP (%) 5.7 12.8 13.7 21.4 17.3 21.4 18.6

(continued)


196

Gambia, TheGDP at market pricesb 3.3 5.1 6.9 6.4 6.1 5.3 5.8Current account balance/GDP (%) �1.6 �11.1 �21.6 �21.5 �20.4 �21.2 �18.3

GhanaGDP at market pricesb 4.3 5.6 5.9 6.2 4.9 5.4 6.1Current account balance/GDP (%) �6.4 �3.6 �6.5 �8.7 �10.3 �10.2 �8.4

GuineaGDP at market pricesb 3.9 2.7 3.3 2.4 2.1 3.7 3.2Current account balance/GDP (%) �5.7 �4.3 �5.2 �6.1 �7.3 �10.7 �11.5

Guinea-BissauGDP at market pricesb 1.5 2.2 3.2 2.7 2.7 2.6 3.3Current account balance/GDP (%) �24.0 �4.9 �7.5 �11.3 �10.1 �4.3 �3.3

KenyaGDP at market pricesb 1.9 4.9 5.8 6.1 6.3 5.3 5.1Current account balance/GDP (%) �1.6 �2.2 �0.8 �2.1 �3.2 �4.1 �3.6

LesothoGDP at market pricesb 3.4 3.1 2.9 6.9 4.1 4.8 4.4Current account balance/GDP (%) �13.3 �5.6 �7.4 3.0 0.1 �0.9 2.4

MadagascarGDP at market pricesb 1.7 5.3 4.6 4.9 5.8 6.3 6.9Current account balance/GDP (%) �7.8 �12.4 �11.9 �8.9 �14.7 �21.7 �21.4

MalawiGDP at market pricesb 3.4 7.1 2.7 7.4 6.3 6.1 5.6Current account balance/GDP (%) �8.5 �4.7 �9.5 �6.2 �4.0 �5.1 �4.1

MaliGDP at market pricesb 4.0 2.2 6.1 4.9 5.1 5.2 5.3Current account balance/GDP (%) �8.9 �8.4 �8.3 �4.9 �4.6 �4.9 �5.2

MauritaniaGDP at market pricesb 2.9 5.2 5.4 11.7 2.1 5.7 6.7Current account balance/GDP (%) �0.3 �20.1 �46.3 �2.1 �4.0 �11.6 �19.3

MauritiusGDP at market pricesb 5.3 4.7 3.1 3.5 4.1 4.4 3.9Current account balance/GDP (%) �1.6 �1.8 �3.7 �5.6 �8.4 �8.9 �7.6

MozambiqueGDP at market pricesb 5.2 7.5 6.2 8.5 9.1 7.4 6.7Current account balance/GDP (%) �18.2 �11.1 �12.6 �7.3 �9.0 �10.2 �9.2

NamibiaGDP at market pricesb 4.2 6.0 4.2 4.6 4.6 4.3 3.8Current account balance/GDP (%) 4.1 10.0 7.2 13.0 15.0 11.3 10.6

NigerGDP at market pricesb 1.8 0.0 7.0 4.2 4.8 4.5 4.6Current account balance/GDP (%) �6.9 �7.6 �9.4 �9.5 �10.3 �11.0 �10.8

NigeriaGDP at market pricesb 2.8 6.0 6.6 5.6 5.9 7.4 6.1Current account balance/GDP (%) �0.8 23.4 10.9 13.7 9.2 12.6 9.5

RwandaGDP at market pricesb 0.2 4.0 6.1 6.4 6.7 5.1 4.7Current account balance/GDP (%) �3.5 �10.8 �3.9 �7.1 �8.3 �8.2 �7.4

SenegalGDP at market pricesb 2.9 6.2 5.5 3.1 4.5 5.1 5.3Current account balance/GDP (%) �6.0 �6.7 �7.0 �9.6 �9.7 �8.4 �7.6


Forecast

1991–2000a 2004 2005 2006 2007 2008 2009


197

SeychellesGDP at market pricesb 4.6 �2.0 1.2 5.3 5.5 5.1 5.7Current account balance/GDP (%) �7.4 �9.1 �23.8 �28.9 �30.4 �28.8 �20.4

