Global Monitoring Report 2011 61311 Improving the Odds of Achieving the MDGs Global Monitoring Report 2011 Improving the Odds of Achieving the MDGs Global Monitoring Report 2011 Improving the Odds of Achieving the MDGs Heterogeneity, Gaps, and Challenges © 2011 The International Bank for Reconstruction and Development / The World Bank 1818 H Street NW Washington DC 20433 Telephone: 202-473-1000 Internet: www.worldbank.org All rights reserved 1 2 3 4 14 13 12 11 This volume is a product of the staffs of The World Bank and The International Monetary Fund. The find- ings, interpretations, and conclusions expressed in this volume do not necessarily reflect the views of the Board of Executive Directors of The World Bank, the Board of Executive Directors of The International Monetary Fund, or the governments they represent. The World Bank and The International Monetary Fund do not guarantee the accuracy of the data included in this work. The boundaries, colors, denominations, and other information shown on any map in this work do not imply any judgement on the part of The World Bank or The International Monetary Fund concerning the legal status of any territory or the endorsement or acceptance of such boundaries. Rights and Permissions The material in this publication is copyrighted. Copying and/or transmitting portions or all of this work without permission may be a violation of applicable law. The International Bank for Reconstruction and Development / The World Bank encourages dissemination of its work and will normally grant permission to reproduce portions of the work promptly. For permission to photocopy or reprint any part of this work, please send a request with complete information to the Copyright Clearance Center Inc., 222 Rosewood Drive, Danvers, MA 01923, USA; telephone: 978-750-8400; fax: 978-750-4470; Internet: www.copyright.com. All other queries on rights and licenses, including subsidiary rights, should be addressed to the Office of the Publisher, The World Bank, 1818 H Street NW, Washington, DC 20433, USA; fax: 202-522-2422; e-mail: pubrights@worldbank.org. ISBN: 978-0-8213-8700-9 eISBN: 978-0-8213-8701-6 DOI: 10.1596/978-0-8213-8700-9 The painting on the cover, “Faces of Diversity,” is by Tola Wewe; the painting is from his ‘Beauty is Everywhere’ series and is part of a private collection. Tola Wewe is a member of the African Artists’ Foundation (AAF), which is a nonprofit organization dedicated to the promotion of African art and artists, and the promotion of relevant social issues through artistic endeavors. AAF plays a significant role in art communities in Africa through its art exhibitions, competitions, art classes, and workshops with the aim of unearthing and developing talent, creating societal awareness, and providing a platform to express creativity. Visit www.africanartists.org for more information on AAF. Cover design by Debra Naylor of Naylor Design Interior photographs: All photographs are from the National Geographic Society Image Collection. Lynn Johnson (page 10); James P. Blair (page 42); Lynn Johnson (page 70); W. E. Garrett (page 102); Randy Olson (page 124). Contents Foreword . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xi Acknowledgments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xiii Acronyms and Abbreviations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xv Goals and Targets from the Millennium Declaration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xvi Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 1 The Diversity of MDG Progress . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11 2 Economic Growth in Developing Countries . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 43 3 Linking Spending and Outcomes: Some Lessons from Impact Evaluations in Education and Health . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 71 4 Assisting the Indigenous and Socially Excluded . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 103 5 Progress in the International Development and Trading Framework. . . . . . . . . . . . . . 125 Appendix 1: Technical Issues in Estimating Progress on the MDGs . . . . . . . . . . . . . . . . . 159 Appendix 2: Policies to Reduce Biodiversity Loss and Enhance Human Development . . . 175 GLOBAL MONITORING REPORT 2011 v vi CONTENTS GLOBAL MONITORING REPORT 2011 BOXES 1.1 Poverty data and projections. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13 1.2 Gaps and issues in measuring development outcomes beyond 2015 . . . . . . . . . . . . 18 1.3 Improving children’s health through sustainable access to food, water, and energy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23 1.4 The impact of violence on the MDGs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30 1.5 Reducing inequality and poverty in Brazil . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 33 2.1 The global food price surge in 2010—causes and prospects . . . . . . . . . . . . . . . . . . 46 2.2 Low-income-country exports during the exit from the crisis: early evidence . . . . . . 50 2.3 Emerging market capital inflows: experience and challenges . . . . . . . . . . . . . . . . . . 58 2.4 Fragile states—experience and implications . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 65 3.1 Some impact evaluations have forced a rethinking of our assumptions about HIV/AIDS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 76 3.2 Spillovers and unintended effects can be important . . . . . . . . . . . . . . . . . . . . . . . . . 77 3.3 The DIME initiative . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 78 4.1 Who are indigenous peoples? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 104 4.2 The Roma of Europe . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 110 4.3 Vulnerable groups deserve more attention in MDG country reports . . . . . . . . . . . 112 4.4 Bilingual education in Guatemala . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 114 4.5 Examples of successful targeted programs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 116 5.1 Aid recipients . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 127 5.2 The donor community is enhancing Africa’s trade facilitation . . . . . . . . . . . . . . . . 140 5.3 Special reconstruction allocation for Haiti . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 146 5.4 World Development Report 2011: Conflict, Security, and Development . . . . . . . . 147 5.5 Open data, open knowledge, and open solutions . . . . . . . . . . . . . . . . . . . . . . . . . 152 5.6 Climate change, natural disasters, and the World Bank Group . . . . . . . . . . . . . . . 153 FIGURES 1 The distance to global MDGs ranges widely . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2 2 More than two-thirds of developing countries are on track or close to being on track . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3 1.1 Current global distance to the MDGs is wide ranging . . . . . . . . . . . . . . . . . . . . . . . 12 1.2 Fewer low-income countries are on track to achieve the MDGs . . . . . . . . . . . . . . . 12 1.3 Countries on target to achieve the MDGs, by region . . . . . . . . . . . . . . . . . . . . . . . . 15 1.4 More than two-thirds of developing countries are on track or close to being on track . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22 1.5 MDG performance is stronger in countries with good initial conditions . . . . . . . . . 26 1.6 Subsequent economic growth and policy seem to matter more . . . . . . . . . . . . . . . . 27 1.7 MDG performance lags in IDA and fragile countries . . . . . . . . . . . . . . . . . . . . . . . 29 GLOBAL MONITORING REPORT 2011 CONTENTS vii 1.8 MDG performance is better in countries with greater export sophistication and shipping connectivity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31 1.9 The odds of getting on target rise substantially with faster growth and better policy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 35 1.10 Growth and policy reforms will put many countries on track . . . . . . . . . . . . . . . . . 36 2.1 Low-income economies’ per capita growth remained positive in 2009, in contrast to elsewhere . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 44 2.2 Economic recovery has taken many commodity prices sharply higher . . . . . . . . . . . 45 2.3 Bond spreads and international bond issuance are virtually back to precrisis levels in emerging and developing countries . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 47 2.4 Share prices, too, have bounced back . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 47 2.5 Emerging market bank financing continues to recover . . . . . . . . . . . . . . . . . . . . . . 47 2.6 Nonperforming loans in emerging and developing countries, 2003–10 . . . . . . . . . . 48 2.7 Global current account imbalances, 2000–11 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 48 2.8 Annual changes in terms of trade, by quintile group . . . . . . . . . . . . . . . . . . . . . . . . 49 2.9 Reserve levels broadly kept pace with the rebound in trade. . . . . . . . . . . . . . . . . . . 49 2.10 Low-income countries’ precrisis macroeconomic policy buffers were stronger than in earlier years . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 50 2.11 Macroeconomic policy responses to the crisis vary . . . . . . . . . . . . . . . . . . . . . . . . . 53 2.12 Monetary policy loosening . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 54 2.13 Average year-on-year growth in money and the money gap in emerging market economies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 54 2.14 Fiscal deficit, 2008–11 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 55 2.15 Low-income countries: fiscal balances and economic crises . . . . . . . . . . . . . . . . . . . 55 2.16 Real primary expenditure in low-income countries has increased in favor of public investment and social sectors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 55 2.17 Macroeconomic policy quality in low-income countries remained little changed. . . 56 2.18 The overall debt outlook in low-income countries seems favorable as fiscal buffers are being rebuilt along with the recovery. . . . . . . . . . . . . . . . . . . . . . . . . . . 63 2.19 Low-income countries projected to cut real spending tend to have high deficits and debt . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 63 2.20 The current account balance is projected to improve gradually in low-income countries as exports rebound . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 64 2.21 Reserve cushions would not improve much among low-income countries with relatively low reserves . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 64 3.1 Government and donor spending on health and education is unprecedented. . . . . . 72 3.2 The relationship between expenditure and learning outcomes is weak . . . . . . . . . . 73 3.3 There are wide differences in the share of public health spending going to the rich and the poor . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 73 3.4 Links between public expenditure and human development outcomes . . . . . . . . . . 74 4.1 Progress is mixed in reducing the poverty of indigenous groups . . . . . . . . . . . . . . 106 4.2 Test results show how stigma can be internalized . . . . . . . . . . . . . . . . . . . . . . . . . 108 viii CONTENTS GLOBAL MONITORING REPORT 2011 4.3 Heterogeneity has a negative effect on test scores . . . . . . . . . . . . . . . . . . . . . . . . . 108 4.4 Both poor and rich countries show gaps in life expectancy . . . . . . . . . . . . . . . . . . 110 5.1 Impact of banking crises on net disbursed aid provided by crisis-affected donors, 1977–2007 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 128 5.2 Exports have yet to recover their precrisis trend . . . . . . . . . . . . . . . . . . . . . . . . . . 132 5.3 Merchandise export volume . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 133 5.4 Trade-restrictive measures have fallen since 2008 . . . . . . . . . . . . . . . . . . . . . . . . . 134 5.5 Newly initiated trade remedy investigations have peaked . . . . . . . . . . . . . . . . . . . 134 5.6 Trade-restrictive and trade-liberalizing measures, November 2009– October 2010 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 135 5.7 Local-to-global connectivity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 142 5.8 Regional trends in aid for trade, 2002–09 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 144 A1.1 Odds of achieving the MDGs improve with growth and better policy . . . . . . . . . . 171 MAPS 1.1 The world is still on track to meet the poverty reduction target . . . . . . . . . . . . . . . 25 1.2 Sub-Saharan Africa and Southern Asia are home to the vast majority of children out of school . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 37 2.1 GDP: Economic growth in 2009 and 2010 has been uneven . . . . . . . . . . . . . . . . . . 60 2.2 Fiscal balance: Countries are diverging in their responses to fiscal challenges . . . . . 61 3.1 The maternal mortality rate is declining only slowly, even though the vast majority of deaths are avoidable . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 79 3.2 Each year of a girl’s education reduces, by as much as 10 percent, the risk of her children dying before age five . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 85 4.1 Almost nine million children still die each year before they reach their fifth birthday . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 109 4.2 For girls in some African countries, secondary education remains elusive . . . . . . 115 5.1 Across the world, 884 million people lack access to safe water—around 80 percent of them in rural areas . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 137 5.2 With half the people in developing regions without sanitation, the 2015 target appears to be out of reach . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 141 A2.1 Average change in species populations, relative to the intact situation (mean species abundance), 2000 and 2050. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 176 TABLES 1 Lagging countries are close to getting on target. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4 2 Growth and CPIA scores are higher in countries on track or close to being on track . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5 3 Global output, 2007–14 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6 1.1 Several low-income countries are achieving the MDGs . . . . . . . . . . . . . . . . . . . . . . 16 1.2 Lagging countries are close to getting on target. . . . . . . . . . . . . . . . . . . . . . . . . . . . 19 GLOBAL MONITORING REPORT 2011 CONTENTS ix 1.3 Many countries are within 10–20 percent of being on target. . . . . . . . . . . . . . . . . . 20 1.4 Lagging countries within 10 percent of being on target to achieve the MDGs . . . . 20 1.5 Growth and CPIA scores are higher in countries on track or close to being on track . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28 1.6 A one-standard-deviation increase in growth is definitely achievable . . . . . . . . . . . 36 2.1 Global output, 2007–14 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 44 2.2 Net financial flows . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 52 2.3 Inflows of international remittances . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 53 4.1 Poverty rates, latest year available . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 105 4.2 Years of schooling, latest year available . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 107 4.3 Status of indigenous groups relative to the MDGs, selected African countries . . . . 113 A1.1 Data availability by MDG, income, and geographic region, fiscal 2011 . . . . . . . . 160 A1.2 MDG performance: share of countries close to the target in the total number of countries off target, by geographic region . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 160 A1.3 MDG performance: share of countries close to the target in the total number of countries off target, by income level . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 161 A1.4 MDG performance: share of countries close to the target in the total number of countries off target, by alternative typologies . . . . . . . . . . . . . . . . . . . . . . . . . . 161 A1.5 Development indicators, average levels circa 1990 . . . . . . . . . . . . . . . . . . . . . . . . 162 A1.6 Growth has a clear, pervasive effect, but better policy also helps improve the odds of achieving the MDGs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 164 A1.7 Small-Hsiao tests of IIA assumption—null hypothesis: odds (outcome-j versus outcome-k) are independent of other alternatives . . . . . . . . . . . . . . . . . . . . . . . . . 166 A1.8 Predicted probabilities and marginal effects . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 167 A1.9 Effects of a one-standard-deviation increase in selected development drivers . . . . 168 A1.10 Probability of being on target or far from the target average . . . . . . . . . . . . . . . . . 169 A1.11 Probability of being on target or close to the target average . . . . . . . . . . . . . . . . . 169 A1.12 Probability of being in percentile (66) or above or in percentile (33) or below . . . 170 A1.13 Classification of economies by region and income, fiscal 2011 . . . . . . . . . . . . . . . 172 Foreword T his year we are facing historic develop- the goals. Moreover, progress could stall with- ment challenges—from natural disasters, out stronger global growth, expanded access to to food and fuel price spikes, and pro- export markets for developing countries, and found change in the Middle East. Despite high adequate assistance from donors. average growth in the developing world, it is We are making headway against poverty: crucial to provide opportunities to those that are based on our best economic projections, the being left behind. Wealthier economies are expe- world remains on track to reduce by half the riencing slower growth—but development assis- number of people living in extreme poverty. We tance needs remain high. In our interconnected project that by 2015, 882.7 million people will world, sustainable recovery means supporting be living on less than $1.25 a day, compared with inclusive growth. 1.4 billion in 2005 and 1.8 billion in 1990. Yet a Only four years remain until the 2015 dead- substantial portion of this progress reflects rapid line for reaching the Millennium Development growth in China and India, while many African Goals. The Global Monitoring Report 2011: countries are lagging behind: 17 countries are far Improving the Odds of Achieving the MDGs— from halving extreme poverty, even as the aggre- Heterogeneity, Gaps, and Challenges underlines gate goals will be reached. the urgency of helping countries that are behind Developing countries as a whole will also on meeting key targets for extreme poverty, hun- likely achieve the MDGs for gender parity in pri- ger, disease, and child and maternal mortality. mary and secondary education, for access to safe The report lays out the challenges that remain; drinking water, and will be very close on hunger analyzes efforts to improve human development; and primary education completion. But progress and assesses the role of growth, policy reforms, is too slow on meeting goals for child and mater- trade, and donor policies in meeting the MDGs. nal mortality as well as access to sanitation. The findings from this year’s report offer rea- Many of the poorest countries—even some son for both hope and concern. that have made substantial progress—face severe Two-thirds of developing countries are on tar- difficulties in achieving the MDGs. This is par- get or close to being on target for all the MDGs. ticularly true for those affected by violence and Among developing countries that are falling fragility. The difficulties reflect the drawbacks of short, half are close to becoming on track; with their initial conditions; by contrast, many middle- improved policies and faster growth, these coun- income countries, which entered the 1990s with tries could still achieve the targets in 2015 or better policies, better institutions, and higher soon after. Yet even those middle-income coun- growth, are now in a better position to achieve tries on track to achieve the MDGs are home to the goals. Yet we have seen how improved poli- indigenous and socially excluded groups that are cies in low-income countries did contribute to still very poor and often well behind in reaching strong growth in the years before the financial GLOBAL MONITORING REPORT 2011 xi xii FOREWORD GLOBAL MONITORING REPORT 2011 crisis. Policy buffers that allowed for active coun- Second, low-income countries need to achieve tercyclical policies softened the impact of the cri- and sustain more rapid economic growth and sis and drove a relatively rapid return to precrisis restore their policy buffers. One major challenge growth rates. To substantially reduce poverty and for growth is addressing the infrastructure gap meet the MDGs, low-income and other develop- through stronger public investment management ing countries need to grow faster and shore up systems and enhanced efforts to mobilize domes- their ability to guard against future shocks. tic revenue. Another is to create private sector To help us better understand results on the opportunities, for local business people as well ground, this year’s report presents findings and as foreign investors. Entrepreneurs’ small enter- lessons from impact evaluations in health and prises can create jobs, opportunity, and hope. education. Such evaluations often show that International organizations can offer policy progress in human development has not improved advice and technical assistance; donors need to as much as might be expected, given substantial meet aid commitments and strengthen the over- recent increases in resources devoted to health all effectiveness of aid. The world can also assist and education services. This is partly because, too by expanding trade opportunities, particularly often, scant attention has been given to improv- for the poorest countries. Key steps include com- ing incentives for rural doctors and health work- pleting the Doha Round, and expanding and ers, for example, so as to improve the quality of improving Aid for Trade, connected to trade services. This might explain the slower progress facilitation and building capacity. toward MDGs measured by outcomes (such as Third, fragile states lag the furthest behind those for health), than those MDGs measuring in reaching the MDGs. These countries require access to services, (such as those for education). additional support in building institutions and A key lesson is that strengthening institutions and moving toward a virtuous circle of development, improving incentives—by enhancing the role of peace, and security, as discussed in the Bank’s performance in setting the pay of health workers, recently released World Development Report for example—are vital to better outcomes. 