Sierra LeoneGDP at market pricesb �4.7 7.4 7.5 7.5 7.3 6.9 7.4Current account balance/GDP (%) �9.0 �13.1 �9.3 �6.0 �6.5 �8.3 �9.0

South AfricaGDP at market pricesb 1.8 4.8 5.0 5.4 5.0 5.1 5.3Current account balance/GDP (%) �0.2 �3.2 �3.8 �6.3 �6.8 �7.8 �7.7

SudanGDP at market pricesb 5.7 5.2 8.6 11.8 10.8 9.7 8.1Current account balance/GDP (%) �6.7 �4.1 �10.9 �14.2 �9.6 �6.5 �6.6

SwazilandGDP at market pricesb 3.1 2.1 2.3 1.7 1.4 1.1 1.2Current account balance/GDP (%) �2.6 4.6 1.5 1.0 0.1 �3.2 �2.4

TanzaniaGDP at market pricesb 2.9 6.7 6.8 6.2 7.1 6.8 6.7Current account balance/GDP (%) �12.5 �2.2 �4.8 �8.0 �9.4 �11.0 �10.2

TogoGDP at market pricesb 2.2 3.0 2.8 1.5 2.3 2.1 2.4Current account balance/GDP (%) �8.5 �10.0 �5.1 �5.7 �6.1 �9.3 �7.8

UgandaGDP at market pricesb 6.8 5.5 6.6 5.4 5.5 5.3 5.9Current account balance/GDP (%) �7.0 �3.2 �2.2 �4.8 �3.0 �5.4 �4.9

ZambiaGDP at market pricesb 0.7 5.4 5.1 5.9 5.3 5.7 5.9Current account balance/GDP (%) �10.5 �10.0 �8.6 1.6 �0.7 �4.6 �6.2

ZimbabweGDP at market pricesb 0.9 �3.8 �5.3 �4.2 �6.3 �4.9 �2.1Current account balance/GDP (%) �7.5 20.7 25.7 45.0 71.9 81.7 82.7

Source: World Bank.Notes: Growth and current account figures presented here are World Bank projections and may differ from targets contained inother World Bank documents. Liberia, Mayotte, S~ao Tomé and Principe, and Somalia are not forecast 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.


Forecast

1991–2000a 2004 2005 2006 2007 2008 2009

Recent turbulence in international finan-cial markets resulted in a moderate deprecia-tion of the rand against the dollar, but thatfollowed a period of strong appreciationcaused by, among other factors, anticipatedcapital inflows related to merger and acquisi-tion activity. The rand has since returned tolevels prevailing before the period of intensefinancial market disruptions of July. Volatilityand declines in high-income financial marketsaffected capital markets in South Africa tosome degree, but the change was limited. The

country avoided sharp asset sell-offs, and theall-share Johannesburg Stock Exchange indexreached new highs. Looking forward, limitedspillovers to consumption and investment areanticipated.

Exceptional outturns in countries like An-gola and Sudan underpinned growth amongSub-Saharan Africa’s oil-exporting economies(figure A22). In Angola, the region’s fastest-growing economy and its second largest oilproducer, growth continues to be exception-ally robust, with both oil and non-oil sectors


198

advancing at double-digit rates. Diamond pro-duction has also been climbing, supplement-ing oil revenues and contributing to an in-crease in the current account surplus.Nevertheless, the increase in oil productionwas an impressive 18.7 percent in the firstnine months of 2007 (year-on-year) after ex-panding 26 percent in 2005 and 13 percent in2006. The new oil fields scheduled to comeon line over 2007 to 2009 will ensure stronggrowth in the oil sector, while robust growthin agriculture, manufacturing, construction,and the power sector will boost activity in thenon-oil sectors to near 20 percent.

Also notable among the oil-exportingeconomies is an improvement in macroeco-nomic stability, with inflation easing, notwith-standing large revenue inflows and strongergrowth in government spending. Furthermore,recent monetary tightening should help bringinflationary pressures in Angola down further.Both fiscal and current account balances haveimproved tremendously as a result of the oilwindfall. The non-oil fiscal balance also hasimproved, but largely as a result of capacityconstraints that limit government spendingincreases.