2011: Conflict, Security, and Development. To regain momentum toward achieving the The MDGs were designed to provide a frame- MDGs, we need international cooperation on work for the entire international community to three fronts: work together toward a common end: making First, low-income countries in particular sure that human development reaches everyone, require a strong and stable global economic envi- everywhere. If these goals are achieved, billions ronment to continue growing. While advanced more people will have the opportunity to benefit economies are facing subdued growth and high from the global economy. We have four years unemployment, many fast-growing emerging left to reach the targets. International financial markets are seeing inflation pressures build. institutions, including regional development With global growth projected at 4.4 percent this banks, are adapting policies, procedures, and year and next, advanced market economies will tools to respond to the postcrisis world and to need to repair and reform their financial systems tailor support for unique country circumstances and tackle their fiscal imbalances in order to and needs—including economies and peoples hit strengthen the recovery. Emerging market econo- hard by banking problems, food and fuel crises, mies must adjust their macroeconomic policies and natural disasters. In doing so, we need to to reorient growth domestically and avoid pos- keep the MDGs firmly in our sights. Reaching sible overheating. these goals is central to global development. Robert B. Zoellick Dominique Strauss-Kahn President Managing Director The World Bank Group International Monetary Fund Acknowledgments T his report has been prepared jointly by Deon Filmer, Ariel Fiszbein, Ann Harrison, the staffs of the World Bank and the Jeffrey Lewis, Mohammad Zia Qureshi, International Monetary Fund (IMF). Martin Ravallion, and Ritva S. Reinikka. The cooperation and support of staff of sev- Several people also provided specific sugges- eral institutions are also gratefully acknowl- tions: Andrew Burns, Jeffrey Chelsky, Punam edged: the African Development Bank (AfDB), Chuhan-Pole, Rui Coutinho, Aart C. Kraay, the Asian Development Bank (ADB), the Barbara W. Lee, Sonia Plaza, Luis Serven, European Bank for Reconstruction and Devel- Lucio Vinas de Souza, Dominique van der opment (EBRD), the Inter-American Develop- Mensbrugghe, and Adam Wagstaff. ment Bank (IDB), as well as the Netherlands Several staff members also made valuable Environmental Assessment Agency (PBL). contributions and inputs, including the follow- Delfin S. Go was the lead author and man- ing from the World Bank: Harold Alderman, ager of the report. Brad McDonald led the Uranbileg Batjargal, Jinzhu Chen, Shaohua team from the IMF. The principal authors Chen, Stacey Tai Sie Chow, Yohana Dukhan, of the chapters were: chapter 1–Delfin S. Yoichiro Ishihara, Lauren Murphy, Victor Go and Jose Alejandro Quijada; chapter Hugo Orozco Olvera, Israel Osorio-Rodarte, 2–Brad McDonald; chapter 3–Gayle Martin Prem Sangraula, Jennifer Sturdy, and Edit and Shwetlena Sabarwal; chapter 4–Harry Velenyi. Other contributors from the IMF Anthony Patrinos; and chapter 5–Amy Hey- included Sibabrata Das, Nisreen Farhan, Shaun man (aid), Mariam Malouche (trade), and Roache, Manrique Saenz, Mika Saito, and Joe Stefano Curto (international financial institu- Thornton. Contributors from other institu- tions). Jose Alejandro Quijada, Sachin Shah- tions included Paul Lucas, Henk Hilderink, ria, and William Shaw were key members of Marcel Kok, Ben ten Brink, and Stefan van der the core team in the overall preparation and Esch (PBL); Patrica Laverley and Agnes Soucat coordination of the report. (AfDB); Indu Bhushan, Valerie Reppelin-Hill, Bruce Ross-Larson was the principal Manju Senapaty, and Gina Marie Umali editor. (ADB); and Murat Jadraliyev, Anita Taci, and The work was carried out under the general James Earwicker (EBRD). guidance of Justin Lin and Hans Timmer at Guidance received from the Executive the World Bank. Supervision at the IMF was Directors of the World Bank and the IMF provided by Hugh Bredenkamp and Cath- and their staffs during discussions of the draft erine Pattillo. Advisers to the report include report is gratefully acknowledged. The report Shantayanan Devarajan, Shahrokh Fardoust, also benefited from many useful comments GLOBAL MONITORING REPORT 2011 xiii xiv ACKNOWLEDGMENTS GLOBAL MONITORING REPORT 2011 and suggestions received from the Bank and communication and dissemination activities. IMF management and staff in the course of its The translation process was coordinated by preparation and review. Jorge del Rosario. The multilingual web sites accompanying The design and production of the Global the report were produced by Vamsee Kanchi, Monitoring Report 2011 was handled by the Roula Yazigi, Rebecca Ong, and Swati Pria- World Bank’s Office of the Publisher, under darshini Mishra; the main GMR Web address the supervision of Stephen McGroarty, Susan is www.worldbank.org/gmr2011. Rebecca Graham, and Denise Bergeron. Ong and Merrell Tuck-Primdahl managed the Abbreviations and Acronyms ADB Asian Development Bank IDA International Development AfDB African Development Bank Association AIDS acquired immune deficiency IFC International Finance syndrome Corporation BRICs Brazil, Russia, India, and IFI international financial China institution CAT DDO post–natural catastrophe IIA independence of irrelevant deferred drawdown option alternatives CCT conditional cash transfer IMF International Monetary Fund CFM capital flow management LDC least-developed country CPIA Country Policy and MDB multilateral development bank Institutional Assessment MDGs Millennium Development CRW Crisis Response Window Goals DAC Development Assistance NTM nontariff measure Committee OECD Organisation for Economic DDO deferred drawdown option Co-operation and DFQF duty free, quota free Development DIME Development Impact PBL Netherlands Environmental Evaluation Assessment Agency DMC developing member-country PCL Precautionary Credit Line DPL development policy lending PRGT Poverty Reduction and EBRD European Bank for Growth Trust Reconstruction and PRST Poverty Reduction Strategy Development Paper FCL Flexible Credit Line QuODA Quality of Official FDI foreign direct investment Development Assistance GDP gross domestic product Assessment G-20 Group of Twenty RPGs regional public goods HIV human immunodeficiency SDRs special drawing rights virus SMEs small and medium enterprises HRF Haiti Reconstruction Fund TFP Trade Finance Program IADB Inter-American Development WBG World Bank Group Bank WEO World Economic Outlook WTO World Trade Organization All amounts are presented in U.S. dollars, unless otherwise indicated. xv Goals and Targets from the Millennium Declaration GOAL 1 ERADICATE EXTREME POVERTY AND HUNGER TARGET 1.A Halve, between 1990 and 2015, the proportion of people whose income is less than $1.25 a day TARGET 1.B Achieve full and productive employment and decent work for all, including women and young people TARGET 1.C Halve, between 1990 and 2015, the proportion of people who suffer from hunger GOAL 2 ACHIEVE UNIVERSAL PRIMARY EDUCATION TARGET 2.A Ensure that by 2015, children everywhere, boys and girls alike, will be able to complete a full course of primary schooling GOAL 3 PROMOTE GENDER EQUALITY AND EMPOWER WOMEN TARGET 3.A Eliminate gender disparity in primary and secondary education, preferably by 2005, and at all levels of education no later than 2015 GOAL 4 REDUCE CHILD MORTALITY TARGET 4.A Reduce by two-thirds, between 1990 and 2015, the under-five mortality rate GOAL 5 IMPROVE MATERNAL HEALTH TARGET 5.A Reduce by three-quarters, between 1990 and 2015, the maternal mortality ratio TARGET 5.B Achieve by 2015 universal access to reproductive health GOAL 6 COMBAT HIV/AIDS, MALARIA, AND OTHER DISEASES TARGET 6.A Have halted by 2015 and begun to reverse the spread of HIV/AIDS TARGET 6.B Achieve by 2010 universal access to treatment for HIV/AIDS for all those who need it TARGET 6.C Have halted by 2015 and begun to reverse the incidence of malaria and other major diseases GOAL 7 ENSURE ENVIRONMENTAL SUSTAINABILITY TARGET 7.A Integrate the principles of sustainable development into country policies and programs and reverse the loss of environmental resources TARGET 7.B Reduce biodiversity loss, achieving by 2010 a significant reduction in the rate of loss TARGET 7.C Halve by 2015 the proportion of people without sustainable access to safe drinking water and basic sanitation TARGET 7.D Have achieved a significant improvement by 2020 in the lives of at least 100 million slum dwellers GOAL 8 DEVELOP A GLOBAL PARTNERSHIP FOR DEVELOPMENT TARGET 8.A Develop further an open, rule-based, predictable, nondiscriminatory trading and financial system (including a commitment to good governance, development, and poverty reduction, nationally and internationally) TARGET 8.B Address the special needs of the least-developed countries (including tariff- and quota-free access for exports of the least-developed countries; enhanced debt relief for heavily indebted poor countries and cancellation of official bilateral debt; and more generous official development assistance for countries committed to reducing poverty) TARGET 8.C Address the special needs of landlocked countries and small island developing states (through the Programme of Action for the Sustainable Development of Small Island Developing States and the outcome of the 22nd special session of the General Assembly) TARGET 8.D Deal comprehensively with the debt problems of developing countries through national and international measures to make debt sustainable in the long term TARGET 8.E In cooperation with pharmaceutical companies, provide access to affordable, essential drugs in developing countries TARGET 8.F In cooperation with the private sector, make available the benefits of new technologies, especially information and communications Source: United Nations. 2008. Report of the Secretary-General on the Indicators for Monitoring the Millennium Development Goals. E/CN.3/2008/29. New York. Note: The Millennium Development Goals and targets come from the Millennium Declaration, signed by 189 countries, including 147 heads of state and government, in September 2000 (http://www.un.org/millennium/declaration/ares552e.htm) and from further agreement by member states at the 2005 World Summit (Resolution adopted by the General Assembly–A/RES/60/1). The goals and targets are interrelated and should be seen as a whole. They represent a partnership between the developed countries and the developing countries “to create an environment—at the national and global levels alike—which is conducive to development and the elimination of poverty.” Overview W ith less than five years remaining Progress on both fronts has been evident to achieve the Millennium Devel- since the 1990s. Indeed, policy responses opment Goals (MDGs), the inter- in the recent global economic downturn national development community has to set have softened the negative impact, particu- priorities that focus on lagging countries and larly for low-income countries (especially sectors and take onboard lessons from past in Africa). interventions, reflecting the limited resources • This substantial progress is not a reason for available. To inform this effort, Global Mon- complacency. Without a stable expansion itoring Report 2011 analyzes the diverse of the global economy, continuing access record in improving human development to advanced and developing-country mar- across and within developing countries; sum- kets, and adequate assistance from donors, marizes the results of impact evaluations of progress could still break down. Enhancing health and education programs; and reviews the resilience to adverse economic shocks, recent developments in global growth, trade, including the provision of social safety and donor policies. programs, will need greater attention and The key messages: support. • Reaching the MDGs is only one milestone, • Global progress toward the various targets for there still is much work to do in fos- continues to be mixed, and country per- tering inclusive growth, reducing inequal- formance is predictably diverse. Among ity and poverty, and improving health and developing countries that are off track, the education outcomes in even the most suc- top half are, on average, within 10 percent cessful countries. Even the middle-income of the on-track trajectory. While countries countries on track to reach the MDGs have close to the target may still miss the 2015 trailing pockets of indigenous and socially deadline, they could achieve the targets excluded groups whose odds of reaching soon after, with improved policies and an the goals are slim. acceleration of growth to precrisis levels. • Unsurprisingly, countries with slower • For countries that are on track, or close to growth and poorer institutions are farthest it, solid economic growth and good policies behind. Many countries far from the tar- and institutions have been the key factors. get are fragile states, reinforcing the need 1 2 OVERVIEW GLOBAL MONITORING REPORT 2011 for the international community to step up education, access to safe drinking water, and support to these countries. eradicating extreme poverty and hunger, in • Impact evaluations show that the quan- that order (figure 1). On current trends, and tity of health and education services has despite the recent global economic crisis, increased, but not the quality. This might developing countries are on track to reach the be one reason that progress has been slower global target of cutting income poverty in half for MDGs measured by outcomes (like by 2015, thanks in large part to rapid growth those for health) than for those measured in China and India. Developing countries by access (like those for education). That will also likely achieve the MDGs for gender is why improving the incentives for service parity in primary and secondary education providers and consumers—especially mak- and for access to safe drinking water, and ing service providers directly accountable will be very close on hunger and on primary to consumers—is so essential to improving education completion. Progress is too slow, outcomes. however, on health-related outcomes—such • The growing assistance from new donors— as child and maternal mortality and access many in the developing world—will not to sanitation—so the world will likely miss compensate for a significant future fall in these MDGs by 2015. Most regions are lag- aid from traditional donors, particularly if ging on these health goals, but East Asia and they pursue different development priori- Pacific, Latin America and the Caribbean, ties and practices. The changing aid land- and Europe and Central Asia are doing better scape could also have implications for the than other regions. transparency of official flows and for the Poor countries and regions tend to lag in policies and programs that aid supports. attaining the MDGs. The low-income coun- tries lag on all the MDGs. While poverty in Sub-Saharan Africa has fallen steeply with the The diversity of country acceleration of growth since 2000, the region performance is not on track to meet the poverty reduction The global numbers tell a mixed story. goal. Middle-income countries—both lower Progress has been substantial in achieving and upper—generally exhibit the best perfor- the MDGs for gender parity in primary and mance. Eighty percent of the upper-middle- secondary education, completion of primary income countries (23 of them) have achieved FIGURE 1 The distance to global MDGs ranges widely 100 90 96% distance to goal achieved by 91% developing countries, % 80 90% 88% 70 80% 77% 60 69% 67% 50 57% 40 30 20 10 0 MDG 1.a MDG 1.c MDG 2.a MDG 3.a MDG 3.a MDG 4.a MDG 5.a MDG 7.c MDG 7.c extreme hunger primary gender gender child maternal access to access to povery completion parity parity mortality mortality safe water sanitation rate (primary) (secondary) under five Source: World Bank staff calculations based on data from the World Development Indicators database. Note: Distance to goal achieved in this graph is a weighted average of the latest indicators, using population weights in 2009. GLOBAL MONITORING REPORT 2011 OVERVIEW 3 or are on track to achieve the extreme poverty have achieved or are on track to achieve the eradication target. And 72 percent of lower- targets for gender parity in primary and sec- middle-income countries (34 of them) are set ondary education. Although half the moni- to reach the target for gender parity in second- tored countries (57) are off target for the pri- ary education. Both low- and middle-income mary education completion goal, two thirds countries have made good progress on gender of them (38) are very close. parity in primary education. Progress is still mixed or poor for access Even so, the targets remain within reach to sanitation, maternal mortality, and child for many developing countries. Although mortality. For example, more than 40 percent more developing countries are off track than of low- to upper-middle-income countries (58 on track in achieving the targets, many off- of them) are lagging significantly on access to track countries are closer to the goals, thanks sanitation. to more than a decade of better policy and The variation among lagging countries faster growth. A country is “close to the is still large, but the average gap is not. target” if its distance to getting on target is Lagging countries are, on average, 23 per- smaller than the average gap of all lagging cent behind being on track to achieve all the countries. Others are “far from the target,” MDGs (table 1). They are close to being on if their distance to getting on target is bigger track for gender parity in primary educa- than the average gap. tion (7 percent); gender parity in secondary Indeed, two-thirds or more of developing education (16 percent); hunger (19 percent); countries are, on average, on target or close primary education completion (20 percent); to being on target for all the MDGs (figure 2). and, to some extent, under-five mortality Many countries are on track to achieve sev- (23 percent). But for each target there are eral MDGs: gender parity in primary educa- countries with scant progress. For example, tion (89 of them), gender parity in secondary 17 countries are far from halving extreme education (82); access to safe drinking water poverty, even as the global goal will be (66); primary completion rate (55); and wip- reached. ing out extreme poverty (47). For instance, More important, among countries that are about 70 percent of developing countries off track, the top half are, on average, about FIGURE 2 More than two-thirds of developing countries are on track or close to being on track 80 proportion of developing countries, % 89 82 70 60 47 55 66 50 25 58 40 52 53 46 48 18 38 39 30 21 36 35 37 12 30 17 25 23 27 20 19 13 15 10 0 MDG 1.a MDG 1.c MDG 2.a MDG 3.a MDG 3.a MDG 4.a MDG 5.a MDG 7.c MDG 7.c extreme hunger primary gender gender child mortality maternal access to access to povery completion parity parity under five mortality safe water sanitation rate (primary) (secondary) on target close to target far from target Source: World Bank staff calculations based on data from the World Development Indicators database. Note: The figure above each bar is the number of countries. 4 OVERVIEW GLOBAL MONITORING REPORT 2011 TABLE 1 Lagging countries are close to getting on target Average distance to getting on target (gaps, %) All off-target countries Countries close to the target Countries far from the target MDG 1.a Extreme poverty 39 (96) 17 67 MDG 1.c Hunger 19 (60) 9 35 MDG 2.a Primary education completion 20 (96) 9 40 MDG 3.a Gender parity in primary education 7 (22) 4 14 MDG 3.a Gender parity in secondary education 16 (52) 8 29 MDG 4.a Child mortality under five 23 (59) 8 38 MDG 5.a Maternal mortality 32 (80) 11 51 MDG 7.c Access to safe drinking water 25 (76) 14 41 MDG 7.c Access to sanitation 27 (50) 16 34 Simple average 23 11 39 Source: World Bank staff calculations based on data from the World Development Indicators database. Note: A country is “close to the target” if its distance to getting on target is smaller than the average gap of all lagging countries. It is “far from the target” if its distance is bigger than the average gap. Figures in parentheses indicate the range of variation (maximum value minus mini- mum value) of countries off target, by MDG. Averages and numbers of countries cover only those with data, which can vary by MDG. 10 percent off the on-track trajectory. Their estimate of poverty at the global and regional mean distance is only 4–9 percent for gender levels for 2008. There is also spotty infor- parity in primary and secondary education, mation about infectious diseases, such as child mortality, primary education comple- malaria and tuberculosis. Other data issues tion, and alleviating hunger. Indeed, these include reporting errors that severely hamper countries close to the target need to increase the accuracy of the maternal mortality ratio. primary education completion by only 9.2 The quality, timeliness, and availability of percent, on average, to become on track. data are gradually improving, which can be Reducing data gaps will improve the seen from the World Bank Statistical Capac- assessment of progress. Poor data availabil- ity Indicator (http://bbsc.worldbank.org); but ity may fail to convey progress or understate additional effort and support for statistical deterioration resulting from shocks and con- capacity-building activities will be required flicts. Among the MDGs, the collection of to improve data and close gaps for all MDG- data on poverty seems to lag the most. Con- related indicators. ducting household income and expenditure surveys in both large and poor countries is difficult, often delaying global or regional The role of growth and policy poverty updates when the countries with Starting points—inherited initial condi- missing data account for a large share of the tions—count in MDG performance, but total. The new household surveys employed subsequent growth and policy also matter in the analysis covered about 43 percent of greatly. Countries with a higher per capita the population in developing countries in GDP in 1990 generally have better MDG 2008 and 7.6 percent in 2009. However, performance. Countries starting with good gaps for 2008–10 household surveys remain policy and institutions also tend to do better. in several regions—particularly, South Asia But initial conditions do not fully determine (pending India’s new household survey, 10.8 outcomes, and economic growth and policy percent of the population is covered), Middle performance after the initial year may be East and North Africa (19.1 percent), and more important in meeting the goals. Coun- Sub-Saharan Africa (20.1 percent). Efforts tries that have reached or are on track to are already under way at the World Bank to reach the targets show, on average, the fast- close the remaining gaps before the end of est per capita GDP growth over 1990–2009 2011 to complete and update the time-series (table 2). Similarly, countries close to the GLOBAL MONITORING REPORT 2011 OVERVIEW 5 TABLE 2 Growth and CPIA scores are higher in countries on track or close to being on track Average values across MDGs (weighted by the number of countries in each MDG category) Close to Far from On target the target the target Average GDP per capita growth (1990–2009) 2.4 1.8 1.2 CPIA index (2009) 3.7 3.5 3.3 Source: World Bank staff calculations based on data from the World Development Indicators database. Note: The pairwise correlation between average GDP per capita growth and the CPIA index is 0.32 (significant at 0.01). A country is “close to the target” if its distance to getting on target is smaller than the average gap of all lagging countries. It is “far from the target” if its distance is bigger than the average gap. target tend to have faster per capita growth in developing countries by one standard devi- than countries far from the target. A strong ation is feasible, although the implied rate of policy and institutional framework, as mea- 3.8 percent in average developing-country per sured by the World Bank’s annual country capita growth is roughly double its historical policy and institutional assessment (CPIA), 1.8 percent rate (see table 2), it is comparable appears to be associated with improved both to recent periods of accelerated growth MDG performance. and to current forecasts of growth for 2010– Both factors—initial conditions and subse- 15. Achieving the assumed improvement in quent growth and policy—also point to why policy and institutions may be more chal- the MDGs are such big challenges for the lenging, given the few years to 2015. Because poorest and most fragile countries. Middle- policy reforms can take time to implement income countries, having grown earlier, often and bear fruit, it is important to begin now. have better policies and institutions. Although many close-to-target countries Growth has an all-encompassing bear- may still miss the 2015 deadline, with higher ing on the progress of MDGs, while policy growth and better policies they could still is vital for outcome-based MDGs. Prelimi- achieve the targets soon afterward. nary econometric results suggest that eco- nomic growth has a pervasive association with the odds of achieving the MDGs. For Prospects and challenges for those countries currently far from the goals, a economic growth strong growth performance seems to catalyze The global economic recovery is proceeding progress toward the MDGs. And for those along multiple tracks. Advanced economies that are closer to becoming on target, growth are slowly recovering from the recent global seems to be significantly correlated with spe- economic crisis and face continued high cific goals such as primary education comple- unemployment. By contrast, emerging econo- tion and gender parity. Improved policy and mies have seen a robust recovery, and some institutions appear to be especially important faster-growing economies are experiencing for health-related MDGs, such as maternal inflation pressures amid signs of overheating. mortality, under-five mortality, hunger alle- Thanks in part to their countercyclical policy viation, and access to safe drinking water; as responses to the crisis, low-income economies outcome-based goals, they depend (in addi- are seeing a relatively rapid return to precrisis tion to growth and resources) on a myriad growth rates. Higher commodity prices are of factors that improve the quality of public supporting growth in commodity-exporting expenditures and service delivery. countries, but are sparking concerns over Overall, an increase in growth and in the the affordability of food for the poorer seg- quality of policies equivalent to one stan- ments of the population in some low- and dard deviation would appear to put 32 more lower-middle-income countries. Global GDP developing countries (44 percent) on track to is forecast to rise by about 4.5 percent in 2011 meet the MDGs. Increasing average growth and 2012, with rates in advanced economies 6 OVERVIEW GLOBAL MONITORING REPORT 2011 TABLE 3 Global output, 2007–14 annual percentage change Projections Region 2007 2008 2009 2010 2011 2012–14 World output 5.4 2.9 –0.5 5.0 4.4 4.6 Advanced economies 2.7 0.2 –3.4 3.0 2.4 2.5 Emerging and Developing Economies 8.8 6.1 2.7 7.3 6.5 6.6 Central and Eastern Europe 5.5 3.2 –3.6 4.2 3.7 3.9 Commonwealth of Independent States 9.0 5.3 –6.4 4.6 5.0 4.6 Developing Asia 11.4 7.7 7.2 9.5 8.4 8.5 Middle East and North Africa 6.2 5.1 1.8 3.8 4.1 4.5 Sub–Saharan Africa 7.2 5.6 2.8 5.0 5.5 5.7 Western Hemisphere 5.7 4.3 –1.7 6.1 4.7 4.0 Emerging Economies 9.2 6.3 2.6 7.5 6.7 6.7 Other Developing Economies 7.2 6.0 5.2 6.2 6.1 6.4 Least Developed Countries (LDCs)a 9.0 6.9 5.2 5.3 6.1 6.4 Source: World Economic Outlook. a. United Nations classification, a subset of developing countries. several percentage points below those in currency appreciation are resulting in a slug- emerging and developing economies (table 3). gish monetary policy response in several Good macroeconomic policies remain emerging economies. Differences in interest crucial to the recovery. Continuing accom- rates and growth prospects have spurred modative monetary policy in most advanced strong capital flows from the advanced to the economies has alleviated the financial cri- emerging economies, adding to the overheat- sis and recession, and monetary conditions ing and complicating the policy response. in most emerging and developing countries The challenge in low-income countries began to normalize in 2010. Fiscal policy is to sustain and accelerate growth through is shifting from supporting recovery to cut- better policies and greater investment in ting deficits, although easy credit conditions infrastructure. Closing the large infrastruc- may be slowing the pace of fiscal consolida- ture gap by raising productivity and encour- tion among major emerging economies. Low- aging private investment could substantially income countries have, appropriately, begun increase per capita income growth. Effective to reverse the unprecedented countercyclical public investment management—better stra- fiscal response that softened the impact of tegic planning and analyses of the feasibility the crisis. In many countries, policy makers of projects—is critical. The financing of addi- can also take steps to improve employment tional investment also needs to be carefully opportunities, including for young adults. managed, to ensure that debt sustainability is Sustaining the global recovery demands preserved, and should be supported by paral- greater effort. To avoid unsustainable debt lel reforms to strengthen tax revenue. burdens, the advanced economies must Rebuilding policy buffers can help protect redress fiscal imbalances and repair and growth in the face of future shocks. Good reform fi nancial systems. Emerging econo- policies in low-income countries contributed mies face policy challenges of overheating to strong growth before the crisis. The policy and strong capital inflows. Although core buffers established then also created the space inflation remains subdued in many of them, for countercyclical policies that softened the inflationary expectations are rising and pol- impact of the crisis and drove a fairly rapid icy targets have been exceeded in some cases. return to precrisis growth rates. To ensure Inflation could threaten otherwise sound pol- that higher growth is sustained in the face icy frameworks, but concerns about further of future shocks—a prerequisite for meeting GLOBAL MONITORING REPORT 2011 OVERVIEW 7 the MDGs—low-income and other develop- facing both service providers and consumers. ing countries need to rebuild their policy buf- Too much effort has been devoted to increas- fers as the recovery proceeds. In addition to ing inputs, and not enough to ensuring that strong, well-designed national policies, inter- institutions provide services efficiently and national cooperation is required to restore a responsively—and that consumers have the global economic environment conducive to ability and incentive to use services efficiently development and poverty reduction, and to and hold service providers accountable for support the most vulnerable countries. quality. Some of the lessons are these: • Reducing user costs and providing cash Making the right policy transfers has improved the uptake of health interventions—lessons from and education services, but issues of qual- impact evaluations in education ity remain. and health • Health inputs have had to grapple with Improving the likelihood of more countries uptake and use by citizens, and school- attaining the MDGs depends not just on based and community-based approaches more resources but also, and quite critically, are tackling this. But more research and on improving the quality of service provision impact evaluations are needed to discover through better policies and stronger insti- when and how these approaches work. tutions. But not much is known about how • Increasing traditional schooling inputs has exactly to do this. To help inform this ques- often been ineffective in improving learn- tion, a special feature of this year’s report ing outcomes—as in Bolivia, Colombia, presents the findings and lessons from impact Kenya, and Nicaragua. evaluations in health and education. • Service delivery too often fails the poor— Development assistance for health and as in Bangladesh, Brazil, Ecuador, Ghana, education has risen to unprecedented lev- India, Indonesia, Morocco, Peru, Tunisia, els in volume, but has not generated the and Uganda. expected improvements in outcomes. And • Pay for performance improved student given the current global economic environ- learning outcomes—as in India, Israel, and ment, citizens in developed and developing Kenya. The Rwanda health center pay-for- countries alike are demanding more value for performance program increased institu- their money. This requires closer attention tional deliveries and preventive care visits to the causal chain linking spending to out- by young children and improved the qual- comes and actions to isolate and strengthen ity of prenatal care. the weak links in this chain. In recent years, • New approaches are being designed to impact evaluations have emerged as a tool to improve service delivery, and an evidence do this; and even though their evidence base base on the effectiveness of these approaches is still far from complete, some interesting is beginning to take shape. lessons are beginning to emerge. Impact evaluations highlight the discon- These are just some of the compelling mes- nect between increased public spending and sages from the evaluations summarized in changes in outcomes. The outcomes have chapter 3. been disappointing, partly because the spend- Although designing the right approach ing focus has been narrowly trained on input to improving outcomes is complex and provision, ignoring other parts of the causal depends on context, a systematic evidence chain that links public spending to better base generated through rigorous impact outcomes. Inputs continue to be important, evaluations can provide useful guidance but alone they are not sufficient for attaining for policy makers. The information base on the goals in many developing countries. Poli- impact evaluations is far from complete, and cies have failed to account for the incentives much more work is needed before the most 8 OVERVIEW GLOBAL MONITORING REPORT 2011 pressing questions can be answered. But some and evaluation of programs. The more suc- new approaches show promise in improving cessful programs, such as those in Asia, have incentives and in strengthening the account- extended economic opportunities and pro- ability of service provision. vided cash or in-kind assistance, thus allow- ing these groups to use the targeted assistance more effectively. Other countries may need to Assisting the poorest of the target these groups more tightly and raise the poor—socially excluded groups quality of services by increasing the account- Reaching the MDGs requires addressing ability of service providers to their clients. the plight of the world’s socially excluded In addition to assisting the indigenous groups, including indigenous people, ethnic peoples, more inclusive growth and equal- minorities, and linguistic groups. They make ity within countries will also benefit people up a sizable proportion of the world’s people in the bottom quintiles and lift more people and an even greater share of the world’s poor. out of poverty. The case of Brazil points to Turning the situation around will therefore two lessons: (1) reforms to social policies and require widespread and sustainable economic programs to make them more pro-poor (if growth, as well as specific interventions to fiscally possible) can play an important role reach these groups. in sustaining poverty reduction, even during Most MDG indicators for indigenous a period of economic stagnation; and (2) sen- people and ethnic minorities are worse than sible macroeconomic and trade policies need population averages. This is true for under- not hurt the poor and, in the specific case of five mortality; adult literacy; school enroll- taming hyperinfl ation, are likely to make a ment, completion, and achievement; gender significant contribution in the fight against equity; water deprivation; child nutrition; poverty, even when that is not the primary and, especially, poverty reduction. Most objective. countries in Latin America with sizable indig- enous populations, for example, show almost no poverty reduction for those groups. Progress in the international Programs can meet the needs of indige- development framework nous people and ethnic minorities. Bilingual The international development framework education programs can promote national functioned well in coping with the crisis. language acquisition, school completion, and Governments and international institutions subsequent earnings gains, contributing to cooperated well in the face of recession, thus multiple goals—and be cost effective. Con- avoiding the perils of a downward protec- ditional cash transfer programs can promote tionist spiral, maintaining aid, and boost- schooling, health, and poverty reduction; ing emergency financial resources. Financing and these programs result in disproportion- from the international institutions jumped ate gains for minority populations, par- sharply, leading to a general increase in their ticularly in countries with large vulnerable resources. Although recent research suggests populations. that declines in aid have tended to deepen for Attaining the MDGs for indigenous peo- several years after banking-related crises, aid ple and ethnic minorities requires innova- levels from countries in 2010 have not yet tive approaches. These people may be hard been constrained by concerns over rising fis- to reach because they live in remote locations cal deficits. Some major-donor governments with poor transportation. They may also remained dedicated to maintaining aid levels suffer from social and economic discrimi- and others announced cuts. Even so, official nation and government neglect. So assisting development assistance (ODA) from members them may require more complex interven- of the Development Assistance Committee tions than those for the general population. (DAC) of the Organisation for Economic Co- The lack of data also complicates the design operation and Development (OECD) rose to GLOBAL MONITORING REPORT 2011 OVERVIEW 9 its highest level, at $128.7 billion, an increase even lower than the level it would have of 6.5 percent over 2009 in real terms. Devel- achieved if it had continued the 1995–2008 oping countries are also benefiting from new trend. The rise in protectionist measures dur- donors (including disbursements of some ing the crisis, which particularly affected $9–$10 billion a year from donors that are the exports of least-developed countries, not members of DAC) and from a sharp appears to be receding. Solidifying an open, rise in donations from the private sector in rules-based international trade regime can be advanced countries. accomplished best by concluding the Doha The past tumultuous decade has ushered Round. in significant changes in the assistance poli- The global community can support pov- cies of international institutions. They have erty reduction through improving trade inte- developed a results-based, country-driven gration in low-income countries. Poor-coun- assistance framework grounded in regular try market access could improve significantly reviews of country strategy and independent if rich countries extended duty-free, quota- evaluations. A more diverse and flexible range free access to all poor-country exports and of fi nancing instruments helped them tailor simplified rules of origin in preference agree- assistance to the needs of particular countries ments. Efforts also are required to ensure (such as those affected by conflict) and in par- that poor countries can access trade fi nance ticular situations (such as disaster relief and at reasonable cost, along with improvements crisis assistance). They are also doing more in data and a review of whether banking in providing knowledge—through loans and regulations impose excessive capital require- technical assistance, and global programs ments on trade finance transactions. Regional and projects—as well as free information to trade agreements should support open trade the global community. through low external tariffs, while technical The growing assistance from emerging assistance to developing-country trade nego- donors (many in the developing world) is tiators would support deeper regional inte- welcome but may not compensate fully for a gration. More financial resources and techni- significant fall in aid from traditional donors, cal assistance are needed to strengthen trade to the extent that the emerging donors pursue facilitation and to reduce supply constraints different development priorities and practices. on poor-country exports. These efforts This changing aid landscape could also have should include increased commitments for implications for the transparency of official aid for trade (which in real terms stagnated flows and the policies and programs that aid in 2009) and greater use of public-private supports. partnerships. Further efforts are required to Trade integration and facilitation remain connect landlocked countries and lagging essential for inclusive growth and poverty regions to regional and international markets. reduction. World trade, now recovering at Logistical improvements at the subnational about double the 2002–08 rate of growth, level are also critical for connecting rural and remains well below the precrisis peak and remote areas in developing countries. 1 The Diversity of MDG Progress W ith less than five years left to Several questions demand answers: How achieve the Millennium Develop- many countries are off target, and how far ment Goals (MDGs), the inter- are they from the goals? Why are some coun- national development community is show- tries behind? And what factors are key to ing renewed urgency to assess the various improving the odds that off-target countries development efforts, especially in light of the can reach the goals? recent global economic crisis and the still- fragile recovery. What are the prospects and challenges for reaching the goals? Answers Looking under global progress are clearly linked to the complex tapestry of The global numbers tell a familiar, mixed progress that lies below the global numbers. story (figure 1.1). The latest information The Global Monitoring Report [GMR] confirms that progress toward the MDGs 2011 extends the forward-looking analy- remains substantial on gender and educa- sis started in last year’s report by examin- tion, access to safe drinking water, extreme ing more carefully the diversity of progress poverty, and hunger, in that order. On cur- and its implications for the remaining gaps rent trends and despite the recent global eco- and challenges. It complements that report’s nomic crisis, the world is on track to reach analysis of the impact of the global crisis on the global target of cutting income poverty the MDGs. in half by 2015. Thanks to rapid growth in A key observation on the progress toward China, the East Asia and Pacific region has the MDGs is its diversity. At the aggregate already halved extreme poverty. Developing and regional levels, low-income countries, countries will also likely achieve the MDGs particularly fragile states and those in Sub- for gender parity in primary and secondary Saharan Africa, lag because of a combination education and for access to safe drinking of low starting points and difficult circum- water, and will be very close on hunger and stances. Behind those aggregate numbers, the primary education completion rate. however, the great diversity of performance Progress continues to lag in health-related across indicators, countries, and groups of development outcomes, such as child and countries requires further analysis. That is maternal mortality and access to sanitation. the subject of this chapter. New data and methodologies indicate much Rendille villagers in northern Kenya scoop the dregs from a water tank filled only the night before by a government truck but already drained below the level 11 of its spigot. They must wait a week for the next delivery. 12 THE DIVERSITY OF MDG PROGRESS GLOBAL MONITORING REPORT 2011 FIGURE 1.1 Current global distance to the MDGs is wide ranging 100 90 96% distance to goal achieved by 91% developing countries, % 80 90% 88% 70 80% 77% 60 69% 67% 50 57% 40 30 20 10 0 MDG 1.a MDG 1.c MDG 2.a MDG 3.a MDG 3.a MDG 4.a MDG 5.a MDG 7.c MDG 7.c extreme hunger primary gender gender child maternal access to access to povery completion parity parity mortality mortality safe water sanitation rate (primary) (secondary) under five Source: World Bank staff calculations based on data from the World Development Indicators database. Note: Distance to goal achieved in this graph is a weighted average of the latest indicators, using population weights in 2009. In this and other graphs in the chapter, the focus is on MDGs with well-defined targets and time-series data to assess progress. more progress than previously thought in Poor countries and regions tend to lag in reducing maternal mortality, but that is still attaining the MDGs. As a group, they lag the MDG that lags the most. On current on all the MDGs and are unlikely to reach a trends, the world will likely miss these three single target by 2015. Generally, fewer low- targets by 2015. Most regions are off track, income than middle-income countries are on but East Asia and the Pacific, Latin America target to achieve each MDG (figure 1.2). The and the Caribbean, and Europe and Central number of countries in each income group Asia are doing somewhat better than others. does not differ greatly, so this is not the result FIGURE 1.2 Fewer low-income countries are on track to achieve the MDGs 100 proportion of developing countries 90 37 23 32 80 28 34 25 by income level, % 70 7 32 60 18 25 23 30 10 19 50 20 40 8 11 16 13 6 15 11 30 10 20 5 2 4 10 1 0 MDG 1.a MDG 1.c MDG 2.a MDG 3.a MDG 3.a MDG 4.a MDG 5.a MDG 7.c MDG 7.c extreme hunger primary gender gender child mortality maternal access to access to povery completion parity parity under five mortality safe water sanitation rate (primary) (secondary) upper-middle-income countries lower-middle-income countries low-income countries Source: World Bank staff calculations based on data from the World Development Indicators database. Note: The number above each bar refers to the number of countries attaining that MDG. GLOBAL MONITORING REPORT 2011 THE DIVERSITY OF MDG PROGRESS 13 of the distribution of countries. In fact, about obscured by poor past performance and by 40 countries are currently classified as low the disproportionate challenges the MDGs income, compared with 54 lower-middle- pose in many Sub-Saharan African and other income and 48 upper-middle-income coun- poor countries.1 The longer the distance to tries (see appendix 1 for the classification). the 2015 targets, the more ambitious the Nor do the results arise from missing data in goals appear and the steeper the path to low-income countries because the availabil- achieve them. Ravallion and the World Bank ity of data also shows little variation among and IMF discussed how higher starting pov- income groups (see appendix table A1.1). At erty rates are generally associated with a the global level, the latest poverty data remain lower responsiveness (elasticity) of poverty at 2005, and box 1.1 explains why. to economic growth. 2 Countries in Sub-Sa- The pattern more likely stems from the haran Africa implemented reforms later than starting points—especially incomes, which others and therefore benefited much later matter greatly in attaining the MDGs. As from accelerating economic growth. How- several studies show, recent achievements are ever, after showing no decline for much of BOX 1.1 Poverty data and projections New poverty projections at the World Bank are a more rapid growth of per capita consump- the result of several changes—new and more tion than previously reported in the national recent household surveys, updates of historical accounts. This estimate will be updated further consumption per capita from national accounts, when India’s forthcoming household survey for and a new forecast of per capita consumption 2008 is completed. That survey will provide a growth. For this report, the poverty forecast more accurate estimate of household consump- includes 62 new household surveys out of a total tion for different income groups. Although the of 123 countries, which reflect methodological poverty numbers for East Asia and the Pacific advances in newer surveys and changes in the remain relatively stable, China’s poverty rate underlying distribution of income not measured and number (at the poverty level of $1.25 a day) by changes in mean income or expenditure. The have decreased further, based on the new 2008 forecast therefore captures changes in income household survey and a higher growth rate of its inequality in the new surveys. However, it household income. Projections for Sub-Saharan assumes inequality is unchanged in other coun- Africa are slightly better than previously esti- tries. Some of the effects of food and fuel price mated: its extreme poverty at $1.25 a day is pro- shocks in 2008 are captured in the new surveys. jected to be 35.8 percent in 2015, lower than the Based on the economic projections of devel- previous forecast of 38.0 percent, because of its oping countries at the International Monetary higher recent growth performance and growth Fund and the World Bank, the world remains on forecast. However, several of Africa’s household track to reduce by half the number of people liv- surveys for 2008 still need to be completed (see ing in extreme poverty. The number of people the discussion of gaps below). living on less than $1.25 a day is projected to be The new household surveys employed in the 882.7 million in 2015, which is lower than the forecast cover about 43.1 percent of the popu- previous estimate of 918 million (see the table lation in developing countries in 2008 and 7.6 on the next page). The decline results mainly percent in 2009 (see the figure on the next page). from data changes for India, which showed However, gaps for 2008–10 household surveys continued 14 THE DIVERSITY OF MDG PROGRESS GLOBAL MONITORING REPORT 2011 BOX 1.1 Poverty data and projections (continued) Estimates of poverty reduction on a poverty line of $1.25 and $2.00 a day, by region Region 1990 2005 2015 1990 2005 2015 Population living on less than $1.25 a day (%) People living on less than $1.25 a day (millions) East Asia and Pacific 54.7 16.8 5.9 873.3 316.2 119.0 China 60.2 15.9 4.8 683.2 207.7 66.1 Europe and Central Asia 2.0 3.7 1.2 9.1 17.3 5.8 Latin America and the Caribbean 11.3 8.2 4.7 49.6 45.1 29.1 Middle East and North Africa 4.3 3.6 1.3 9.7 11.0 4.8 South Asia 51.7 40.3 22.4 579.2 595.6 379.3 India 51.3 41.6 22.4 435.5 455.8 276.8 Sub-Saharan Africa 57.6 50.9 35.8 295.7 388.4 344.7 Total 41.7 25.2 14.4 1,816.6 1,373.5 882.7 Population living on less than $2.00 a day (%) People living on less than $2.00 a day (millions) East Asia and Pacific 79.8 38.7 19.7 1,273.7 728.7 399.4 China 84.6 36.3 15.4 960.8 473.7 213.4 Europe and Central Asia 6.9 8.9 4.5 31.9 41.9 21.4 Latin America and the Caribbean 19.7 16.6 10.7 86.3 91.3 66.3 Middle East and North Africa 19.7 16.9 7.2 44.4 51.5 26.2 South Asia 82.7 73.9 57.1 926.0 1091.5 967.2 India 82.6 75.6 56.9 701.6 827.7 702.0 Sub-Saharan Africa 76.2 73.0 57.7 391.2 556.7 555.6 Total 63.2 47.0 33.1 2,753.5 2,561.5 2,036.1 Source: World Bank staff calculations from PovcalNet database. remain at the regional level: South Asia (pending and Europe and Central Asia, 94.4 percent. India’s new household survey), 10.8 percent of Efforts are already under way at the World Bank population is covered; Middle East and North to close remaining gaps before the end of 2011 Africa, 19.1 percent; Sub-Saharan Africa, 20.1 to complete and update the time-series estimate percent; Latin America and the Caribbean, 84.1 of poverty at the global and regional levels for percent; East Asia and the Pacific, 90.3 percent; 2008. Population coverage of the latest available household surveys is increasing 50 45 43.10 population in developing 40 countries, 2009, % 35 30 25.09 25 20 15 10.90 10 7.65 4.66 4.46 5 1.50 0.19 0.08 1.13 0.11 0.02 0 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 Source: World Bank staff calculations based on household data from the PovcalNet database. GLOBAL MONITORING REPORT 2011 THE DIVERSITY OF MDG PROGRESS 15 the 1990s, Sub-Saharan Africa’s poverty has parity in secondary education, where 82 per- fallen steeply since 2000. cent of countries (14 countries) are on target. Middle-income countries—both lower South Asia is also closing the development and upper—generally exhibit the best per- gap. Its performance is encouraging for pri- formance. For example, 80 percent of coun- mary education completion, gender parity, tries in the upper-middle-income bracket (23 maternal mortality, and access to safe drink- countries) have achieved or are on track to ing water, although performance needs to achieve the extreme poverty target. Similarly, improve for extreme poverty and hunger. 