Meanwhile, performance in Nigeria, the re-gion’s second largest economy, has improvedmarginally over the course of 2007. GDP

growth rose modestly from 5.6 percent in2006 to 5.9 percent, as strong gains in thenon-oil sector offset substantial underperfor-mance in oil. A 9.5 percent expansion in thenon-oil sector during the third quarter of 2007(year-on-year), driven by strong performancein agriculture and financial services, pushedoverall growth up to 6.1 percent from 5.7 per-cent in the previous quarter. Crude oil pro-duction, including condensates and naturalgas liquids, declined 7.1 percent during thethird quarter (year-on-year), bringing outputdown 4.2 percent for the first nine months of2007. This unfavorable result followed a con-traction in production of 5.3 percent in 2006.Economic growth in Mauritania, the region’snewest oil producer, has also been disappoint-ing. Growth there was subdued largely becauseof a halving in oil output tied to a changeoverin equipment.

Members of the Central African Economicand Monetary Community experienced mixedeconomic performance. For example, stronggrowth in the non-oil sector in Gabon led to aconsiderable pickup in GDP despite a declinein oil production. Growth also accelerated inCameroon, underpinned by strong domesticdemand, and on the supply side, by strongperformance in the transport and telecommu-nications sectors. Meanwhile in the Republicof Congo, growth performance fell off as ex-pansion in the oil sector softened, and delaysin public investment affected growth in thenon-oil sector. In Chad, another sharp declinein oil production and continued decelerationin growth within the non-oil sector caused theeconomy to contract during the year.

Among the region’s oil importers, GDP ad-vances were robust, especially in those coun-tries that are further along the reform path,have achieved greater trade openness, andhave improved macroeconomic stability (fig-ure A23). Increases in nontraditional manu-factured exports have helped propel growth inseveral countries. In East Africa, Tanzaniarecorded stronger growth in the agriculturesector and improved its performance in manu-facturing, mining, and construction, pushing


Figure A22 …reaching a 35-year high in oil-exporting countries…

�6

1970

1973

1976

1979

1982

1985

1988

1991

1994

1997

2000

2003

2006

2009

8

10

12

14

6

4

2

0

�2

�4

Source: World Bank.

Forecast


199

overall GDP gains to 7.1 percent in 2007. InKenya, at 6.3 percent, growth should surpassa 30-year record set in 2006 (6.1 percent)based on improved tourism arrivals and re-ceipts, stronger gains in transport and com-munications, and better agriculture output.Growth accelerated to 7.1 percent in the sec-ond quarter of 2007 (year-on-year) as manu-facturing expanded 8.6 percent, hotels andrestaurants gained 11.1 percent, and commu-nications expanded 11.8 percent. In Madagas-car, large inflows of FDI to the mining sectorare generating an important shift in the struc-ture of the economy, while Uganda is benefit-ing from strong expansion in construction,transport, and communications, as well asfrom recovery in agriculture.

Growth among members of the WestAfrican Economic and Monetary Union alsopicked up in 2007, rising by almost a full per-centage point to 3.7 percent, as GDP gains ex-ceeded 4 percent in five of the eighteconomies. Yet inadequate energy supplies inmany member countries and difficulties inseveral agriculture subsectors have keptgrowth from rising even further. Agriculturaloutput and private consumption will takesome time to recover in several countries inthe union that sustained significant damage

from flooding, slowing growth toward theend of 2007 into 2008.

In Mozambique, growth continues apace,with output up 8.8 percent in the first half ofthe year, but damage from cyclones and flood-ing slowed the rate of growth from 10 percentrecorded a year earlier (year-on-year). Thecountry boosted its GDP with investments of$1.4 billion in the Moatize Coal Project and$390 million in other projects during the firsthalf of 2007.