72 percent of lower-middle-income coun- Eastern Europe shows important progress tries (34 countries) are reaching the target on extreme poverty (the region’s countries for gender parity in secondary education. In account for 32 percent of countries on tar- one area, however—gender parity in primary get). Advances are also significant for child education—progress in both low-income and and maternal mortality. Challenges remain, middle-income countries is substantial. however—particularly in Central Asia, Latin America and the Caribbean is show- where progress on extreme poverty, child and ing excellent results and leading in several maternal mortality, and access to safe drink- indicators: hunger, primary education com- ing water is relatively low. For instance, no pletion, gender parity in secondary educa- Central Asian country is on track to achieve tion, access to safe drinking water, and access the child mortality goal; but the target is still to sanitation. Even so, it faces serious chal- within reach or close to becoming on target lenges on maternal mortality, with just 10 per- for five countries in the subregion (see below cent of countries (three countries) that have and appendix 1 for a discussion of the con- reached or are on track to reach the 2015 tar- cept, close to becoming on target). get (figure 1.3). The Middle East and North Africa is per- East Asia and Pacific is progressing in forming relatively well on gender parity in sec- many areas, particularly on education, gender ondary education and on access to sanitation. parity, and access to safe drinking water. Its However, it needs faster progress on extreme performance is particularly good for gender poverty, hunger, and maternal mortality. FIGURE 1.3 Countries on target to achieve the MDGs, by region 100 proportion of developing countries 90 80 70 in each region, % 60 50 40 30 20 10 0 MDG 1.a MDG 1.c MDG 2.a MDG 3.a MDG 3.a MDG 4.a MDG 5.a MDG 7.c MDG 7.c extreme hunger primary gender gender child mortality maternal access to access to povery completion parity parity under five mortality safe water sanitation rate (primary) (secondary) East Asia and Pacific Europe and Central Asia Latin America and the Caribbean Middle East and North Africa South Asia Sub-Saharan Africa Source: World Bank staff calculations based on data from the World Development Indicators database. 16 THE DIVERSITY OF MDG PROGRESS GLOBAL MONITORING REPORT 2011 Sub-Saharan Africa lags the other regions access to sanitation, and primary education but can point to some encouraging results. completion require stepped-up efforts. Progress is quite good on extreme poverty (9 Several low-income countries are doing countries), hunger (8 countries), gender par- well. A look beneath the aggregate global ity in primary education (27 countries), and statistics shows not just middle-income access to safe drinking water (15 countries). countries doing well, but many low-income Goals related to child and maternal mortality, countries, too (table 1.1). A recent study by TABLE 1.1 Several low-income countries are achieving the MDGs Low-income countries that have Low-income countries that are on track to Selected MDG achieved the goal achieve the goal Poverty Cambodia Central African Republic Kenya Ethiopia Mauritania Ghana Universal primary education Myanmar None Tajikistan Tanzania Gender parity in primary Bangladesh Benin education Gambia, The Burkina Faso Ghana Burundi Haiti Cambodia Kenya Comoros Kyrgyz Republic Ethiopia Madagascar Guinea Malawi Nepal Mauritania Sierra Leone Myanmar Solomon Islands Rwanda Togo Tanzania Uganda Zambia Zimbabwe Gender parity in secondary Bangladesh Gambia, The education Kyrgyz Republic Malawi Myanmar Mauritania Nepal Rwanda Under-five mortality None Bangladesh Eritrea Lao PDR Madagascar Nepal Access to safe drinking water Afghanistan Benin Burkina Faso Cambodia Comoros Guinea Gambia, The Uganda Ghana Korea, Dem. People’s Rep. Kyrgyz Republic Malawi Nepal Access to sanitation Lao PDR Rwanda Myanmar Tajikistan Source: World Development Indicators database (as of March 2011). Note: List of low-income countries is based on the new World Bank classification for fiscal year 2011 (see appendix 1). GLOBAL MONITORING REPORT 2011 THE DIVERSITY OF MDG PROGRESS 17 Leo and Barmeier confirmed that progress in countries are attaining the goals, and how individual African and poor countries was many are behind? Are lagging countries far surprisingly strong.3 from the goals? And how many are already Variation behind the aggregates. The close? Answers from available information typical global and regional summaries amass are surprisingly hopeful. data for countries of dissimilar development To examine them, we distinguish coun- and types—fragile, low-income, and middle- tries that are on target (or on track)—that is, income countries. For example, the Europe their annual rate of progress between the ref- and Central Asia region covers such middle- erence year of MDGs in 1990 (or the closest income countries as Albania and Bulgaria available) to the latest year of data implies the and such low-income countries as Tajiki- right trajectory or trend to meet or exceed the stan and Uzbekistan. Among the developing goals—and those that are off target (or lag- countries in Sub-Saharan Africa, some are ging). We look fi rst at the gaps of off-target middle-income countries (such as Mauri- countries because they are the countries that tius and South Africa); some lower-middle- need most attention. income countries (such as Angola and the First, the variation among lagging coun- Democratic Republic of Congo) are resource tries is large, but the average gap is not rich, but their levels of development may (table 1.2). Lagging countries are, on average, be closer to those of low-income countries. only 23 percent away from being on track to Progress has been more heterogeneous than achieve all the MDGs. They are especially is shown by the aggregate figures.4 close to the targets for gender parity in pri- mary education (average gap is 7 percent); gender parity in secondary education (16 Country progress, not just global or percent gap); hunger (19 percent gap); pri- regional mary education completion (20 percent gap); Although the MDGs were conceived as global and, to some extent, under-five mortality (23 targets to spur development efforts and sup- percent gap). But for each target there are port to poor countries, it is necessary to mea- countries where progress has been scant. For sure and describe progress at the country or example, several countries are far from halv- other level to better understand advances and ing extreme poverty, even as the global goal remaining gaps (see box 1.2). To untangle will be reached. the aggregate numbers, we introduce alterna- Progress is mixed or poor on access to safe tive ways of analyzing MDG progress with drinking water, access to sanitation, mater- a focus on lagging countries. Our approach nal mortality, and extreme poverty. Even so, characterizes MDG progress by country per- the mean gaps of all lagging countries are less formance and by different typologies—such than 50 percent from the targets on access to as initial income and policy-institutional con- safe drinking water (25 percent) and access ditions, subsequent growth and policy-insti- to sanitation (27 percent), and no worse than tutional achievement, the poorest countries 40 percent on maternal mortality (32 percent) versus the others, level of fragility, and export and extreme poverty (39 percent). sophistication and shipping connectivity Different starting points will imply a (broadly following Collier and O’Connell5 ). unique trajectory for each country to reach This approach provides some empirical expla- a specific goal. Hence, comparing the slope nations of the links between development’s or growth rate of the historical path with the drivers and different rates of MDG progress. required one is a good way to assess progress. Leo and Barmeier define lagging countries as close to target if their trajectory is within 50 Where do countries stand? percent of the required progress to reach the Several global targets will be missed, but the goals, earning half a full score.6 In our meth- following questions also matter: How many odology, we do not assign numerical scores in 18 THE DIVERSITY OF MDG PROGRESS GLOBAL MONITORING REPORT 2011 BOX 1.2 Gaps and issues in measuring development outcomes beyond 2015 Measuring broad development outcomes through teachers are at stake, partly because the outcomes specific indicators is never precise, so the diversity in also depend on other factors beyond the control of MDG performance is partly the result of indicator teachers. Take Tanzania as an example. It received or measurement issues. Although the purpose of this an MDG award at the 2010 United Nations sum- report is not to focus on these issues, they are impor- mit for its rapid progress on the primary education tant for future deliberations of MDGs beyond 2015. completion rate, but the ultimate development goal Some of the issues are the following: of improving learning outcomes of Tanzanian chil- dren remains problematic.a The point is that some • Outcomes versus outputs. Some MDG indicators meaningful development outcomes rest outside the (such as under-five mortality and maternal mortal- scope of the MDGs, and it is important to begin ity) quantify development outcomes, whereas oth- to develop the auxiliary indicators and information ers (such as the primary education completion rate base to monitor them. and ratios of girls to boys in education) are specific • Missing targets. Some specific development out- intermediate outputs. It is not surprising that the comes are not defined or are missing. Economic more outcome-based health MDGs often progress growth, particularly inclusive growth, is also more slowly than others. But poverty reduction, a excluded from the MDGs, as noted by many com- broad development outcome, is progressing rap- mentators. Its centrality as a means to achieve the idly. Because of the inherent and varying difficulty MDGs is nonetheless monitored in the Global of demonstrating progress and results, choosing Monitoring Report framework. Missing global between outcomes and outputs may touch on issues public goods include climate change–related goals of political economy. For example, a target like uni- (such as reducing carbon emissions) and market versal primary school enrollment is easily embraced access for developing countries. because it is easier to show progress by getting chil- • Targets versus trends. Although Sub-Saharan dren through school than it is to make sure they countries as a group will fi nd it difficult to meet learn something, which may also be harder to mea- the poverty goal by 2015, they are progressing far sure. By contrast, learning outcomes as an objec- above historical trends. Therefore, deviations from tive are often resisted when the jobs and pay of the targets versus deviations from past trends also this manner or use an arbitrary cutoff point of More important, among countries that are 50 percent. Although we use the trend devia- off track, the top half are, on average, only tions or differences in the two growth rates to about 11 percent away from being on target. defi ne the gaps, the actual gaps are retained The mean distance of this subgroup is only to classify countries into groups according 4–9 percent for gender parity in primary and to their progress, to measure the mean gaps secondary education, child mortality, primary of each group, and to identify countries that education completion, and hunger. Indeed, are within 10 (or 20) percent of becoming on countries close to the target need to increase target. The mean gaps of lagging countries primary education completion only by 9.2 are all less than 50 percent across the MDGs, percent (or 1.5 percent a year), on average, to and they provide data-specific cutoff points be on track to reach the 2015 target. to split the off-target countries into two sub- However, the rather uneven distribution groups: above average and below average. also points to serious problems for the bot- Countries in the top half are “close to the tar- tom half of the off-target countries, those get,” whereas countries in the bottom half are far from the target among the lagging coun- “far from the target” (appendix 1 explains tries—they are disproportionately far from the approach in more detail). the targets, especially for poverty (67 percent) GLOBAL MONITORING REPORT 2011 THE DIVERSITY OF MDG PROGRESS 19 BOX 1.2 (continued) matter. A related, wider issue is whether the MDGs the hard-to-reach populations (chapter 4). The ratios should be absolute or relative targets.b of girls to boys in schools may be stagnant at 97–99 • Data gaps. Poor data availability may fail to con- percent in some higher-middle-income countries vey progress or understate deterioration resulting because of different enrollment ratios between girls from shocks and conflicts. Among the MDGs, the and boys. Boys may be lagging in completion rates collection and dissemination of poverty aggregates (for example, at the primary level in Uruguay and seem to lag the most. Conducting and complet- at the secondary level in the Russian Federation). ing household income and expenditure surveys in In addition, country figures can vary from interna- large countries is difficult, often delaying global or tional ones because of differences between national regional poverty updates because of the weights education systems and the International Standard of these countries in the aggregates. There is also of Education Classification (ISCED) used in multi- spotty information about infectious diseases, such lateral development agencies—as well as differences as malaria and tuberculosis. Other data issues in coverage and even population estimates. include reporting errors that severely compromise • Averages and weights. Large countries such as the accuracy of the maternal mortality ratio. And China, India, and Nigeria dominate their respec- hunger is assessed only indirectly, through mini- tive regional averages, especially on measures, such mum food intake and its deprivation. as poverty, where population is the weight. So the • Metrics issues in some MDGs. When targets are pattern would likely change if the reference unit measured in proportionate amounts, such as reduc- pertained to individual countries, not regional or ing by half the proportion of people living on global aggregates. For example, many poor coun- less than $1.25 a day, problems can arise at both tries, including Bangladesh, Bolivia, the Lao Peo- extremes. For low-income countries with high ini- ple’s Democratic Republic, Malawi, Mozambique, tial poverty rates, many of which are in Africa, the Nepal, and Niger are registering major achieve- greater distance to the goal makes the target harder ments on difficult MDGs such as child mortality. to reach; for middle-income countries where poverty rates are less than 10 or even 5 percent, reducing the a. Uwezo Tanzania 2010. rate further is also difficult and may entail assisting b. ODI 2010. TABLE 1.2 Lagging countries are close to getting on target Average distance to getting on target (gaps, %) All off-target countries Countries close to the target Countries far from the target MDG 1.a Extreme poverty 39 (96) 17 67 MDG 1.c Hunger 19 (60) 9 35 MDG 2.a Primary education completion 20 (96) 9 40 MDG 3.a Gender parity in primary education 7 (22) 4 14 MDG 3.a Gender parity in secondary education 16 (52) 8 29 MDG 4.a Child mortality under five 23 (59) 8 38 MDG 5.a Maternal mortality 32 (80) 11 51 MDG 7.c Access to safe drinking water 25 (76) 14 41 MDG 7.c Access to sanitation 27 (50) 16 34 Simple average 23 11 39 Source: World Bank staff calculations based on data from the World Development Indicators database. Note: A country is “close to the target” if its distance to getting on target (that is, its gap of trajectory) is smaller than the average gap of all lagging countries. Otherwise, it is “far from the target” (that is, its distance is greater than the average gap). Figures in parentheses indicate the range of variation (Maximum value—Minimum value) of countries off target, by MDG. Averages and numbers of countries cover only those with data—and that may vary by MDG. See appendix 1 for more details. 20 THE DIVERSITY OF MDG PROGRESS GLOBAL MONITORING REPORT 2011 TABLE 1.3 Many countries are within 10–20 percent of being on target Distribution of lagging countries Gap ≤ 10 percent Gap ≤ 20 percent Number of Proportion of Number of Proportion of countries countries (%) countries countries (%) MDG 1.a Extreme poverty 9 24 13 34 MDG 1.c Hunger 10 33 18 60 MDG 2.a Primary education completion 23 40 39 68 MDG 3.a Gender parity in primary education 28 74 36 95 MDG 3.a Gender parity in secondary education 16 42 23 61 MDG 4.a Child mortality under five 33 31 48 46 MDG 5.a Maternal mortality 20 21 37 39 MDG 7.c Access to safe drinking water 10 15 32 48 MDG 7.c Access to sanitation 6 6 25 26 Simple average 17 32 30 53 Source: World Bank staff calculations based on data from the World Development Indicators database. and maternal mortality (51 percent). And or 20 percent of getting on target. A third of the range of variation is considerably large off-target countries have, on average, a gap among countries off target. For extreme pov- of 10 percent or less from being on target erty and primary education completion, the across the MDGs. Countries like Bangladesh gap between the countries closest to and far- (extreme poverty, hunger, and maternal mor- thest from being on target is 96 percent, a fact tality), Indonesia (hunger, child and maternal that clearly illustrates the diversity of perfor- mortality, access to safe drinking water), and mance. This is the case for El Salvador and Mali (gender parity in primary education and Uzbekistan on extreme poverty reduction and access to safe drinking water) are in this cate- for Bhutan and Djibouti on primary comple- gory. (Table 1.4 lists these countries by MDG.) tion rates. More than half have a gap of 20 percent or Indeed, many lagging countries are less. Of the countries within 20 percent of tar- already within striking distance. From get, the best results are for gender parity in another perspective, table 1.3 provides the primary education, primary education com- proportion of countries within 10 percent pletion, gender parity in secondary education, TABLE 1.4 Lagging countries within 10 percent of being on target to achieve the MDGs MDG 2.a MDG 3.a MDG 3.a MDG 4.a MDG 1.a Primary Gender parity Gender parity Child MDG 5.a MDG 7.c MDG 7.c Extreme MDG 1.c education in primary in secondary mortality Maternal Access to safe Access to poverty Hunger completion education education under five mortality drinking water sanitation Bangladesh Bangladesh Bhutan Belize Bulgaria Algeria Algeria Azerbaijan Botswana Burkina Faso Bolivia Cambodia Cape Verde Congo, Rep. Antigua and Bangladesh Colombia Brazil Barbuda El Salvador Egypt, Arab Comoros Chile Georgia Argentina Brazil Eritrea Dominican Rep. Republic Guinea Indonesia Cuba Congo, Dem. Grenada Belarus Cambodia Haiti Morocco Rep. India Jordan El Salvador Congo, Rep. Guatemala Bhutan Cape Verde Indonesia Peru Lao PDR Kenya Gambia, The Djibouti Macedonia, Cape Verde Dominican Iran, Islamic Turkey FYR Republic Rep. Lesotho Nigeria Ghana El Salvador Madagascar Colombia Egypt, Arab Kiribati Rep. Philippines Pakistan Guatemala Grenada Morocco Dominican Ethiopia Mali Republic Uganda Rwanda Honduras Guatemala Pakistan Ecuador Haiti Myanmar GLOBAL MONITORING REPORT 2011 THE DIVERSITY OF MDG PROGRESS 21 TABLE 1.4 Lagging countries within 10 percent of being on target to achieve the MDGs (continued) MDG 2.a MDG 3.a MDG 3.a MDG 4.a MDG 1.a Primary Gender parity Gender parity Child MDG 5.a MDG 7.c MDG 7.c Extreme MDG 1.c education in primary in secondary mortality Maternal Access to safe Access to poverty Hunger completion education education under five mortality drinking water sanitation Zambia Iraq Guinea- Russian Ethiopia India Venezuela, Bissau Federation R.B. de Jamaica Jamaica Senegal Guatemala Indonesia Kyrgyz Lao PDR Solomon Honduras Lao PDR Republic Islands Lebanon Lebanon Sudan Indonesia Mongolia Lithuania Maldives Swaziland Kazakhstan Morocco Macedonia, Mali Vanuatu Kiribati Nepal FYR Mauritius Mozambique Zimbabwe Kyrgyz Peru Republic Moldova Nigeria Liberia Rwanda Morocco Paraguay Libya Syrian Arab Republic Nepal South Africa Malawi Tunisia Philippines St. Vincent Moldova Yemen, Rep. and the Grenadines South Africa Sudan Montenegro Tanzania Suriname Niger Turkey Swaziland Paraguay Tajikistan Russian Federation Tonga Samoa Uruguay Sri Lanka Vanuatu St. Vincent and the Grenadines Venezuela, Suriname R.B. de Syrian Arab Republic Tajikistan Turkmenistan Uzbekistan Yemen, Rep. Source: World Bank staff calculations based on data from the World Development Indicators database. Note: See also box 1.2, which covers metrics issues concerning some MDG targets. and hunger. The worst results are for access than a decade of better policy and growth to sanitation, extreme poverty, and maternal (figure 1.4). Many countries are making sub- mortality, with access to safe drinking water stantial progress in several MDGs: gender and under-five mortality in the middle. parity in primary education (89 of them), All told, developing countries have gender parity in secondary education (82), been doing much better than thought until access to safe drinking water (66), primary recently. Although many more develop- completion rate (55), and extreme poverty ing countries are off track than on track to (47). For instance, about 70 percent of devel- achieve the targets, two thirds or more of oping countries have achieved or are on track developing countries are actually on target to achieve the targets for gender parity in or close to being on target, thanks to more primary and secondary education. Although 22 THE DIVERSITY OF MDG PROGRESS GLOBAL MONITORING REPORT 2011 FIGURE 1.4 More than two-thirds of developing countries are on track or close to being on track 80 proportion of developing countries, % 89 82 70 60 47 55 66 50 25 58 40 52 53 46 48 18 38 39 30 21 36 35 37 12 30 17 25 23 27 20 19 13 15 10 0 MDG 1.a MDG 1.c MDG 2.a MDG 3.a MDG 3.a MDG 4.a MDG 5.a MDG 7.c MDG 7.c extreme hunger primary gender gender child mortality maternal access to access to povery completion parity parity under five mortality safe water sanitation rate (primary) (secondary) on target close to target far from target Source: World Bank staff calculations based on data from the World Development Indicators database. Note: A country is “close to the target” if its distance to getting on target (that is, its gap of trajectory) is smaller than the average gap of all lagging countries. Otherwise, it is “far from the target” (that is, its distance is greater than the average gap). The number above each bar indicates the number of countries. See appendix 1 for more details. half the monitored countries (57) are off tar- water (69 percent of countries are close to the get for the primary education completion target). goal, two-thirds of them (38) are very close Many South Asian countries that face dif- to being on track. ficulties in reaching the 2015 targets are also Progress is mixed or poor on access to performing better than average in the off- sanitation, maternal mortality, and child target group. For instance, all the region’s coun- mortality (box 1.3). Unfortunately, more tries are within reach of the extreme poverty than 40 percent of low-income to upper- target, and 80 percent are close to the access to middle-income countries in the sample (58 sanitation goal. For 67 percent of South Asian countries) are significantly off target for countries, the hunger target is within range. access to sanitation. In the Middle East and North Africa, 67 Regional patterns vary. The maps in percent of countries are close to the target this report show global and regional pat- for gender parity in primary education. Even terns (see map 1.1). Because the gaps in the more encouraging: 80 percent of countries in top half of lagging countries are small, the the region are within reach of the primary ratio of countries close to the target to the education goal. total number of off-target countries also pro- In Latin America and the Caribbean, vides a relative measure of progress among Europe and Central Asia, and East Asia and MDG underachievers (table A1.2 and related Pacific—the regions with better MDG perfor- tables in appendix 1). Within this group, Sub- mance—most off-target countries are close to Saharan Africa shows encouraging results. several goals, such as hunger, primary educa- For instance, 81 percent of the region’s lag- tion completion, and gender parity. Results ging countries are relatively close to the are less promising for maternal mortality and extreme poverty target. Fifty percent of the access to sanitation. Interestingly, for coun- region’s countries are very close to the target tries that have low poverty rates (that is, less for gender parity in secondary education, and than 10 percent), reducing extreme poverty at 53 percent are quite close for hunger. Prog- $1.25 a day to a much greater extent may not ress is also evident in access to safe drinking be easy (see box 1.2). GLOBAL MONITORING REPORT 2011 THE DIVERSITY OF MDG PROGRESS 23 BOX 1.3 Improving children’s health through sustainable access to food, water, and energy Sustainable access to food, safe drinking water, basic provision of food, water, and energy becomes more sanitation, and modern energy sources can decrease difficult when natural resources are not properly child mortality significantly.a Of the estimated 10.5 managed or degrade as a result of global environ- million child deaths annually, the vast majority are mental change.c For example, climate change induces from preventable and treatable diseases and condi- changes in rainfall and temperature patterns, poten- tions, including low dietary energy consumption tially increasing the likelihood of short-term crop (underweight), unsafe drinking water and the lack failures and long-term production declines as well of basic sanitation (diarrhea), and indoor air pollu- as deterioration in water quality. The most vulner- tion related to solid fuel use for cooking and heating able are poor and food-insecure countries at lower (pneumonia). latitudes (especially in seasonally dry and tropical Based on socioeconomic and environmental regions) that largely depend on rainfed farming— trends, we project the population that lacks adequate again in Sub-Saharan Africa and South Asia. Many access to food, water, and energy, and the resulting such pressures are slow moving and cannot easily be effect on child mortality. Except for Middle East and stopped because of major inertia, including the pres- North Africa, progress is not enough to reach the sures of fertility transition and greenhouse gas accu- MDG target on child mortality; and Sub-Saharan mulation. They become apparent only in the long Africa does not even come close by 2030. Acceler- term, after 2015 or even after 2030, trapping people ated progress significantly improves this situation (see in their poverty and reversing progress. the figure below). This is especially the case in Sub- Therefore, to achieve the MDG target on child Saharan Africa, where approximately one-third of mortality, accelerated growth in the other MDGs is the child mortality gap can be achieved by achieving key—particularly access to food, improved drinking other MDGs. water, basic sanitation, and improved energy sources. Rising demand for food, water, and modern Furthermore, policies addressing increased access to energy will put pressure on scarce natural resources food, water, and energy should take global scarcities (for example, fertile land for food and bioenergy; oil and global environmental changes into account. This resources for clean fuels, such as kerosene for cook- could both help in achieving the MDGs and in mak- ing). This will increase the prices of (especially) food ing those achievements more sustainable. and energy. And it will hurt poor people in import- ing countries in Sub-Saharan Africa and South Asia a. PBL 2009b. whose governments are unable to guarantee afford- b. PBL 2011. able prices when global prices increase.b Furthermore, c. PBL 2009b. Projected child mortality results from various causes, 2015 and 2030 deaths per 1,000 live births 140 120 100 80 60 40 20 0 2015 2015a 2030 2015 2015a 2030 2015 2015a 2030 2015 2015a 2030 2015 2015a 2030 Latin America Middle East Sub-Saharan East Asia and the Caribbean and North Africa Africa South Asia and Pacific water supply and sanitation indoor air pollution child underweight Malaria other causes MDG target Source: PBL 2009a. Note: Because being underweight is usually not a direct cause of a child’s death (although it increases the risks of dying from pneumonia, malaria, and diarrhea), it is reported separately. a. The accelerated progress scenario for 2015 achieves the MDG targets on food, water, malaria, and energy (GISMO model, PBL 2008). 