Medium-term outlookGrowth in Sub-Saharan Africa is poised to re-main buoyant by historic standards, nearing6 percent over the near and medium terms, de-spite softening of demand in the Euro Area in2008 and corrections to non-oil commodityprices, particularly metals, over the mediumterm (table A11). For many countries, the im-petus for growth will come in part from ro-bust domestic demand, in particular, stronggrowth in public and private fixed investment(figure A24).

Fixed investment is expected to remainstrong, despite the credit crunch in interna-tional financial markets and the easing in


Figure A23 ...and a 10-year high inoil-importing countries

�2

1970

1973

1976

1979

1982

1985

1988

1991

1994

1997

2000

2003

2006

2009

8

10

6

4

2

0

Source: World Bank.

Forecast

Contributions to GDP growth (percent)

Figure A24 Contributions of investment and consumption have increased

�4

�2

2000

2001

2002

2003

2004

2005

2006

2007

2008

2009

8

10

6

4

2

0

Source: World Bank.

Government consumptionNet exports

Private consumption

Fixed investment


200

non-oil commodity prices. Large, strategic,capital projects initiated by big, rapidly–grow-ing, developing economies such as China andIndia are likely to underpin continued stronginvestment. Among recent notable deals is the$5 billion loan agreement reached betweenChina and the Democratic Republic ofCongo. According to the signed accord, about$3 billion will fund large infrastructure pro-jects, including 2,000 miles of railway and2,125 miles of highway. The project will alsofinance the construction of 31 hospitals, 145health centers, and 2 universities. The remain-ing $2 billion will be used to set up a jointventure mining company between the twocountries. Madagascar and Mozambique arealso experiencing large-scale FDI flows intotheir economies.

Several years of strong non-oil commodityprices, together with rising private capital in-flows and remittances, are supporting privateconsumption, although spending is antici-pated to flatten over the medium term as in-come growth eases. In South Africa, higher in-terest rates and an erosion of real incomes islikely to curb real spending. Furthermore, thesharp decline in farmers’ incomes in thosecountries affected by recent floods will consti-tute a drag on consumption, although private,government, and donor transfers may mitigatesome of the effects. According to the latestestimates, the worst-affected countries areBurkina Faso, Ghana, and Uganda. Adversespillovers to personal spending will be felt wellinto 2008. In other countries, recovery fromdrought is translating into improved perfor-mance in agriculture, adding impetus togrowth, while the income effect resulting fromseveral years of high prices for non-oilcommodities continues to stimulate privateconsumption.

Growth among Africa’s oil exporters is ex-pected to slow to 6.9 percent by 2009, as oilprices are forecast to decline that year and asinvestment projects begin to unwind. How-ever, this transition occurs from peak 9 per-cent GDP gains in 2008 led by continued

advances in Angolan and Sudanese crude oiloutput and anticipated recovery in Nigerianproduction. Lower oil prices by 2009 couldyield supply responses, at minimum slowingthe pace of oil output growth for manysmaller producers, where production costs re-main high. For the region’s major producers,however, output is projected to continue toincrease.

Growth in Nigeria is projected to pick upmarkedly in 2008, buoyed by recovery in theoil sector and strong gains in the non-oil sec-tor. Stabilization in the Niger Delta shouldallow oil production to recover gradually andto exceed 2005 levels, supporting gains near10 percent in Nigeria’s oil sector, following thecontraction recorded in 2007. Investmentgrowth should moderate, notwithstanding thegovernment’s plans to invest substantialamounts in roads, railways, and electricity in-frastructure and in the dredging of the NigerRiver. This investment appears highly import-intensive, with import sourcing accounting forsome 75–80 percent of expenditure, therebylimiting the direct positive effects on Nigeria’sgrowth.

Difficulties in the cotton and groundnutsubsectors in several West African countriesare having negative spillover effects ongrowth, while electricity shortages and highenergy costs are constraining faster expansionamong oil-importing economies. Some coun-tries are seeing surges in their oil import bills,in part because many companies use fuel-runpower generators to supplement grid electric-ity. Growth among West African Economicand Monetary Union economies will be sup-ported by improved performance in the pri-mary sector and increased public investmentspending, financed from savings made underthe Highly Indebted Poor Countries and Mul-tilateral Debt Relief initiatives. Furthermore,growth in oil-importing economies shouldbenefit from strong investment growth, aswell as continued demand for non-oil com-modities from the rapidly expanding emergingmarket economies.