24 THE DIVERSITY OF MDG PROGRESS GLOBAL MONITORING REPORT 2011 Country diversity generally softens the What important lessons therefore emerge gloomy global picture. All these statistics are from the complex tapestry of progress? Why remarkable, revealing progress that is much are some countries on target, but others are more diversified and much more hopeful not? Of the lagging countries, why are some than the recent pessimism about achieving close to target and others far away? The two the MDGs. That pessimism was likely col- main drivers often cited as key to attaining ored by the gaps at the global level, the dif- MDG-related development outcomes are eco- ficult circumstances of poor countries, the nomic growth and sound policies and institu- potential negative impact of the recent global tions (fundamental to effective service deliv- crisis, and the lack of recent data to assess ery to the poor).7 Although it is easy to cite outcomes. For example, although only 27 these two drivers, it is hard to provide empiri- percent of low-income countries are on track cal substance to their impact on achieving the to achieve or have achieved the extreme pov- MDGs. erty target, almost 90 percent of these coun- We pursue this tack in the next section tries with data are in the top half of the lag- by examining the country pattern against ging group and, therefore, have the poverty growth and policy accomplishments, con- goal within their reach. Similarly, around 40 tinuing the forward-looking analytical work percent of low-income countries are close to started in the 2010 Global Monitoring the primary education completion goal, even Report.8 More specifically, we ask whether though only 7 percent of the countries in this initial conditions or subsequent growth and income group are on target. policy improve the odds of reaching the goals. What the data also suggest is that the ref- The analysis looks at these elements in two erence unit matters. Simple country averages ways: using prima facie evidence from graph- that give equal importance to each country ical associations and patterns, which point qualify the global story, which uses weighted to these elements’ likely association with the averages that give more importance to coun- diverse progress of countries; and using a sta- tries with large populations. This can work tistical investigation of their significance in in both directions. For example, the progress increasing the likelihood of attaining MDG- in reducing world poverty and in meeting the associated outcomes. goal is essentially the result of rapid advances by China and India, with the absolute num- ber of poor people falling rapidly in China. The role of growth and policy Too many countries, however, still lag on Initial conditions count in MDG perfor- the poverty goal, and their average shortfall mance, but subsequent growth and policy of 39 percent to be on target is the biggest also matter greatly—or more. In most cases, among the MDGs. In contrast, the average countries that are doing better (those on or distance to becoming on target for under-five close to the target) exhibited favorable start- mortality is only 23 percent for lagging coun- ing conditions around 1990 (the reference tries, somewhat less daunting than the global year). A higher per capita GDP in 1990 is distance derived from the population of all generally associated with better MDG per- under-five children. formance (figure 1.5). Country analysis generally brightens the Although there is no perfect indicator of global picture, but both global and country the overall quality of policy and institutions perspectives are necessary for the complete in developing countries, the World Bank’s view. In an unprecedented manner, the MDGs annual Country Policy and Institutional as global targets have galvanized development Assessment (CPIA) provides a broadly consis- efforts to help the world’s poor; however, the tent framework for assessing country perfor- country unit is relevant because these efforts mance on 16 items grouped in four clusters: are still geared to individual countries, given economic management, structural policies, that country-owned strategy and capacity are policies for social inclusion and equity, and important aspects of development assistance. public sector management and institutions. GLOBAL MONITORING REPORT 2011 THE DIVERSITY OF MDG PROGRESS 25 MAP 1.1 The world is still on track to meet the poverty reduction target IBRD 38444 APRIL 2011 Poverty rate Greenland (Den) Faeroe Norway Iceland Islands Finland (Den) Sweden The Netherlands Russian Federation Estonia Canada Isle of Man (UK) Denmark Russian Latvia Fed. Lithuania United Ireland Kingdom Germany Poland Belarus Channel Islands (UK) Belgium Ukraine Luxembourg Moldova Kazakhstan Liechtenstein Romania Mongolia Switzerland France Italy Andorra Bulgaria Georgia Azer- Uzbekistan Kyrgyz D.P.R. Spain Turkey Armenia baijan Turkmenistan Rep. of Korea United States Portugal Monaco Tajikistan Japan Greece Syrian Gibraltar (UK) Cyprus Rep. of Tunisia Malta Lebanon Arab Islamic Rep. Afghanistan China Korea Israel Rep. Iraq of Iran Bermuda Morocco Kuwait (UK) West Bank and Gaza Jordan Bahrain Pakistan Nepal Bhutan Former Algeria Libya Arab Rep. Qatar The Bahamas Spanish of Egypt Saudi Cayman Is. (UK) Sahara Arabia United Arab Bangladesh Cuba Turks and Caicos Islands (UK) India Mexico Emirates Myanmar Lao Haiti Mauritania Oman P.D.R. Belize Jamaica Cape Verde Mali N. Mariana Islands (US) Niger Rep. of Thailand Vietnam Guatemala Honduras Senegal Chad Eritrea Yemen Guam (US) El Salvador Nicaragua The Gambia Burkina Sudan Marshall Faso Djibouti Cambodia Philippines Islands Guinea-Bissau Guinea Benin Federated States of Costa Rica Panama Sri Micronesia R.B. de Guyana Sierra Leone Côte Ghana Nigeria Central Ethiopia Lanka Venezuela Suriname Liberia D’Ivoire African Rep. Brunei Cameroon Somalia Palau Colombia French Guiana (Fr) Togo Malaysia Equatorial Guinea Uganda Maldives Ecuador São Tomé and Príncipe Gabon Congo Rwanda Kenya Singapore Nauru Dem. Rep. of Burundi Seychelles Papua Solomon Congo Indonesia New Guinea Islands Tanzania Comoros Peru Brazil Timor-Leste Angola Malawi Mayotte Zambia (Fr) Vanuatu Bolivia Zimbabwe Mauritius Namibia Madagascar Paraguay Botswana Mozambique Réunion (Fr) New Caledonia Australia (Fr) Swaziland South Lesotho Africa Chile Uruguay Argentina Dominican Puerto British Virgin Poland Poverty rate New Zealand Germany Republic Rico (US) Islands (UK) Share of population living on less than $1.25 a day Ukraine Anguilla (UK) Czech Rep. U.S. Virgin Islands Antigua and Barbuda Slovak Rep. (US) St. Maarten (Neth) St. Kitts Guadeloupe (Fr) Austria Hungary ≥50 and Nevis Slovenia 25–49.9 Dominica Romania Curaçao Montserrat (UK) Martinique (Fr) Croatia (Neth) St. Lucia Bosnia and 10–24.9 Aruba Bonaire St. Vincent and San Herz. Serbia Bulgaria (Neth) (Neth) The Grenadines Barbados Marino Kosovo 2–9.9 Montenegro Grenada FYR ≤2 Italy Macedonia Trinidad Vatican Albania R.B. de Venezuela and Tobago City Greece no data MDG performance Greenland (Den) Faeroe Norway Iceland Islands Finland (Den) Sweden The Netherlands Russian Federation Estonia Canada Isle of Man (UK) Denmark Russian Latvia Fed. Lithuania United Ireland Kingdom Germany Poland Belarus Channel Islands (UK) Belgium Ukraine Luxembourg Moldova Kazakhstan Liechtenstein Romania Mongolia Switzerland France Italy Andorra Bulgaria Georgia Azer- Uzbekistan Kyrgyz D.P.R. Armenia baijan Turkmenistan Rep. United States Portugal Spain Monaco Turkey Tajikistan of Korea Japan Greece Syrian Gibraltar (UK) Cyprus Rep. of Tunisia Malta Lebanon Arab Islamic Rep. Afghanistan China Korea Israel Rep. Iraq of Iran Bermuda Morocco Kuwait (UK) West Bank and Gaza Jordan Pakistan Bhutan Bahrain Nepal Former Algeria Libya Arab Rep. Qatar The Bahamas Spanish of Egypt Saudi Cayman Is. (UK) Sahara Arabia United Arab Bangladesh Cuba Turks and Caicos Islands (UK) India Mexico Emirates Myanmar Lao Haiti Mauritania Oman P.D.R. Belize Jamaica Cape Verde Mali N. Mariana Islands (US) Niger Rep. of Thailand Vietnam Guatemala Honduras Senegal Chad Eritrea Yemen Guam (US) El Salvador Nicaragua The Gambia Burkina Sudan Marshall Faso Djibouti Cambodia Philippines Islands Guinea-Bissau Guinea Benin Federated States of Costa Rica Panama Sri Micronesia R.B. de Guyana Sierra Leone Côte Ghana Nigeria Central Ethiopia Lanka Venezuela Suriname Liberia D’Ivoire African Rep. Brunei Cameroon Somalia Palau Colombia French Guiana (Fr) Togo Malaysia Equatorial Guinea Uganda Maldives Ecuador São Tomé and Príncipe Gabon Congo Rwanda Kenya Singapore Nauru Dem. Rep. of Burundi Seychelles Papua Solomon Congo Indonesia New Guinea Islands Tanzania Comoros Peru Brazil Timor-Leste Angola Zambia Malawi Mayotte (Fr) Vanuatu Bolivia Zimbabwe Mauritius Namibia Madagascar Paraguay Botswana Mozambique Réunion (Fr) New Caledonia Australia (Fr) Swaziland South Lesotho Africa Chile Uruguay Argentina Dominican Puerto British Virgin Poland MDG performance New Zealand Germany Republic Rico (US) Islands (UK) MDG 1.A Halve between 1990 and 2015, the proportion Ukraine Anguilla (UK) Czech Rep. U.S. Virgin Islands Antigua and Barbuda Slovak Rep. of people whose income is less than $1.25 a day (US) Austria St. Maarten (Neth) St. Kitts Guadeloupe (Fr) Hungary and Nevis Slovenia on target Dominica Romania Curaçao Montserrat (UK) Martinique (Fr) Croatia close to target (Neth) Bosnia and Bonaire St. Vincent and St. Lucia San Herz. Serbia Aruba far from target Bulgaria (Neth) (Neth) The Grenadines Barbados Marino This map was produced by the Map Design Unit of The World Bank. Montenegro Kosovo no data The boundaries, colors, denominations and any other information Grenada FYR Italy Macedonia shown on this map do not imply, on the part of The World Bank Trinidad Vatican Albania Group, any judgment on the legal status of any territory, or any and Tobago City Greece R.B. de Venezuela endorsement or acceptance of such boundaries. Source: World Bank staff calculations based on data from the World Development Indicators database. 26 THE DIVERSITY OF MDG PROGRESS GLOBAL MONITORING REPORT 2011 FIGURE 1.5 MDG performance is stronger in countries with good initial conditions a. Initial income conditions 7,000 6,000 GDP per capita, 1990 5,000 4,000 3,000 2,000 1,000 0 MDG 1.a MDG 1.c MDG 2.a MDG 3.a MDG 3.a MDG 4.a MDG 5.a MDG 7.c MDG 7.c extreme hunger primary gender gender child mortality maternal access to access to povery completion parity parity under five mortality safe water sanitation rate (primary) (secondary) on target close to target far from target b. Initial institutional conditions 3.60 3.40 3.20 CPIA index, 1996 3.00 2.80 2.60 2.40 2.20 2.00 MDG 1.a MDG 1.c MDG 2.a MDG 3.a MDG 3.a MDG 4.a MDG 5.a MDG 7.c MDG 7.c extreme hunger primary gender gender child mortality maternal access to access to povery completion parity parity under five mortality safe water sanitation rate (primary) (secondary) on target close to target far from target Source: World Bank staff calculations based on data from the World Development Indicators database. Note: A country is “close to the target” if its distance to getting on target (that is, its gap of trajectory) is smaller than the average gap of all lagging countries. Otherwise, it is “far from the target” (that is, its distance is greater than the average gap). The score is from 1 (low) to 6 (high) for each that countries starting with good policies and policy that covers a wide range of issues.9 institutions tend to do better in the MDGs. The index focuses on policies and institu- Starting points—inherited initial condi- tional arrangements—the key elements that tions—explain why middle-income countries are within the country’s control—rather than generally do better than low-income coun- on actual outcomes (for example, growth tries. Having grown earlier, they also tend to rates) that are influenced by elements outside have implemented earlier a better set of poli- the country’s control. Over time, good poli- cies and institutions. But there are variations. cies and institutions are expected to lead to For extreme poverty and gender parity in pri- favorable growth and poverty reduction out- mary education, countries making the fastest comes, notwithstanding possible yearly fluc- progress are those that experienced medium tuations caused by external factors.10 Using poverty and female-to-male primary enroll- the 1996 CPIA, the earliest index with com- ment ratios in the 1990s (see table A1.5 in the parable scale and criteria available,11 suggests appendix). The latter results draw attention GLOBAL MONITORING REPORT 2011 THE DIVERSITY OF MDG PROGRESS 27 FIGURE 1.6 Subsequent economic growth and policy seem to matter more a. MDG performance and growth in GPD per capita growth in GDP per capita, 1990–2009, % 3.50 3.00 2.50 2.00 1.50 1.00 0.50 0 MDG 1.a MDG 1.c MDG 2.a MDG 3.a MDG 3.a MDG 4.a MDG 5.a MDG 7.c MDG 7.c extreme hunger primary gender gender child mortality maternal access to access to povery completion parity parity under five mortality safe water sanitation rate (primary) (secondary) on target close to target far from target b. MDG performance and recent institutional conditions 3.95 3.75 CPIA index, 2009 3.55 3.35 3.15 2.95 2.75 MDG 1.a MDG 1.c MDG 2.a MDG 3.a MDG 3.a MDG 4.a MDG 5.a MDG 7.c MDG 7.c extreme hunger primary gender gender child mortality maternal access to access to povery completion parity parity under five mortality safe water sanitation rate (primary) (secondary) on target close to target far from target Source: World Bank staff calculations based on data from the World Development Indicators database. Note: A country is “close to the target” if its distance to getting on target (that is, its gap of trajectory) is smaller than the average gap of all lagging countries. Otherwise, it is “far from the target” (that is, its distance is greater than the average gap). to the challenges of poverty reduction in the track to reach the targets (excluding gender proportionate way that MDGs are defined parity in primary education) show the fast- at low-income and middle-income levels— est per capita GDP growth over 1990–2009. for poor countries, the distance to the goal In the same way, countries close to the target is long; for middle-income countries, halving tend to have grown faster, in per capita terms, already low poverty rates is not easy. than countries far from the target. Likewise, a So, although starting points (given their strong policy and institutional framework in inherited nature) do not say much about what the most recent year, 2009, tends to facilitate countries can or should do, they need not pre- service delivery to the poor and to improve ordain outcomes. The good news is that eco- MDG performance (table 1.5). nomic growth and policy performance after Both factors—initial conditions and sub- the initial year appear to count greatly, if not sequent growth and policy—also point to more than the starting points (figure 1.6). On why the MDGs are such big challenges for average, countries that have reached or are on the poorest and most fragile countries. The 28 THE DIVERSITY OF MDG PROGRESS GLOBAL MONITORING REPORT 2011 TABLE 1.5 Growth and CPIA scores are higher in countries on track or close to being on track Average values across MDGs (weighted by the number of countries in each MDG category) Close to Far from On target the target the target Average GDP per capita growth (1990–2009) 2.4 1.8 1.2 CPIA index (2009) 3.7 3.5 3.3 Source: World Bank staff calculations based on data from the World Development Indicators database. Note: The pairwise correlation between average GDP per capita growth and the CPIA index is 0.32 (significant at 0.01 level). GDP per capita, purchasing power parity constant 2005 international dollars. A country is “close to the target” if its distance to getting on target (that is, its gap of trajectory) is smaller than the average gap of all lagging countries. Otherwise, it is “far from the target” (that is, its distance is greater than the average gap). world’s 79 poorest countries serviced by the performance in developing countries. Coun- World Bank’s International Development tries with a higher level of export sophistica- Association (IDA) have a threshold per capita tion are generally more on track to achieve gross national income of $1,165 for fiscal year the targets (figure 1.8a). Likewise, countries 2011, with average per capita growth and more integrated into global shipping net- recent institutional performances well below works are more likely to be on or close to the average.12 Half the IDA countries are in Sub- target (figure 1.8b). Saharan Africa. With lower incomes and a We also looked at simpler dimensions of late start in policy reforms and growth, IDA trade—commodity versus noncommodity countries’ MDG performance tends to lag exporters as well as landlocked versus other that of middle-income and non-IDA countries countries—but the associations tend to be less (figure 1.7a). Despite the greater distance to defined than those shown in figure 1.8. In any the MDGs set by low starting points, the pov- case, export sophistication, shipping connec- erty target is within reach for more than 70 tivity, and state fragility are likely to be cor- percent of IDA countries as a result of more related with a country’s level of development, recent economic growth and policy improve- growth performance, infrastructure, and with ment. That is also true of the hunger target for its policies and institutions for trade, private 58 percent of IDA countries. Results are also sector development, and doing business.15 good for gender parity in primary education. Fragile conditions in conflict-affected Assessing the odds of achieving the countries are also associated with very poor MDGs MDG performance because these countries may experience growth collapses and disas- Is it possible to link and simulate the impact trous policy and institutional environments of growth and policy to the likelihood of (see box 1.4).13 In broad terms, the propor- achieving the MDGs in a manner more rigor- tion of on-target countries tends to rise with ous and statistical than with graphical asso- declining state fragility (figure 1.7b). Fragility ciations? Although formal econometric analy- in the graph is the index from the Center for sis, in principle, can isolate partial effects that Global Policy, which ranges from 0 (no fra- are not apparent from the simple correlations gility) to 25 (high fragility), divided into four in the previous section, there are caveats: the categories ranging from little to extreme fra- direction of impact between development out- gility (for more information on fragile states comes as measured by the MDGs and the two and the policy toolbox, see box 5.4).14 basic drivers (growth and policy) can go both Export sophistication and shipping con- ways; the two drivers themselves are likely to nectivity also have positive associations be correlated; and some factors that affect the with MDG performance. In a period of rapid progress of MDGs are not readily measur- globalization and trade expansion, these two able and available. Data constraints are also trade dimensions tend to be related to MDG problematic. For these reasons, the fi ndings GLOBAL MONITORING REPORT 2011 THE DIVERSITY OF MDG PROGRESS 29 FIGURE 1.7 MDG performance lags in IDA and fragile countries a. IDA countries on target 100 proportion of developing countries, % 90 57 80 53 35 46 70 36 43 60 12 25 50 13 23 40 12 25 25 26 30 9 10 20 10 10 5 0 MDG 1.a MDG 1.c MDG 2.a MDG 3.a MDG 3.a MDG 4.a MDG 5.a MDG 7.c MDG 7.c extreme hunger primary gender gender child mortality maternal access to access to povery completion parity parity under five mortality safe water sanitation rate (primary) (secondary) non-IDA countries IDA countries b. Level of fragility in countries on target 100 30 proportion of developing countries 90 24 25 by level of state fragility, % 80 20 21 18 24 7 70 8 19 14 16 60 18 18 12 13 50 9 8 8 11 13 40 5 9 8 8 30 6 6 4 5 5 20 4 4 10 1 1 0 MDG 1.a MDG 1.c MDG 2.a MDG 3.a MDG 3.a MDG 4.a MDG 5.a MDG 7.c MDG 7.c extreme hunger primary gender gender child mortality maternal access to access to povery completion parity parity under five mortality safe water sanitation rate (primary) (secondary) little to low moderate serious high to extreme Source: World Bank staff calculations based on data from the World Development Indicators database and Marshall and Cole 2010. Note: The number above each bar indicates the number of countries. See appendix 1 for more details. in this section are preliminary; they are spe- falling into one of the three defined categories, cific to the approach and presentation of data linking performance to the two drivers. For a taken; and they may not apply when using given development indicator associated with other approaches or treatment of the MDG each MDG, the likelihood of a country being variables. on target, close to the target, or far from the With these caveats and building on the target is expressed as a function of empirical patterns, previously defi ned mea- sures of MDG progress, and the basic driv- • economic growth (annual per capita GDP ers of progress in the GMR framework,16 we growth, 1990–2009); introduce a simple and intuitive model that is • recent quality of the policy and institutional suited to assessing the probability of a country framework approximated by the current 30 THE DIVERSITY OF MDG PROGRESS GLOBAL MONITORING REPORT 2011 BOX 1.4 The impact of violence on the MDGs The World Development Report 2011: Conflict, to lack access to basic health services and safe drink- Security, and Development depicts the arrested social ing water. Children born in these countries also tend development in countries affected by violence. Devel- to miss out on schooling; they are twice as likely to opment in these countries is lagging on nearly every be undernourished and nearly twice as likely to die MDG (top figure below). before age 5 (bottom figure below). People in fragile and confl ict-affected states are more liable to be impoverished and malnourished and Violence poses a major challenge to meeting the MDGs 3.5 3.0 countries not affected by violence incidence ratio relative to 2.5 2.0 1.5 1.0 0.5 0.0 under- impoverished primary secondary infant child deaths births HIV/AIDS access to access to nourished (not enrolled) (not enrolled) deaths under five unattended improved improved water sanitation fragile and conflict-affected states (FCS) recovering from fragility non-FCS (including Brazil, Russian Federation, India, and China) Countries affected by violence account for: 60% undernourished 60 61% impoverished 61 77% children not in primary school 77 59% children not in 59 secondary school 70% infant deaths 70 71% child deaths under five 71 64% births unattended 64 43% persons living with 43 HIV/AIDS 65% people without access 65 to safe water 54% people without access 54 to sanitation 0 50 100 percent GLOBAL MONITORING REPORT 2011 THE DIVERSITY OF MDG PROGRESS 31 FIGURE 1.8 MDG performance is better in countries with greater export sophistication and shipping connectivity a. Level of export sophistication 2 10 100 17 proportion of developing countries, % 90 16 9 8 80 70 13 5 8 19 16 15 60 9 5 13 12 14 50 8 10 5 14 40 5 6 3 4 8 9 8 9 9 30 6 6 20 5 4 2 2 10 0 MDG 1.a MDG 1.c MDG 2.a MDG 3.a MDG 3.a MDG 4.a MDG 5.a MDG 7.c MDG 7.c extreme hunger primary gender gender child mortality maternal access to access to povery completion parity parity under five mortality safe water sanitation rate (primary) (secondary) high medium-high medium-low low b. Level of shipping connectivity 30 liner shipping connectivity index 25 20 15 10 5 0 MDG 1.a MDG 1.c MDG 2.a MDG 3.a MDG 3.a MDG 4.a MDG 5.a MDG 7.c MDG 7.c extreme hunger primary gender gender child mortality maternal access to access to povery completion parity parity under five mortality safe water sanitation rate (primary) (secondary) on target close to target far from target Source: World Bank staff calculations based on data from the World Development Indicators database, Lall 2000 for export sophistication, and UNCTAD 2010 for liner shipping connectivity. Note: The index of export sophistication measures the technological content of exports from developing countries, standardized and distributed into four export groups. The ship- ping connectivity index has a base year of 2004. The number above each bar indicates the number of countries. A country is “close to the target” if its distance to getting on target (that is, its gap of trajectory) is smaller than the average gap of all lagging countries. Otherwise, it is “far from the target” (that is, its distance is greater than the average gap). See appendix 1 for more details. CPIA, which assesses recent changes in using the multinomial logit model. Estima- policies and institutions and, by design, tions are performed for each of the nine does not correlate with recent growth17; development targets under consideration • initial conditions (per capita GDP in 1990 using “far from the target” as the reference and CPIA index in 1996); and group or base category (see appendix 1 for a • controls (specific development indicators technical discussion). The statistical analy- around 1990). sis therefore pools all country information and focuses on the probability of a country The probability function across the differ- being in one of the three states of MDG ent states of MDG performance is estimated performance. 