RisksAmong the most important downside risks togrowth in Sub-Sahara African is a larger-than-expected falloff in high-income countrygrowth and import demand. Such a falloffcould be tied to continuing difficulties in fi-nancial markets, or it could be triggered byweaker growth outturns in major emergingmarket economies, such as in China and India,which might result in weaker export growthfor Sub-Saharan Africa in general, and for theoil-importing countries in particular.

The risk of increasing inflationary pressureis also palpable. Should oil prices increase fur-ther or drought conditions affect the food sup-ply, inflation could be the consequence. In-creases in international prices of staplecommodities such as cereals and vegetableoils, to a large degree already in the pipeline,could spill over to fuel domestic price pres-sures, particularly in countries that are heavilyreliant on imports of wheat and vegetable oils.

In some countries, strong domestic demandand upward adjustments in public sectorwages are already fueling inflationary pres-sures. Capacity constraints and overheating insome economies, especially the oil-producingeconomies, constitute further upside risks forinflation. Similarly, in South Africa, high con-fidence, strong domestic expenditures, risingemployment, buoyant credit extension, and

increasing asset prices all boost consumptionthrough positive wealth effects and make therisk of a further buildup in inflationary pres-sures very apparent.

South Africa also faces a specific source ofrisk in the way it has financed its sizable cur-rent account deficit, anticipated to be in excessof 7 percent of GDP in 2008 and 2009: thecountry has relied largely on portfolio invest-ments, which could potentially be reversed.Moreover, a marked depreciation of the randin the event of a sell-off of South African assetsby nonresidents would present the authoritieswith additional financial management issues toaddress as well as inflation pressures.

Finally, negative terms-of-trade shocks(lower non-oil commodity prices, furtherincreases in oil prices, or both) may come to af-fect disposable incomes in oil-importing coun-tries, undermining private consumption whilealso yielding worse external balances. Con-versely, higher oil prices would benefit growthin oil-exporting countries. Sociopolitical ten-sions also remain a source of downside risk forcountries such as Côte d’Ivoire, the Demo-cratic Republic of Congo, Ethiopia, Guinea,Guinea-Bissau, Somalia, and Togo. In Nigeria,the risks associated with the activity of militantgroups in the Niger Delta remain substantial,with oil production still about 25 percentbelow the 2.9 million barrels daily capacity.


201

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 2008 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:• 21 trees• 1247 lbs. of solid

waste• 7,539 gallons of

wastewater• 2,300 lbs. of net

greenhouse gases• 14 million BTUs

of total energy

Global Economic Prospects 2008: Technology Diffusion in the Developing

World examines the state of technology in developing countries and

the pace with which it has advanced since the early 1990s.

It reveals both encouraging and cautionary trends. On the one

hand, the pace of technological progress in developing countries

has been much faster than in high-income countries—reflecting

increased exposure to foreign technology as a result of linkages

with high-skilled diasporas and the opening of these countries

to international trade and foreign direct investment. On the other

hand, the technology gap remains large, and the domestic factors that determine

how quickly technologies spread within developing countries often stymie progress, especially among

low-income countries.

This year’s Global Economic Prospects comes on the heels of an extended period of strong growth and a 15

year period of strong performance in much of the developing world, which has contributed to substantial

declines in global poverty. While high oil prices and heightened market volatility may signal a coming pause

in this process, over the longer term continued technological progress should continue to push back poverty.

WORLD BANKPublications

Visit our Web site at

www.worldbank.org/publications

See other side for order form.

New! Prospects for the GlobalEconomy onlineYour multilingual source (English, French, Spanish) for• Macroeconomic indicators and forecast data, from 1980 to 2008• Insightful interactive calculators and simulation tools• One-page PDF briefs summarizing countries’ external financial

positions and trade• 16 individual commodity reports and price forecasts• Timely analysis of worldwide economic prospects and risks

Please visit the new interactive Prospects for the Global Economy Web site at www.worldbank.org/globaloutlook.