32 THE DIVERSITY OF MDG PROGRESS GLOBAL MONITORING REPORT 2011 Most of the literature on the determi- contrast, policy improvements as defi ned by nants of MDGs, at country or regional the CPIA cover myriad areas and interventions level, focuses on demand-side factors (such that need a longer time to come through. In as income and growth, demographic char- any case, given the short time left until 2015, acteristics, and cultural values and prefer- the statistical results confi rm the centrality ences) and on supply-side interventions (such of growth in improving countries’ odds of as public social expenditures, infrastructure, achieving the MDGs. (In addition to growth, institutional quality, and civil service perfor- reducing inequality also helps decrease pov- mance). Accordingly, empirical cross-country erty in the case of Brazil [box 1.5]). analyses usually relate supply and demand factors to development indicators in levels.18 Improving the odds of achieving the In the 2010 Global Monitoring Report, the MDGs scenarios of the impact of the recent crisis on the MDGs also used such an approach, How much will higher growth and better and those scenarios remain generally valid. policy improve these odds? We consider an For this report, the emphasis is shedding light increase of one standard deviation in growth on the likelihood of countries attaining the and in the quality of policy institutional development goals. assessment to be equivalent to about 1.8 per- The statistical probability of achiev- centage points in added growth and to the ing the MDGs is positively correlated with CPIA index rising by 0.5 points. growth and improved policy. The correla- Economic growth can jump-start coun- tion is significant at a 10 percent or better tries particularly far from the goals. For level of statistical significance (see figure A1.1 countries that are far from the target (start- and marginal probabilities in the appendix). ing from a low base), the effects of a one- As expected, growth and policy effectiveness standard-deviation simulated increase in per are positively related with the predicted prob- capita growth on the probabilities of reach- abilities of achieving or being on track for ing some MDGs tend to be distinct and large achieving the MDGs. (figure 1.9b). It would raise 12-fold the prob- Both development drivers count, but ability of reaching the targets for primary growth has an all-encompassing bearing on completion and gender parity in secondary progress toward the MDGs. A closer look education, more than double it for under- at estimation results (table A1.6 in appendix five child mortality and sanitation, almost 1) reveals that economic growth has a per- double it for extreme poverty and hunger, vasively significant and positive impact on and increase it by more than half for access to the odds of achieving all MDGs under con- safe drinking water. sideration, apart from gender parity in pri- For countries close to the target, higher mary education. The quality of policy and growth rates still appear to have a significant institutions also has a positive and statisti- impact on primary education completion and cally significant relation with MDGs for hun- gender parity, but not to the same extent as ger reduction, gender parity, and child and for countries far from the target (note the maternal health. scale of the y-axis in figure 1.9a). This is Consequently, based on the average pattern doubtless because growth is already higher in thus far and at the aggregate level, growth this group (see table 1.5), which likely needs might have a broader impact on attaining better policy to move to a higher plane. MDGs than the quality of policy and insti- Good policies and institutions are vital tutions. This is likely because growth has a for outcome-based MDGs. For lagging more immediate effect and can be generated countries far from the target, this seems true from several sources, including better policy for several health-related MDGs—under- as well as beneficial exogenous shocks and five mortality, maternal mortality, and hun- flows in the global economic environment. By ger—as well as for gender parity in primary GLOBAL MONITORING REPORT 2011 THE DIVERSITY OF MDG PROGRESS 33 education. A one-standard-deviation simu- Why do policies and institutions seem to lated improvement in the quality of policies play a greater role in the chances of reaching and institutions would increase the probabil- health-related MDGs in both groups of off- ity of achieving the hunger target nearly four- target countries? The reason is likely because fold. For the remaining targets, the impact the targets are outcome-based measures that ranges from 152 percent to 67 percent. depend not only on growth and resources For lagging countries close to the target, but also on myriad factors in the system: the effective policies and stronger institutions flow of budgets to localities where resources also appear important to the progress on are needed, accountability and transparency, health-related MDGs. For instance, the odds incentives of service providers and clients, and of reaching targets such as maternal mortality other institutions for service delivery. If the and access to safe drinking water improve by goals for education and gender parity were more than 30 percent after a one-standard- also outcome based (for example, based on deviation increase in the CPIA index. learning outcomes or equal pay for workers BOX 1.5 Reducing inequality and poverty in Brazil In addition to faster growth and better policies, more developing world. According to the World Develop- inclusive growth and equality within countries will ment Indicators database, Brazil’s Gini coefficient of benefit people in the bottom quintiles and lift more income inequality peaked at 63.0 in 1989—the high- people out of poverty. However, despite reduced est among 70 countries with data, just above Sierra inequality across countries over the last 25 years, Leone’s 62.9 and Zambia’s 60.5. Since then, Brazil has within-country inequality has generally increased experienced extraordinary progress in reducing both in most developing countries since 1980.a Brazil is a poverty and inequality—in fact, the country’s Gini notable exception. coefficient has fallen 10 points, while gross domestic At the end of the 1980s, Brazil was one of the product and real household consumption have grown most unequal countries in Latin America and the steadily since 2003, after earlier stagnation. Reduction of income inequality has significant effects on poverty in Brazil a. Distributional effect b. Poverty head count 18 0.5 16 % of population living on less 0.4 14 than $1.25 a day 12 kernel density 0.3 10 8 0.2 6 0.1 4 2 0 PPP $1.25 poverty line 0 0 500 1,500 89 90 92 93 95 96 97 98 99 01 02 03 04 05 06 07 19 19 19 19 19 19 19 19 19 20 20 20 20 20 20 20 log of income per capita, 2005 US$ PPP Brazil 1989a Brazil 2007 Source: World Bank staff calculations based on PovcalNet database. a. Brazil 1989 is a projection of the structure of income distribution in 1989 with the average income in 2007. continued 34 THE DIVERSITY OF MDG PROGRESS GLOBAL MONITORING REPORT 2011 BOX 1.5 Reducing inequality and poverty in Brazil (continued) The 1990s marked the expansion of social safety stagnation. Second, sensible macroeconomic and nets in Brazil. Public social expenditure, including trade policies need not hurt the poor and, in the spe- conditional cash transfers such as the Bolsa Família, cific case of taming hyperinflation, are likely to make targeted to poor families rose from 17.6 percent of a significant contribution in the fight against poverty, GDP in 1990 to 26 .0 percent of GDP in 2008—an even when that is not the primary objective.”d increase of almost 50 percent in education, health, Brazil is also taking steps to reach its indigenous housing, and social security. Recent evidence sug- peoples. Between 2002 and 2007, the number of gests that this increase in social spending and better indigenous schools rose by about 45 percent, and targeting contributed much to reducing poverty and school enrollment among indigenous populations inequality.b From 1994 to 2004, sound macroeco- increased by 50.7 percent. The “Indigenous Portfo- nomic policies also lowered inflation from high levels lio” (Carteira Indígena) Initiative implemented in and improved economic growth in more recent years. 2004 has approved over 250 projects to support eco- Without the reduction in income equality from nomically and environmentally sustainable productive 1989 to 2007, the $1.25-a-day poverty rate would activities for the benefit of 12,000 indigenous families have been 14.5 percent, instead of the 5.2 percent and 60 ethnic groups across 18 Brazilian states; the attained. An additional 17.5 million Brazilians would participation rate by indigenous organizations and have remained in extreme poverty. Using regression communities is more than 80 percent.e Within-coun- decomposition methods, Ferreira, Leite, and Raval- try variation and inequality being important, chapter lion calculate that without the social transfer policies 4 of this report examines one segment of the popu- and programs, the poverty head count index in Brazil lation—indigenous and socially excluded groups of would have been about 5 percentage points higher in people—that is lagging behind even in countries that 2004.c are already on track to achieve the poverty goal. Brazil’s encouraging results illustrate the impor- tance of inclusive economic growth and reduction a. See World Bank (2007, ch. 3) for a more detailed dis- of inequality in the fight against poverty and social cussion of the trends and issues of global and within-country exclusion. Ravallion identifies two main lessons from inequality. the Brazilian experience: “First, reforms to social b. Ferreira, Leite, and Litchfield 2008. policies to make them more pro-poor [if fiscally pos- c. Ferreira, Leite, and Ravallion 2010. sible] can play an important role in sustaining pov- d. Ravallion 2011, p. 17. erty reduction, even during a period of economic e. Government of Brazil 2009. of similar characteristics), the results could service delivery to the poor—appear crucial be similar. The lack of data and defined for achieving health-related MDGs. goals in these areas makes it hard to test this Many more developing countries can get more systematically. (See appendix 1 for more on track. A one-standard-deviation rise in discussion.) both growth and the CPIA would mean that The simulation results generally show as many as 32 more developing countries that economic growth and policy effective- can get on track for the MDGs—an average ness can contribute significantly to achiev- increase of 44 percent in the number of on- ing the MDGs. Although per capita GDP track countries (figure 1.10). This forecast is growth tends to have a broader impact on based on a greater than 50 percent probability development targets, sound policies and of each country getting on track. Statistically, institutions—basic dimensions of effective the probability of lagging countries can only GLOBAL MONITORING REPORT 2011 THE DIVERSITY OF MDG PROGRESS 35 FIGURE 1.9 The odds of getting on target rise substantially with faster growth and better policy Effects of a one-standard-deviation increase in selected development drivers from the multinomial logit estimates a. Change in odds for countries close to the target 200 180 180 160 143 percentage change 140 128 120 100 80 58 60 40 34 20 0 MDG 1.c MDG 2.a MDG 3.a MDG 5.a MDG 7.c hunger primary gender parity maternal access to completion rate (secondary) mortality safe water because of increase in GDP per capita growth because of increase in CPIA index b. Change in odds for countries far from the target 1,111 1,191 500 450 400 percentage change 350 300 281 250 200 163 152 150 120 93 102 100 88 67 61 50 0 MDG 1.a MDG 1.c MDG 2.a MDG 3.a MDG 3.a MDG 4.a MDG 5.a MDG 7.c MDG 7.c extreme hunger primary gender gender child mortality maternal access to access to povery completion parity parity under five mortality safe water sanitation rate (primary) (secondary) because of increase in GDP per capita growth because of increase in CPIA index Source: World Bank staff calculations. See appendix 1 for details. Note: For simplicity of presentation, only those countries with significant percentage changes at the 0.10 level or better are shown. Marginal probabilities and odds ratios are presented in appendix 1. Percentage variations are not comparable across indicators. Average standard deviation increase in GDP per capita growth ≈ 1.8. Average standard deviation increase in CPIA index ≈ 0.5. GDP per capita, purchasing power parity constant 2005 international dollars. reach 100 percent as an upper (asymptotic) percent), maternal mortality (37 percent), and limit, but a 95 percent confidence interval of access to safe drinking water (36 percent). For a 50 percent increase will generally cover that the other MDGs (such as poverty, primary upper limit. The percentage increase in the education completion, and gender equality number of countries getting on track generally in primary and secondary education), the rises most for the targets farthest behind—tar- increase in the number of countries is about gets such as under-five mortality (89 percent), 30 percent, still substantial. Individual coun- hunger (64 percent), access to sanitation (54 tries that are good candidates to get on track 36 THE DIVERSITY OF MDG PROGRESS GLOBAL MONITORING REPORT 2011 FIGURE 1.10 Growth and policy reforms will put many countries on track 27 22 24 16 15 32 89 82 19 16 11 66 47 55 36 30 35 25 MDG 1.a MDG 1.c MDG 2.a MDG 3.a MDG 3.a MDG 4.a MDG 5.a MDG 7.c MDG 7.c extreme hunger primary gender gender child mortality maternal access to access to povery completion parity parity under five mortality safe water sanitation rate (primary) (secondary) number of countries currently on target number of additional countries becoming on target Source: World Bank staff calculations based on a one-standard-deviation increase or improvement in growth and policy. See appendix 1 for details. are those currently very close—that is, within the recent global crisis years (2008–09). The 10 percent of getting on track (table 1.4). increase, in fact, is very much within the How achievable are these gains? Recent realm of actual performance for Sub-Saharan history suggests they may be. Achieving the African countries during periods of growth growth assumption for developing countries acceleration (3.9 percent), including the high- appears possible. To put the one-standard- growth period 2000–07.19 deviation growth increase in context, per cap- For the two off-target groups, growth dur- ita GDP growth will need to double from its ing the recent global crisis did not fall below historical rate of 1.9 percent a year. Even so, the rates in the reference period (1990–2009), the historical rate is an average covering all corroborating other economic assessments types of developing countries and the uneven that low-income countries did relatively well subperiods during 1990 –2009, including (table 1.6). Three factors explain why the TABLE 1.6 A one-standard-deviation increase in growth is definitely achievable Developing countries (GDP per capita growth rate) Growth period Years covered on target close to the target far from the target I. Reference period 1990–2009 2.42 1.77 1.22 II. Recent growth accelerations Modern trend-break 1995–2007 3.46 2.61 2.01 New millennium 2000–07 3.97 2.90 2.25 Boom years 2003–07 4.82 3.65 3.07 III. Recent global economic crisis Crisis years 2008–09 1.48 1.79 1.48 Peak crisis 2009 –1.09 0.28 0.65 IV. Growth prospects 2010–15 3.58 3.33 3.22 V. Growth assumption n.a. n.a. 3.57 3.02 Sources: World Development Indicators database. Growth prospects are from the IMF’s World Economic Outlook.World Economic Outlook. Note: The growth assumption is growth during reference period plus one standard deviation, or 1.8. Growth rates are all simple averages, giving equal weight to each country GDP per capita, purchasing power parity constant 2005 international dollars. n.a. = not applicable. GLOBAL MONITORING REPORT 2011 THE DIVERSITY OF MDG PROGRESS 37 MAP 1.2 Sub-Saharan Africa and Southern Asia are home to the vast majority of children out of school IBRD 38446 APRIL 2011 Primary completion rate Greenland (Den) Faeroe Norway Iceland Islands Finland (Den) Sweden The Netherlands Russian Federation Estonia Canada Isle of Man (UK) Denmark Russian Latvia Fed. Lithuania United Ireland Kingdom Germany Poland Belarus Channel Islands (UK) Belgium Ukraine Luxembourg Moldova Kazakhstan Liechtenstein Romania Mongolia Switzerland France Italy Andorra Bulgaria Georgia Azer- Uzbekistan Kyrgyz D.P.R. Turkey Armenia baijan Turkmenistan Rep. United States Portugal Spain Tajikistan of Korea Japan Monaco Greece Gibraltar (UK) Cyprus Syrian Rep. of Tunisia Malta Lebanon Arab Islamic Rep. Afghanistan China Korea Israel Rep. Iraq of Iran Bermuda Morocco Kuwait (UK) West Bank and Gaza Jordan Bahrain Pakistan Nepal Bhutan Former Algeria Libya Arab Rep. Qatar The Bahamas Spanish of Egypt Saudi Cayman Is. (UK) Sahara Arabia United Arab Bangladesh Cuba Turks and Caicos Islands (UK) India Mexico Emirates Myanmar Lao Haiti Mauritania Oman P.D.R. Belize Jamaica Cape Verde Mali N. Mariana Islands (US) Niger Rep. of Thailand Vietnam Guatemala Honduras Senegal Chad Eritrea Yemen Guam (US) El Salvador Nicaragua The Gambia Burkina Sudan Marshall Faso Djibouti Cambodia Philippines Islands Guinea-Bissau Guinea Benin Federated States of Costa Rica Panama Sri Micronesia R.B. de Guyana Sierra Leone Côte Ghana Nigeria Central Ethiopia Lanka Venezuela Suriname Liberia D’Ivoire African Rep. Brunei Cameroon Somalia Palau Colombia French Guiana (Fr) Togo Malaysia Equatorial Guinea Uganda Maldives Ecuador São Tomé and Príncipe Gabon Congo Rwanda Kenya Singapore Nauru Dem. Rep. of Burundi Seychelles Papua Solomon Congo Indonesia New Guinea Islands Tanzania Comoros Peru Brazil Timor-Leste Angola Zambia Malawi Mayotte (Fr) Vanuatu Bolivia Zimbabwe Mauritius Namibia Madagascar Paraguay Botswana Mozambique Réunion (Fr) New Caledonia Australia (Fr) Swaziland South Lesotho Africa Chile Uruguay Argentina Primary completion rate New Dominican Puerto British Virgin Poland Percentage of students completing Zealand Germany Republic Rico (US) Islands (UK) Ukraine Anguilla (UK) Czech Rep. the last year of primary school U.S. Virgin Islands Antigua and Barbuda Slovak Rep. (US) St. Maarten (Neth) St. Kitts Guadeloupe (Fr) Austria Hungary <49.9 and Nevis Dominica Slovenia Romania 50–74.9 Curaçao Montserrat (UK) Martinique (Fr) Croatia (Neth) St. Lucia Bosnia and 75–89.9 Aruba Bonaire St. Vincent and San Herz. Serbia Bulgaria (Neth) (Neth) The Grenadines Barbados Marino 90–99.9 Montenegro Kosovo Grenada FYR ≥100 Italy Macedonia Trinidad Vatican Albania R.B. de Venezuela and Tobago City Greece no data MDG performance Greenland (Den) Faeroe Norway Iceland Islands Finland (Den) Sweden The Netherlands Russian Federation Estonia Canada Isle of Man (UK) Denmark Russian Latvia Fed. Lithuania United Ireland Kingdom Germany Poland Belarus Channel Islands (UK) Belgium Ukraine Luxembourg Moldova Kazakhstan Liechtenstein Romania Mongolia Switzerland France Italy Andorra Bulgaria Georgia Azer- Uzbekistan Kyrgyz D.P.R. Turkey Armenia baijan Turkmenistan Rep. United States Portugal Spain Tajikistan of Korea Japan Monaco Greece Gibraltar (UK) Cyprus Syrian Rep. of Tunisia Malta Lebanon Arab Islamic Rep. Afghanistan China Korea Israel Rep. Iraq of Iran Bermuda Morocco Kuwait (UK) West Bank and Gaza Jordan Bahrain Pakistan Nepal Bhutan Former Algeria Libya Arab Rep. Qatar The Bahamas Spanish of Egypt Saudi Cayman Is. (UK) Sahara Arabia United Arab Bangladesh Cuba Turks and Caicos Islands (UK) India Mexico Emirates Myanmar Lao Haiti Mauritania Oman P.D.R. Belize Jamaica Cape Verde Mali N. Mariana Islands (US) Niger Rep. of Thailand Vietnam Guatemala Honduras Senegal Chad Eritrea Yemen Guam (US) El Salvador Nicaragua The Gambia Burkina Sudan Marshall Faso Djibouti Cambodia Philippines Islands Guinea-Bissau Guinea Benin Federated States of Costa Rica Panama Sri Micronesia R.B. de Guyana Sierra Leone Côte Ghana Nigeria Central Ethiopia Lanka Venezuela Suriname Liberia D’Ivoire African Rep. Brunei Cameroon Somalia Palau Colombia French Guiana (Fr) Togo Malaysia Equatorial Guinea Uganda Maldives Ecuador São Tomé and Príncipe Gabon Congo Rwanda Kenya Singapore Nauru Dem. Rep. of Burundi Seychelles Papua Solomon Congo Indonesia New Guinea Islands Tanzania Comoros Peru Brazil Timor-Leste Angola Zambia Malawi Mayotte (Fr) Vanuatu Bolivia Zimbabwe Mauritius Namibia Madagascar Paraguay Botswana Mozambique Réunion (Fr) New Caledonia Australia (Fr) Swaziland South Lesotho Africa Chile Uruguay Argentina MDG performance New Dominican Puerto British Virgin Poland MDG 2.A Ensure that by 2015, children everywhere, Zealand Germany Republic Rico (US) Islands (UK) boys and girls alike, will be able to complete a full Ukraine Anguilla (UK) Czech Rep. U.S. Virgin Islands Antigua and Barbuda Slovak Rep. course of primary schooling (US) Austria St. Maarten (Neth) St. Kitts Guadeloupe (Fr) Hungary and Nevis Slovenia on target Dominica Romania Curaçao Montserrat (UK) Martinique (Fr) Croatia close to target (Neth) Bosnia and Bonaire St. Vincent and St. Lucia San Herz. Serbia Aruba far from target Bulgaria (Neth) (Neth) The Grenadines Barbados Marino This map was produced by the Map Design Unit of The World Bank. Montenegro Kosovo Grenada Italy FYR no data The boundaries, colors, denominations and any other information shown on this map do not imply, on the part of The World Bank Trinidad Macedonia and Tobago Vatican Albania Group, any judgment on the legal status of any territory, or any R.B. de Venezuela City Greece endorsement or acceptance of such boundaries. Source: World Bank staff calculations based on data from the World Development Indicators database. 38 THE DIVERSITY OF MDG PROGRESS GLOBAL MONITORING REPORT 2011 recent crisis was different for low-income broadly consistent over time. For instance, countries. First, policies and institutions from 1998 to 2003, 32 countries (24 percent improved before the crisis, and economic of developing countries for which scores are growth accelerated after the mid-1990s— available) experienced an improvement of particularly after 2000. Second, unlike previ- 0.5 points or better, especially countries in ous crises, the recent one was not caused by Eastern Europe. More recently, during the domestic policy failure, which would have period 2004–09 when the new system has severe impacts on human development out- been stable, countries that have achieved an comes—particularly on child and maternal improvement of 0.5 include Georgia, Nigeria, mortality. Third, spending on social safety and Seychelles. As one of the few broad mea- nets was protected by governments with the sures available for policy and institutions, it assistance of international fi nancial institu- is a proxy for the point that significant policy tions and the donor community.20 Even dur- reforms are needed, especially for outcome- ing the peak in 2009, their average growth based or system-oriented MDGs. Because stayed in positive territory. Their recovery policy reforms can take time to implement is now expected to be strong, with growth and bear fruit, it is also important to under- prospects for 2010–15 similar to the growth take significant reforms sooner than later. assumption (which is higher than in the refer- ence period, but as good as the recent growth accelerations, variously defined). What is in the rest of this report The global crisis struck the on-track Developing countries are doing better when developing countries much harder. At its looking at country-level figures than at global peak, growth in this group was negative. figures. Lagging countries, on average, are However, many of the countries are higher- very close to the targets, and their odds of get- middle-income ones—particularly in Eastern ting on track can improve dramatically with Europe, where the MDGs were less of a chal- stronger growth and sounder policy. Eco- lenge. The growth forecast for 2010–15 is still nomic growth has a pervasive effect on all the higher than the record in the reference period MDGs and can jump-start countries far from or the recent trend-break (1995–2007). the target. The implications are clear. With Where a problem may surface is in improv- 2015 only a few years away, growth in devel- ing policy and institutions, given the few oping countries needs to be taken quickly to years left until 2015. A one-standard-devia- a higher plane, the fastest way to lift more tion improvement in the CPIA is equivalent countries to the MDGs. Chapter 2 examines to a 0.5-point increase, or about the differ- the prospects and challenges for economic ence between the CPIA for on-target coun- growth in developing countries, how they are tries and for countries far from the target (see recovering from the recent global economic table 1.5). A 0.5-point increase in a CPIA crisis, and what they need to do to boost rating is the normal award for an improve- growth further. ment in any policy area in a country. But to The quality of policies and institutions do this consistently for all the 16 questions in is more crucial for MDGs that are health the CPIA is much harder. In any given year, related or are lagging the most, as well as for a 0.1-point increase in the overall score rep- countries close to the target. What constitutes resents a significant policy improvement for good policies and institutions in developing a country; a 0.2- or 0.3-point increase rep- countries is complex, however, covering a resents a substantial policy shift or regime wide range of areas. change—rare for any country. Interventions can be broad and wide rang- But it is certainly conceivable over time. ing or specific to local circumstances and The World Bank’s CPIA has undergone problems. Much has been written about the changes to improve its assessment and is only broad issues, but less about microeconomic GLOBAL MONITORING REPORT 2011 THE DIVERSITY OF MDG PROGRESS 39 interventions, precisely because of specificity For example, the assessment of governance and local context. But thanks to the World issues was not included in the earlier CPIA. Bank’s Development Impact Evaluation Ini- 12. Average GDP per capita growth in IDA tiative and similar efforts, more documen- countries (1990–2009) is 1.36, a point below tation has emerged in recent years. To help average growth in non-IDA countries (2.38). untangle the difficult challenge of improving The CPIA index in 2009 is, on average, 3.26 policies and institutions in developing coun- in IDA countries versus 3.69 in non-IDA tries, chapter 3 presents findings and lessons countries. Fragile or confl ict-affected coun- tries (one or more years, 2006–09) exhibit from impact evaluations in health and educa- average per capita GDP growth (1990 – tion, where several of the studies have already 2009) close to 1.03 percent and a CPIA been completed. index of 3.00 in 2009. However, nonfragile Chapter 4 examines what countries can do states have grown, in per capita terms, at to help the bottom 10 percent of society, usu- an average rate of 2.27 percent since 1990. ally indigenous and socially excluded groups. The CPIA index for these countries is 3.68 Because a beneficial global environment in 2009. (Goal 8) also helps support MDG progress 13. World Bank 2010. Harttgen and Klasen in developing countries, chapter 5 reviews (2010) show that fragile countries perform global trade, aid, and the actions of interna- worse on MDG levels, but that their MDG tional financial institutions for developing progress is not necessarily slower. countries. Goal 7 on environmental sustain- 14. Marshall and Cole 2010. ability and biodiversity has no well-defi ned 15. It is important to point out that these simple targets but it affects other MDGs, such as graphical patterns can be driven by more child health and human development more fundamental development factors, such as generally. Appendix 2 describes some aspects growth and institutions. The next section based on contributions from the Netherlands tackles some of these issues. Environmental Assessment Agency. 16. World Bank 2004. 17. A study (IDA/DECVP 2007) found the cor- relation between contemporaneous CPIA and Notes growth to be weak and the correlation between 1. World Bank 2010; Easterly 2009; Clemens, CPIA and future growth to be strong. The Kenny, and Moss 2007. CPIA measures the level of policies, not the 2. Ravallion 2009; World Bank 2010. change; and it focuses on actual implementa- 3. Leo and Barmeier 2010. tion, not just introduction or announcement. 4. See, for example, Bourguignon et al. (2008), It is therefore backward looking. The inclu- Leo and Barmeier (2010), and ODI (2010). sion of the 1996 CPIA is an attempt to capture 5. Collier and O’Connell 2006. the policy achievements close to the reference 6. Leo and Barmeier 2010. year in 1990, and the 2009 index will include 7. See, for example, the first Global Monitoring the more recent record. See appendix 1 for a Report (World Bank 2004). discussion of the use of the CPIA index and 8. World Bank 2010. other alternatives. See also the guide for CPIA 9. Issues include macroeconomic and fiscal pol- in World Bank (2009). icy, debt policy, trade, human development 18. A review of the determinants of MDGs is in policy in education and health, gender equal- chapter 3. See also Lay (2010) and Lofgren ity, social protection, budgetary and financial (2010) for extensive reviews on the determi- management, and corruption in the public nants of education-related and health-related sector. MDG indicators. 10. See World Bank (2009) and footnote 17. 19. See, for example, Arbache, Go, and Page 11. An earlier version of the CPIA goes back to (2008). 1970s but uses a different scale and criteria. 20. See World Bank (2010). 40 THE DIVERSITY OF MDG PROGRESS GLOBAL MONITORING REPORT 2011 References and Development Effectiveness.” IDA 15. Washington, DC. Arbache, J., D. Go, and J. Page. 2008. “Is Afri- Lall, S. 2000. “The Technological Structure and ca’s Economy at a Turning Point” In Africa at Performance of Developing Country Manufac- a Turning Point? Growth, Aid and External tured Exports, 1985–1998.” Working Paper Shocks, ed. D. Go and J. Page, 14–86. Wash- 44. Queen Elizabeth House, University of ington, DC: World Bank. Oxford, Oxford, UK. Bourguignon, F., A. Bénassy-Quéré, S. Dercon, Lay, J. 2010. “MDG Achievements, Determinants A. Estache, J. W. Gunning, R. Kanbur, S. Kla- and Resource Needs: What Has Been Learnt?” sen, S. Maxwell, J.P. Platteau, and A. Spadaro. Policy Research Working Paper 5320, World 2008. “Millennium Development Goals at Bank, Washington, DC. Midpoint: Where Do We Stand and Where Do Leo, B., and J. Barmeier. 2010. “Who Are the We Need to Go?” European Report on Devel- MDG Trailblazers? A New MDG Progress opment, European Commission, Brussels. Index.” Working Paper 222, Center for Global Clemens, M. A., C. J. Kenny, and T. J. Moss. Development, Washington, DC. 2007. “The Trouble with the MDGs: Con- Lofgren, H. 2010. “What Determines the Evolu- fronting Expectations of Aid and Development tion of MDG Indicators? A Selective Review Success.” World Development 35 (5): 735–51. of the Literature.” Unpublished manuscript, Collier, P., and S. A. O’Connell. 2006. “Oppor- World Bank, Washington, DC. tunities and Choices.” Explaining African Marshall, M., and B. Cole. 2010. Global Report Economic Growth, chapter 2 of synthesis vol- 2009: Confl ict, Governance, and State Fra- ume. African Economic Consortium, Nairobi, gility. Washington, DC: Center for Global Kenya: Policy. Easterly, W. 2009. “How the Millennium Devel- ODI (Overseas Development Institute). 2010. opment Goals Are Unfair to Africa.” World “Millennium Development Goals Report Development 37 (1): 26–35. Card: Measuring Progress across Countries.” London. Ferreira, F., P. Leite, and J. Litchfield. 2008. PBL (Netherlands Environmental Assessment “The Rise and Fall of Brazilian Inequality: Agency). 2008. Towards a Global Integrated 1981–2004.” Macroeconomic Dynamics 12: Sustainabilit y Model: GISMO 1.0 Sta- 199–230. tus Report. Report 550025002. Bilthoven: Ferreira, F., P. Leite, and M. Ravallion. 2010. Netherlands. “Pover t y Reduction without E conomic ——— . 2009a. Beyond 2015: Long-Term Devel- Growth? Explaining Brazil’s Poverty Dynam- opment and the Millennium Development ics 1985–2004.” Journal of Development Eco- Goals. Bilthoven, Netherlands. nomics 93 (1): 20–36. ———. 2009b. Growing within Limits. A Report Government of Brazil. 2009. “Comments of the to the Global Assembly 2009 of the Club of Government of Brazil to the report of the Spe- Rome. Bilthoven, Netherlands. cial Rapporteur on the Situation of Human ———. 2011. “Scarcities in a Sea of Plenty: Global Rights and Fundamental Freedoms of Indig- Resource Scarcities and Policies in the Euro- enous Peoples,” James Anaya. Presentation at pean Union and the Netherlands.” Bilthoven, the United Nations on July 9, 2009. Netherlands. Harttgen, K., and S. Klasen. 2010. “Fragility and Ravallion, M. 2009. “Why Don’t We See Poverty MDG Progress: How Useful Is the Fragility Convergence?” Policy Research Working Paper Concept.” Working Paper 20, Robert Schu- 4974, World Bank, Washington, DC. man Centre for Advanced Studies, European ——— . 2011. “A Comparative Perspective on University Institute, Florence, Italy. Poverty Reduction in Brazil, China and IDA/DECVP (International Development Asso- India.” World Bank Research Observer 26 (1): ciation and Development Economics, Office 71–104. of the Chief Economist). 2007. “Selectivity UNCTAD (United Nations Conference on Trade and Performance: IDA’s Country Assessment and Development). 2010. “Liner Shipping GLOBAL MONITORING REPORT 2011 THE DIVERSITY OF MDG PROGRESS 41 Connectivity Index.” Division on Technology ——— . 2009. “Country Policy and Institutional and Logistics, Geneva, Switzerland. Assessments—2009 Assessment Question- Uwezo Tanzania. 2010. Are Our Children Learn- naire.” Operations Policy and Country Services. ing? Annual Learning Assessment Report Washington DC: World Bank. Tanzania 2010. Dar es Salaam, Tanzania. ——— . 2010. Global Monitoring Report 2010: World Bank. 2004. Global Monitoring Report: The MDGs after the Crisis. Washington, DC. Policies and Actions for Achieving the Millen- ——— . 2011. World Development Report 2011: nium Development Goals and Related Out- Conflict, Security and Development. Wash- comes. Washington, DC. ington, DC. ——— . 2007. Global Economic Prospects 2007: Managing the Next Wave of Globalization. Washington, DC. 2 Economic Growth in Developing Countries I n chapter 1, economic growth is seen policies that softened the impact of the cri- as critical to attaining the Millennium sis and drove a relatively rapid return to pre- Development Goals (MDGs). Prospects crisis growth rates. To substantially reduce for further progress on the MDGs should poverty and meet the MDGs, however, low- be seen in light of macroeconomic develop- income and other developing countries need ments in emerging and developing econo- to grow faster and to rebuild buffers so as mies, and in the global economic environ- to guard against future shocks. In addition ment they face. to strong, well-designed policies in the coun- Economic recovery is proceeding along tries themselves, international cooperation is two tracks. Advanced economies are expe- required to restore a global economic envi- riencing subdued growth with high unem- ronment conducive to poverty reduction and ployment, while many fast-growing emerg- development and to provide adequate assis- ing economies are seeing infl ation pressures tance, with special attention to the most vul- build amid some signs of overheating. Pro- nerable countries. jections put annual global growth at about 4.5 percent in 2011 and 2012. But sustain- ing the global recovery demands divergent Economic recovery policy responses: in the advanced economies, Global economic activity has gathered redressing fiscal imbalances and repairing and pace, unevenly reforming financial systems; in the emerging economies, facilitating external rebalancing The World Economic Outlook of the Inter- and checking overheating pressures, in many national Monetary Fund (IMF) estimates cases allowing further exchange rate appre- that after contracting by 0.5 percent in 2009, ciation, and in some, more ambitious fi scal global GDP expanded by 5.0 percent in tightening. 2010, compared with the projection of 4.2 Good policies among low-income coun- percent a year ago (table 2.1). The recovery tries contributed to strong growth before is markedly uneven, however, with rates the crisis. The policy buffers established of growth in advanced market economies then also allowed for active countercyclical several percentage points less than those in A girl receives a tuberculosis vaccination as part of the effort to vaccinate Baluchi nomads. 43 44 ECONOMIC GROWTH IN DEVELOPING COUNTRIES GLOBAL MONITORING REPORT 2011 TABLE 2.1 Global output, 2007–14 annual percentage change Projections Region 2007 2008 2009 2010 2011 2012–14 World output 5.4 2.9 –0.5 5.0 4.4 4.6 Advanced economies 2.7 0.2 –3.4 3.0 2.4 2.5 Emerging and developing economies 8.8 6.1 2.7 7.3 6.5 6.6 Central and Eastern Europe 5.5 3.2 –3.6 4.2 3.7 3.9 Commonwealth of Independent States 9.0 5.3 –6.4 4.6 5.0 4.6 Developing Asia 11.4 7.7 7.2 9.5 8.4 8.5 Middle East and North Africa 6.2 5.1 1.8 3.8 4.1 4.5 Sub-Saharan Africa 7.2 5.6 2.8 5.0 5.5 5.7 Western Hemisphere 5.7 4.3 –1.7 6.1 4.7 4.0 Emerging economies 9.2 6.3 2.6 7.5 6.7 6.7 Other developing economies 7.2 6.0 5.2 6.2 6.1 6.4 Least developed countries (LDCs)a 9.0 6.9 5.2 5.3 6.1 6.4 Source: World Economic Outlook. a. United Nations classification, a subset of developing countries. emerging market and developing economies.1 and in other countries with strong precrisis Although below the levels existing immedi- fundamentals. But in many advanced mar- ately before the crisis, per capita real growth ket economies, lower property values and rates in low-income countries remained posi- stubbornly high unemployment reduced tive even in the depths of the economic crisis household wealth and incomes, while weak- in 2009, and they strengthened further in nesses in the fi nancial system still constrain 2010 (figure 2.1). credit; together, these factors are restraining Spurred by expanding inventories and consumption. Exceptional policy stimulus fi xed investments, global merchandise trade helped advanced market economies achieve 3 rose 13 percent in 2010, while industrial pro- percent growth in 2010—still a modest rate, duction also rebounded. Household spending considering that they are emerging from a fi rmed in many emerging market economies deep recession. Price inflation remains low in advanced economies, where unemployment is high and FIGURE 2.1 Low-income economies’ per capita growth remained excess capacity is considerable, but is becom- positive in 2009, in contrast to elsewhere ing a policy concern in some emerging market 6 economies. Inflation rates in emerging mar- ket and some developing economies average 4 % change, median country about 6 percent, reflecting rising commodity 2 prices and other factors. Most low-income countries do not yet seem to have experienced 0 an increase in inflation. –2 –4 Commodity prices have rebounded, –6 leading to concerns over food price 2007 2008 2009 2010 2011 inflation projected emerging markets low-income countries Together with other factors, economic recov- advanced market economies ery brought sharp increases in metals, food, and other commodity prices in 2010 (figure Sources: World Economic Outlook; IMF 2010a; IMF staff estimates. 2.2). This helped support a strong growth GLOBAL MONITORING REPORT 2011 ECONOMIC GROWTH IN DEVELOPING COUNTRIES 45 FIGURE 2.2 Economic recovery has taken many commodity prices sharply higher 600 copper 500 450 index (2001Q1 = 100) 400 fuel 350 all commodities 300 cereals 250 nonfuel 200 food 150 100 50 0 Q1 Q1 Q1 Q1 Q1 Q1 Q1 Q1 Q1 Q1 Q1 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 projected Sources: World Economic Outlook; IMF staff estimates. Note: Indexes are in U.S. dollars. Data for 2011 are projected. rebound among low-income commodity uptick in mid-2010, the fall in international exporters. However, surging food prices are bond spreads toward 2006–07 levels suggests again sparking concerns over the affordability largely restored market access for sovereign of food for the poorer segments of the popu- borrowers and investment-grade corporations lation in some low- and lower-middle-income (figure 2.3). Market access for subinvestment- countries (box 2.1).2 Food price increases are grade borrowers has also been substantially also adding to inflationary pressures in many restored, though high-yield spreads remain emerging market economies, while in parts well above those before the crisis—even if of Sub-Saharan Africa good harvests have yields are now lower. Share prices have also muted food price increases. With continued increased (figure 2.4). Bank financing has economic recovery in 2011, overall commod- continued to recover, notably to emerging ity prices may rise further—though food markets in Asia and Latin America. With price levels will depend greatly on weather the protracted economic problems of emerg- patterns during the year. ing Europe, bank lending to them contracted modestly in 2010 (figure 2.5). Improved financial regulation and cri- Policy challenges remain for global sis policy measures helped banking systems financial stability, but most low-income in most emerging and developing countries countries weathered the crisis well withstand the crisis well. The share of nonper- In advanced economies, leverage is still high, forming loans did, however, increase some- and balance sheet restructuring and regula- what further in 2010, as it did in advanced tory reform are incomplete. The heightened economies (figure 2.6). Banking sectors sensitivity of fi nancial markets to fiscal out- have also benefited from financial market looks is leading to funding pressures in some resilience, including somewhat more stable markets. Differences in interest rates and exchange and interest rates. Even so, linger- growth prospects have spurred strong capital ing concerns about systemic risks to bank and flows from advanced to emerging (and some corporate solvency highlight the need to fur- developing) economies. ther strengthen the financial supervisory and Emerging and developing countries’ institutional framework. To safeguard finan- access to international financial markets has cial stability, the problems of impaired assets returned broadly to precrisis levels. After an need to be addressed. 46 ECONOMIC GROWTH IN DEVELOPING COUNTRIES GLOBAL MONITORING REPORT 2011 BOX 2.1 The global food price surge in 2010—causes and prospects The IMF food price index has increased sharply since recessions have been associated with lower demand the middle of 2010 and has surpassed its prereces- growth. During the most recent recession, how- sion peaks (left figure). (For the index’s definition ever, demand growth remained steady and has now and methodology, see “IMF Primary Commodity risen to around 2.5 percent a year for the major Prices” at http://www.imf.org.) Price increases have crops. China and other emerging market economies been broad based and led by a surge in grain prices account for 70–80 percent of demand growth dur- by 80 percent. Prices of key grains, including corn ing the past three years. The U.S. corn ethanol sector and wheat, have risen sharply to close to their 2008 has rebounded from the 2008–09 crisis. Almost 40 peaks. For more income-elastic food groups, such as percent of the 2010 U.S. corn crop is estimated to vegetable oils, meat, and seafood, prices are pushing have been used as ethanol feedstock, an increase of 5 past previous highs (right figure). percentage points over the previous year. The catalyst for the recent food price rise has been Low global inventories and trade restrictions a series of weather-related supply shocks to major have exacerbated the price response to supply disap- grains. Wheat prices were the fi rst to surge, as the pointments in recent quarters. Global food invento- improving supply picture reversed sharply in mid- ries were run down to very low levels in 2002–08 2010. Drought and wild fi res in the Russian Federa- as demand picked up and the global supply response tion and Kazakhstan and heavy rain in Ukraine led lagged. That response has now begun, most evident to significant downward revisions in global produc- in the increase in acreage harvested; but stocks will tion estimates. Corn prices began to rise as estimates likely only recover gradually. The imposition and of the 2010/11 U.S. harvest were progressively down- extension of grain export restrictions by Russia and graded because of adverse weather during the sum- Ukraine in 2010 added to price volatility. mer months. Output should recover quickly from recent sup- La Niña has contributed to adverse weather con- ply shocks, and increased global acreage indicates ditions around the Pacific Rim. Recent satellite evi- that more normal weather conditions will bring large dence indicates that this is among the most intense harvests in 2011. Weather-related supply shortfalls La Niña episodes of the last 50 years and that its tend to be followed by sharp recoveries in output for effects are likely to be close to their peak during many major crops, including wheat. This would ease the southern hemisphere summer. This has already market tightness and allow prices to decline modestly affected rice harvests across Asia and food products over the next 12 months, although they will remain typically not traded globally, such as local fruits and high compared with 10 –15-year averages. One vegetables. important risk is that the La Niña weather pattern While supply has disappointed, demand for major may adversely affect southern hemisphere harvests. food crops has remained robust, largely reflecting More generally, food markets will remain vulner- growth in emerging economies but also demand able to more widespread supply shocks for as long as for biofuel feedstock. Many important food items inventories are low. (See also related discussion in the are relatively income inelastic, but previous global trade section of chapter 5.) IMF food price index, 2005–11a IMF food price index groups, 2005–11a 190 300 170 250 150 200 130 index index 150 110 100 90 70 50 50 0 2005 2006 2007 2008 2009 2010 Jan 2005 2006 2007 2008 2009 2010 Jan 2011 2011 cereals meat seafood vegetable oils Source: International Monetary Fund. a. U.S. dollar price index rebased at 2005 = 100. Data for 2011 are current as of February 2011. GLOBAL MONITORING REPORT 2011 ECONOMIC GROWTH IN DEVELOPING COUNTRIES 47 FIGURE 2.3 Bond spreads and international bond issuance are virtually back to precrisis levels in emerging and developing countries 70 1,000 60 900 800 spreads, basis points 50 700 US$, billions 40 600 500 30 400 20 300 200 10 100 0 0 Q2 Q1 Q4 Q3 Q2 Q1 Q4 Q3 Q2 Q1 Q4 Q3 Q2 Q1 Q4 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 EMBI global (right axis) issues of international bonds (left axis) Sources: Dealogic; Bloomberg; IMF staff estimates. FIGURE 2.4 Share prices, too, have bounced back 900 800 index (Jan 2000 = 100) 700 600 500 400 300 200 100 0 Jan Jun Nov Apr Sep Feb Jul Dec May Oct Mar Aug Jan Jun Nov Apr Sep Feb Jul Dec May Oct Mar Aug Jan Jun Nov Jan 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 emerging markets developing countries Sources: International Financial Statistics; IMF staff estimates. Note: Prices are presented in local currency. FIGURE 2.5 Emerging market bank financing continues to recover 20 % change from previous quarter 15 10 5 0 –5 –10 –15 –20 Q1 Q2 Q3 Q4 Q1 Q2 Q3 Q4 Q1 Q2 Q3 Q4 Q1 Q2 Q3 2007 2008 2009 2010 Asia Emerging Europe Latin America Middle East/Africa Sources: Bank for International Settlements; IMF staff estimates. Note: Liabilities to BIS-reporting banks, adjusted for exchange rate changes. Changes are calculated as flow adjusted for exchange rate changes as a share of the stock in the previous quarter. 