The reference of choice on development

The World BankGLOBALECONOMIC PROSPECTS

Global Economic Prospects 2008:Technology Diffusion in the Developing World

Name _________________________________________________

Title ___________________________________________________

Organization ____________________________________________

Address________________________________________________

City ___________________________________________________

State __________________________________________________

Zip/Postal Code__________________________________________

Country ________________________________________________

Phone _________________________________________________

Fax ___________________________________________________

E-mail _________________________________________________

■■ Charge $__________________ to my:

■■ American Express ■■ Mastercard ■■ Visa

Card no. _______________________________________________

Expiration date _____________ /____________

Name _________________________________________________as it appears on the card

Signature _______________________________________________required for all credit card charges

■■ Check no. _______________

in the amount of $__________________ is enclosed. When ordering

directly from the World Bank, make check payable in U.S. funds drawn on

a U.S. bank to: The World Bank. Please send your check with your order.

Institutional customers in the U.S. only:

■■ Bill me. Please include purchase order.

To Order:

World Bank Publications


By phone: +1-703-661-1580 or 800-645-7247

By fax: +1-703-661-1501

By mail: P.O. Box 960, Herndon, VA 20172-0960, USA

Questions? E-mail us at [email protected]

Order Form

DCGP8

PAYMENT METHOD

Orders from individuals must be accompanied by payment or credit card information. Credit cards are accepted only for orders addressed

to the World Bank. Check with your local distributor about acceptable credit cards. Please do not send cash.

WORLD BANKPublications

Visit our Web site at


* Geographic discounts apply - up to 75% for some countries. Please see our Web site for information

(http://publications.worldbank.org/discounts).

** Shipping and handling charges within the US are US$8.00 per order. If a purchase order is used, actual shipping will be

charged. Outside of the US, customers have the option to choose between nontrackable airmail delivery (US$7.00 per

order plus US$6.00 per item) and trackable courier airmail delivery (US$20.00 per order plus US$8.00 per item).

Title Stock # Price* Qty. Total US$

Global Economic Prospects 2008: Technology Diffusion in the Developing World D17365 US$38

January 2008. (ISBN 978-0-8213-7365-1).

Shipping and Handling**

Total US$

The reference of choice on development

(Please Print)

THE WORLD BANK

ISBN 978-0-8213-7365-1

Technology and technological progress are central to economic and social well-being. The creation and diffusion of goods and services are critical drivers of economic growth, rising incomes,

social progress, and medical progress. Global Economic Prospects 2008: Technology Diffusion in the Developing World examines the state of technology in developing countries and the pace with which it has advanced since the early 1990s.

It reveals both encouraging and cautionary trends. On the one hand, the pace of technological progress in developing countries has been much faster than in high-income countries—reflecting increased exposure to foreign technology as a result of linkages with high-skilled diasporas and the opening of these countries to international trade and foreign direct investment.

On the other hand, the technology gap remains large, and the domestic factors that determine how quickly technologies spread within developing countries often stymie progress, especially among low-income countries. Repeating the rapid progress of this past decade will be difficult and may require that basic technological literacy is improved further; that government efforts to adapt and disseminate preexisting technologies throughout the economy are strengthened; and that regulatory regimes are modified to encourage business innovation.

This year’s Global Economic Prospects comes on the heels of an extended period of robust growth and a 15-year period of strong performance in much of the developing world that has contributed to substantial declines in global poverty. While high oil prices and heightened market volatility may signal a coming pause in this process, over the longer term sustained technological progress should continue to push back poverty.

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.

“Rapid technological progress in developing countries has

been key to the reduction of poverty in recent decades.

While the integration of global markets has played and

will continue to play a vital role in this, future success

will increasingly depend on strengthening technical

competencies and the business environment for innovative

firms in developing countries.”

— Graeme Wheeler Managing Director

The World Bank

Date post:	18-Mar-2016
Category:	Documents
Upload:	world-bank-russia
View:	230 times
Download:	5 times

Global Economic Prospects 2008

Documents