48 ECONOMIC GROWTH IN DEVELOPING COUNTRIES GLOBAL MONITORING REPORT 2011 FIGURE 2.6 Nonperforming loans in emerging and developing reserves; the size of their current account countries, 2003–10 surpluses did not fall and may have widened somewhat. After large swings in 2009, terms- 12 of-trade changes in 2010 seem to have been 10 relatively benign (figure 2.8). In emerging and low-income countries in 2010, levels of official share of total loans, % 8 reserves roughly kept pace with the rebound in 6 trade, with only modest declines in the median ratio of reserves to imports and, for emerging 4 economies, increasing somewhat relative to 2 short-term external debt (figure 2.9).3 The crisis affected low-income countries 0 mainly through export demand, foreign direct 2003–07 2008 2009 2010 investment (FDI), and remittances, more than emerging markets developing countries global interest rates or the terms of trade.4 advanced economies Exports from low-income countries contracted very sharply in late 2008 and early 2009. Crit- Sources: Global Financial Stability Report; IMF staff estimates. ically, though, relative to previous crises, these countries had entered this crisis with lower external debt, narrower fiscal and current Trade and capital flows have bounced account deficits, and higher external reserves back, into a changed picture (figure 2.10). (See also chapter 1, regarding Globally, current account imbalances wid- growth and these countries’ progress toward ened in 2010 after narrowing considerably reaching the MDGs.) These proved to be in 2009 in the midst of the crisis (figure 2.7). important policy buffers that could be drawn They remain below the levels of 2004–08. A upon to help cushion the shock (see lower number of emerging market economies that external debt, narrower fiscal and current experienced renewed surges in capital inflows account deficits, and higher external reserves in 2010 accumulated more foreign exchange below). By late 2010, monthly exports to the FIGURE 2.7 Global current account imbalances, 2000–11 3.5 2.5 1.5 % of world GDP 0.5 –0.5 –1.5 –2.5 –3.5 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011a low-income countries with surplus United States emerging markets with deficit emerging markets with surplus other advanced countries with deficit low-income countries with deficit other advanced countries with surplus Sources: World Economic Outlook; IMF staff estimates. Note: The global statistical discrepancy is not shown. a. projected. GLOBAL MONITORING REPORT 2011 ECONOMIC GROWTH IN DEVELOPING COUNTRIES 49 FIGURE 2.8 Annual changes in terms of trade, by quintile group 35 30 25 20 15 % change 10 5 0 –5 –10 –15 –20 2007 2008 2009 2010 2011 2007 2008 2009 2010 2011 2007 2008 2009 2010 2011 lowest quintile countries highest quintile countries all countries Sources: World Economic Outlook; IMF staff estimates. Note: Quintile groups are based on the average (mean) of terms-of-trade changes in 2009 and 2010. FIGURE 2.9 Reserve levels broadly kept pace with the rebound in trade a. Reserves, in months of exports b. Reserves, in terms of the stock of external short-term debta 10 2.5 9 8 2.0 7 6 1.5 months index 5 4 1.0 3 2 0.5 1 0 0.0 2006 2007 2008 2009 2010 2006 2007 2008 2009 2010 emerging economies developing countries emerging economies median, emerging median, developing economies countries Sources: World Economic Outlook; IMF staff estimates. Sources: World Economic Outlook; IMF staff estimates. Note: Bars represent the range between the 25th and 75th percentiles. a. The median ratio is shown. Stock of short-term external debt includes amortization paid to official creditors. Group of Twenty (G-20) countries approached developing countries is mixed. In aggregate, their peak precrisis levels—but with interest- there was a sharp recovery in trade credit ing compositional shifts (box 2.2). and strengthening in portfolio equity and Despite important signs of a recovery in bond flows in 2010, and modest increases financial flows, the picture for emerging and in FDI flows.5 However, (simple average) net 50 ECONOMIC GROWTH IN DEVELOPING COUNTRIES GLOBAL MONITORING REPORT 2011 FIGURE 2.10 Low-income countries’ precrisis macroeconomic policy buffers were stronger than in earlier years 6 45 40 5 35 4 30 percent 25 3 20 2 15 10 1 5 0 0 current account deficit fiscal deficit/GDP reserve coverage inflation external public adjusted for FDI/GDP (percent, left axis) (months of imports, (right axis) debt/GDP (percent, left axis) left axis) (right axis) previous crises 2005–07 Sources: World Economic Outlook; IMF staff estimates. BOX 2.2 Low-income-country exports during the exit from the crisis: early evidence Exports from low-income to G-20 countries recov- Direction of trade ered steadily in 2010, after a sharp fall in the second Of total low-income-country exports to G-20 coun- half of 2008 and early 2009. a Using three-month tries, about two-thirds go to advanced market econo- moving averages (non-seasonally adjusted), their mies and one-third goes to emerging market econo- exports grew from US$15 billion in January 2007 to mies (figure below, right).b But low-income-country US$25 billion in August 2008, fell to US$13 billion exports to emerging market economies were growing a month in early 2009, then reached US$22 billion more quickly before the crisis, fell less abruptly during in October 2010 (figure below, left). the crisis, and in 2010 accounted for more than 40 per- cent of their export growth. Among exports to emerg- Low-income country exports, by destination ing market economies, however, China accounted for 28 about one-third prior to the crisis, but nearly half by late 2010. While China accounted for about one-sixth 24 of low-income-country export growth in the precrisis US$, billions 20 period, it accounted for one-quarter in 2010. 16 12 Contribution to export growth, by destination 40 8 29.6 31.5 30 4 19.4 18.6 20 0 10 4.4 5.8 7.4 5.5 percent Jan Apr Jul Oct Jan Apr Jul Oct Jan Apr Jul Oct Jan Apr Jul Oct 2007 2008 2009 2010 0 –2.2 –10 –4.8 advanced and emerging market economies, including China advanced and emerging market economies, excluding China –20 advanced market economies –21.2 –30 –28.1 Sources: Global Trade Atlas; IMF staff estimates. –40 Note: Based on three-month backward moving averages of imports reported by 2008 2009 2010 Argentina, Australia, Brazil, Canada, China, EU-27, India (estimated for July–October 2010), Indonesia, Japan, Republic of Korea, Mexico, Russian Federation, South China emerging markets Africa, Turkey, and the United States; advanced market economies total Sources: Global Trade Atlas; and IMF staff estimates. GLOBAL MONITORING REPORT 2011 ECONOMIC GROWTH IN DEVELOPING COUNTRIES 51 BOX 2.2 (continued) Product composition Contribution to export growth, by product type 40 Minerals constituted about two thirds of low-income- 30 29.6 31.5 22.2 country exports prior to the crisis, followed by light 20 18.8 manufactures such as textiles and clothing (fi gure 10 percent 5.9 3.8 below and at top right).c Exports of other (“heavy”) 0 1.6 1.5 3.0 1.4 2.2 0.8 manufactures were under 10 percent before the crisis –1.5 –1.7 –0.2 –0.4 – –10 but close to 15 percent by end-2010 (figure at bottom –20 right). Minerals exports were the source of the great- –30 –24.3 –28.1 est changes in exports and remain the largest export –40 category. They accounted for three-quarters of pre- 2008 2009 2010 crisis low-income-country export growth (mainly others heavy manufacturing light manufacturing fuel exports by oil exporters), but were also highly minerals agriculture total vulnerable in the downturn—nearly 90 percent of the Sources: Global Trade Atlas; IMF staff estimates. export contraction in 2009 was in minerals. Miner- als exports remained the largest export item with the highest contribution to export growth in the postcri- Low-income country exports: heavy manufacturing 3.0 sis period: in 2010, about two-thirds of low-income- 2.5 country export growth was in mineral products. US$, billions 2.0 Interestingly, the exit from the crisis is also see- 1.5 ing increased exports of heavy manufactures. These 1.0 rose from March 2009, driven by exports to China. This reflects exports of chemical products, plastics, 0.5 rubber, and metal products, mainly from Asian low- 0.0 Jan Apr Jul Oct Jan Apr Jul Oct Jan Apr Jul Oct Jan Apr Jul Oct income-countries. 2007 2008 2009 2010 Aggregate low-income-country exports will advanced and emerging market economies, including China remain sensitive to demand for minerals and other advanced and emerging market economies, excluding China advanced market economies commodities, an area where continued strong growth in emerging markets will be important. But Sources: Global Trade Atlas; IMF staff estimates. Note: Based on three-month backward moving averages of imports reported there is also tentative evidence that emerging market by Argentina, Australia, Brazil, Canada, China, EU-27, India (estimated for July– growth may be helping some low-income countries October 2010), Indonesia, Japan, Republic of Korea, Mexico, Russian Federation, diversify their export composition. South Africa, Turkey, and the United States. a. The analysis in this box is based on detailed bilateral Low-income country exports, by product type monthly trade statistics provided for 19 of the G-20 countries 28 under subscription by the Global Trade Information Service’s 24 Global Trade Atlas. 20 b. Countries are classifi ed as follows: advanced markets US$, billions refer to 32 countries included in World Economic Outlook 16 classification 110; low-income countries refer to 71 countries 12 eligible for the Poverty Reduction and Growth Trust; and 8 emerging markets refer to all the countries in the sample that 6 are neither advanced markets nor low-income countries. 4 c. Products are classified as follows: agriculture (HS codes Jan Apr Jul Oct Jan Apr Jul Oct Jan Apr Jul Oct Jan Apr Jul Oct 01– 05 [animal and animal products] and 06–15 [vegetable 2007 2008 2009 2010 products]), minerals (25–27 [mineral products]); light manu- agriculture heavy manufacturing minerals factures (16–24 [foodstuffs], 50 – 63 [textiles], and 64 – 67 others light manufacturing [foot wear/headgear]); heavy manufactures (28–38 [chemicals Sources: Global Trade Atlas; IMF staff estimates. and allied industries], 39–40 [plastics/rubbers], 72–83 [met- Note: Based on three-month backward moving averages of imports reported als], 84–85 [machinery/electrical], and 86–89 [transporta- by Argentina, Australia, Brazil, Canada, China, EU-27, India (estimated for July– tion]); and others. October 2010), Indonesia, Japan, Republic of Korea, Mexico, Russian Federa- tion, South Africa, Turkey, and the United States. 52 ECONOMIC GROWTH IN DEVELOPING COUNTRIES GLOBAL MONITORING REPORT 2011 TABLE 2.2 Net financial flows percent of GDP Flows 2007 2008 2009 2010 2011 projection Emerging market economies 12.0 10.4 8.1 8.0 7.9 Private capital flows, net 8.9 6.4 1.7 1.9 2.7 of which: private direct investment 5.3 4.8 3.5 3.3 3.0 private portfolio flows 1.2 –0.9 –0.9 0.5 0.2 Private current transfers 4.3 4.0 3.9 3.8 3.6 Official capital flows and transfers (net) –1.2 0.0 2.4 2.3 1.6 Memorandum item: Reserve assets –3.9 –1.7 –2.4 –2.0 –1.7 Developing countries 13.9 13.2 11.7 10.5 13.2 Private capital flows, net 3.4 1.6 0.8 –0.1 0.6 of which: private direct investment 7.0 6.6 5.4 5.6 5.8 private portfolio flows –1.9 –1.6 –1.6 –1.5 –1.3 Private current transfers 3.9 3.9 3.9 4.1 4.1 Official capital flows and transfers (net) 6.5 7.7 7.0 6.5 8.5 Memorandum item: Reserve assets –4.3 –2.3 –3.4 –2.9 –3.5 Least developed countriesa 12.9 10.5 8.2 9.3 11.8 Private capital flows, net –0.5 –4.3 –3.6 –3.2 –4.1 of which: private direct investment 6.1 5.8 4.6 5.3 5.4 private portfolio flows –2.5 –2.4 –2.6 –2.5 –2.4 Private current transfers 1.9 2.0 2.2 2.5 2.3 Official capital flows and transfers (net) 11.4 12.9 9.6 10.0 13.5 Memorandum item: Reserve assets –4.3 –2.2 –5.0 –4.5 –4.8 Fragile statesb 14.7 6.2 0.1 1.8 10.3 Private capital flows, net –1.5 –7.4 –8.8 –8.7 –6.4 of which: private direct investment 6.1 5.7 3.7 4.5 5.4 private portfolio flows –1.2 –0.9 –0.9 –1.4 –1.8 Private current transfers 1.9 1.7 2.4 3.1 3.0 Official capital flows and transfers (net) 14.3 12.0 6.5 7.5 13.7 Memorandum item: Reserve assets –4.3 –2.8 –7.6 –7.0 –9.9 Sources: World Economic Outlook; IMF staff estimates. Note: Unweighted simple averages. a. United Nations classification, a subset of developing countries. b. World Bank classification, a subset of emerging and developing countries. financial inflows, expressed relative to GDP, Twin-track macroeconomic policies actually fell farther in 2010—suggesting that reflect the uneven economic recovery many recipient countries may not yet be shar- ing in these increases (table 2.2). Net fi nan- In advanced and some emerging market cial flows to developing countries in 2011 are economies, as the impact of fi scal stimulus projected to reach 13 percent of GDP, near wanes and active fiscal adjustment begins, the average for 2007–08. highly supportive monetary policy remains Inflows of FDI are lagging other signs important for a more vigorous recovery. But of recovery, both in emerging market and policy makers must address the social costs developing countries (table 2.2), 6although of continued high unemployment (including increases in workers’ remittances to precrisis high youth unemployment in some countries) levels have partly offset this (table 2.3).7 at the same time.8 Most emerging economies GLOBAL MONITORING REPORT 2011 ECONOMIC GROWTH IN DEVELOPING COUNTRIES 53 TABLE 2.3 Inflows of international remittances US$, billions 2003–07 2008 2009 2010 2011 Source of inflows annual average annual annual estimate projections Emerging market economies 150.2 233.2 224.1 238.3 252.5 Developing countries 28.9 55.4 53.1 56.8 60.9 Least-developed countriesa 12.8 22.9 24.2 25.9 27.9 Fragile statesb 9.0 15.6 14.2 15.7 17.0 Sources: World Bank; IMF staff estimates. Note: Remittances include workers’ remittances, compensation of employees, and migrant transfers. a. United Nations classification, a subset of developing countries. b. World Bank classification, a subset of emerging and developing countries. face different challenges, including incipient Continued accommodative monetary inflationary pressures and, often, a need to policy in most advanced economies has been rebalance growth toward domestic sources. critical to alleviating the fi nancial crisis and Accordingly, among developing countries recession. In contrast, monetary conditions and particularly among emerging market in most emerging and developing countries economies, the macroeconomic policy mix began to normalize in 2010, but remained shifted notably in 2010. Following the com- too accommodative in many cases. While bined monetary and fiscal loosening by most some 80 percent of emerging market econo- emerging market economies in the face of mies undertook monetary policy loosen- contracting global output in 2009, one-third ing in 2009, only about one-third did so in undertook both monetary and fiscal tighten- 2010. Many developing countries continued ing in 2010 (figure 2.11). As global conditions with some monetary loosening (figure 2.12). improved, developing countries also began to Resumed strong growth in emerging market reverse their countercyclical policies. economies underpinned declining measures of FIGURE 2.11 Macroeconomic policy responses to the crisis vary a. Emerging markets b. Developing countries 75 75 60 60 45 45 percent percent 30 30 15 15 0 0 monetary monetary monetary fiscal loosening monetary monetary monetary fiscal loosening and fiscal and fiscal loosening and monetary and fiscal and fiscal loosening and monetary loosening tightening and fiscal tightening loosening tightening and fiscal tightening tightening tightening 2008 2009 2010 2008 2009 2010 Sources: International Financial Statistics; IMF staff estimates. Note: Fiscal conditions are defined based on annual change in government balance as a percent of GDP in 2008, 2009, and 2010. Monetary conditions are based on the change in the monetary conditions index; changes are calculated Q4 over Q4, except for 2010, which uses Q3. The monetary conditions index is a linear combination of nominal short-term interest rate and the nominal effective exchange rate (with a one-third weight for the latter). 54 ECONOMIC GROWTH IN DEVELOPING COUNTRIES GLOBAL MONITORING REPORT 2011 FIGURE 2.12 Monetary policy loosening a. Emerging markets with monetary b. Developing countries with monetary policy loosening policy loosening 80 80 70 70 60 60 50 50 percent percent 40 40 30 30 20 20 10 10 0 0 MCI discount rate exchange rate MCI discount rate exchange rate 2008 2009 2010 2008 2009 2010 Sources: World Economic Outlook; IMF staff estimates. Note: MCI = monetary conditions index. Monetary policy loosening is based on MCI calculations. MCI is a linear combination of nominal short-term interest rate and the nominal effective exchange rate (with a one-third weight for the latter). FIGURE 2.13 Average year-on-year growth in money and the money gap in emerging market economies 25 20 15 M2 percent 10 5 money gap 0 –5 Q1 Q3 Q1 Q3 Q1 Q3 Q1 Q3 Q1 Q3 Q1 Q3 Q1 Q3 Q1 Q3 Q4 2003 2004 2005 2006 2007 2008 2009 2010 Sources: International Financial Statistics; IMF staff estimates. Note: The money gap is the difference between year-on-year growth rates of the monetary aggregate M2 and nominal GDP. The sample includes emerging market economies that have data on both for the whole sample period shown. excess liquidity (such as growth of the money market economies, deficits fell modestly supply relative to nominal GDP), even as and cyclically adjusted deficits were broadly monetary aggregates expanded (figure 2.13). unchanged, suggesting that easy credit con- Fiscal policy has begun to shift from sup- ditions may be discouraging the more rapid porting recovery to cutting deficits.9 The aver- consolidation that could help dampen cur- age deficit among advanced economies fell by rency appreciation and inflationary pressures about 1 percentage point of GDP in 2010. But (figure 2.14).10 because this mainly reflected better growth Low-income countries have also begun to and reduced fi nancial sector support rather reverse the unprecedented countercyclical fis- than narrower cyclically adjusted balances, cal response that helped soften the impact of the need remains for robust, specific fiscal the crisis. As in past downturns, fiscal rev- consolidation plans. Among major emerging enue declined during the crisis. This time, GLOBAL MONITORING REPORT 2011 ECONOMIC GROWTH IN DEVELOPING COUNTRIES 55 FIGURE 2.14 Fiscal deficit, 2008–11 FIGURE 2.15 Low-income countries: fiscal balances and economic crises 0 0 –1 –1 –2 % of GDP % of GDP –2 –3 –3 –4 –4 –5 2008 2009 2010 2011 projected –5 T–1 crisis year (T) T+1 Sources: World Economic Outlook; IMF staff estimates. 2009 crisis 1991 crisis 1982 crisis Note: General government balance data from World Economic Outlook based on Government Finance Statistics Manual 2001 definitions of fiscal balance. Sources: World Economic Outlook; IMF 2010a; IMF staff estimates. however, most of these countries did not The quality of low-income-country curtail spending—indeed, about half of them macroeconomic policies in 2010 was accelerated growth of real primary expen- broadly in line with that in recent years ditures in 2009, undertaking a vigorous countercyclical fiscal response that required Monetary policy, fiscal policy, access to for- additional fi nancing (figure 2.15).11 As rev- eign exchange, and the consistency of macro- enues grew in 2010, fi scal deficits declined economic policies are areas that IMF country by about 1 percentage point of GDP (more desks judged strong in most countries in 2010 among Asian low-income countries), and (figure 2.17).12 Expenditure composition and low-income countries began to rebuild policy public sector governance are seen as areas of buffers, even as they broadly sustained rates relative weakness. Whether the result of an of expenditure growth (figure 2.16). actual deterioration or because the global FIGURE 2.16 Real primary expenditure in low-income countries has increased in favor of public investment and social sectors a. Real primary expenditurea b. Real growth rate of primary expenditureb 130 18 125 15 index, 2007 = 100 120 12 percent 115 9 110 6 105 3 100 0 2007 2008 2009 2010 primary education and capital low-income countries emerging economies expenditure health advanced market economies 2006–08 2009 Sources: World Economic Outlook; IMF 2010a; IMF staff estimates. a. Median. b. Median. 56 ECONOMIC GROWTH IN DEVELOPING COUNTRIES GLOBAL MONITORING REPORT 2011 FIGURE 2.17 Macroeconomic policy quality in low-income countries remained little changed a. Fiscal policy b. Composition of public spending 90 90 80 80 70 70 in each category in each category % of countries % of countries 60 60 50 50 40 40 30 30 20 20 10 10 0 0 unsatisfactory adequate good unsatisfactory adequate good c. Fiscal transparencya d. Monetary policy 90 90 80 80 70 70 in each category in each category % of countries % of countries 60 60 50 50 40 40 30 30 20 20 10 10 0 0 unsatisfactory adequate good unsatisfactory adequate good e. Consistency of macroeconomic policy f. Governance in monetary and financial institutions 90 90 80 80 70 70 in each category in each category % of countries % of countries 60 60 50 50 40 40 30 30 20 20 10 10 0 0 unsatisfactory adequate good unsatisfactory adequate good g. Governance in the public sector h. Access to foreign exchange 90 90 80 80 70 70 in each category in each category % of countries % of countries 60 60 50 50 40 40 30 30 20 20 10 10 0 0 unsatisfactory adequate good unsatisfactory adequate good 2003 2008 2009 2010 Source: IMF staff estimates. a. Available from 2005. GLOBAL MONITORING REPORT 2011 ECONOMIC GROWTH IN DEVELOPING COUNTRIES 57 crisis has enhanced awareness of the inher- price bubbles. Robust domestic demand and ent risks, IMF teams appear to have become the recovery of global output are leading to a more concerned with the adequacy of low-in- rapid closure of output gaps, even as food and come countries’ financial sector governance. commodity prices have returned to precrisis peaks.14 Inflationary expectations are rising and policy targets have been exceeded in a Exiting from the crisis: achieving number of Asian and Latin American coun- sustainable growth tries. Infl ationary pressures could threaten some otherwise sound policy frameworks, Meeting the policy challenges in but some countries’ concerns about further emerging market economies currency appreciation are slowing their mon- Despite their impressive growth and declin- etary policy responses. Strong capital inflows ing poverty rates, emerging market econo- that exacerbate overheating pressures are mies remain home to many of the world’s complicating the policy response. poor people. A continued strong and stable Timid policy responses to overheating economic performance, complemented by would pose risks to the real economy and targeted poverty-reduction policies, will help could derail growth prospects, especially in these countries meet the MDGs. In addition, the large economies. Inadequate responses as these economies have become important to large capital inflows could result in asset trade and fi nancial partners for low-income price overvaluations and an increased risk of countries, this performance will also contrib- a hard landing in credit and property mar- ute to those countries’ MDG progress. kets, possibly triggering broader macroeco- Fiscal balances improved in most emerg- nomic and financial instability (box 2.3). To ing market economies in 2010, driven in part lower risks of a hard landing, many emerging by higher growth. Their overall budget def- economies will need to tighten policies. This icit in 2010 is projected at about 4 percent may include a further removal of monetary of GDP (around 0.75 percent of GDP lower accommodation, further macroeconomic than in 2009), in some cases partly the result prudential tightening, and, often, exchange of one-off receipts. Cyclically adjusted bal- rate appreciation. Some will also need to ances in the six largest economies were only tighten fiscal policies. marginally tighter in 2010, but an adjustment of around 0.5 percent of GDP is expected in Tackling obstacles to growth in 2011. Government debt rose during the crisis low-income countries (especially in emerging Europe), but on aver- age remains at less than 40 percent of GDP The crisis has set back progress toward the (below the average in advanced economies). MDGs in many low-income countries, after The policy challenges of the emerging an extended period of strong economic economies involve overheating and strong growth that had made substantial inroads capital inflows—issues quite different from into poverty. Although growth has since those of other countries. (The advanced picked up and poverty rates are falling again, market economies, for example, must for- recovery at the projected pace will not make mulate and implement detailed fiscal con- up the progress lost because of the crisis. To solidation plans to avoid unsustainable fis- regain momentum, action is needed to boost cal situations, even while they face sluggish growth, emphasizing growth patterns that growth and job creation.) For these econo- benefit subnational regions and sectors where mies, growth is generally strong and debt is poor people are most numerous—and push- manageable,13 although many are at risk of ing ambitious microeconomic and sectoral overheating pressures associated with rapid reforms to promote access to social services credit growth, infl ation, and possible asset and basic infrastructure.15 Economic growth 58 ECONOMIC GROWTH IN DEVELOPING COUNTRIES GLOBAL MONITORING REPORT 2011 BOX 2.3 Emerging market capital inflows: experience and challenges Capital flows to emerging market economies have market economies therefore face the challenge of rebounded with the ebbing of the global financial cri- absorbing the benefits of capital inflows while limit- sis. The largest recipients are Asian and Latin Ameri- ing the attendant macroeconomic and financial stabil- can emerging market economies, South Africa, and ity risks. Turkey. In several countries, net inflows are close to A close look at selected countries that have received all-time highs; on a gross basis, however, total inflows large capital inflows provides useful insights regarding to emerging markets have yet to reach their precri- the effects of and policy responses to the recent epi- sis peak. Compared with other waves of inflows, the sode of inflows (see the figure on the next page). These current episode is characterized by a predominance of countries are facing large capital inflows mainly in the volatile portfolio inflows. Gross inflows have reached form of long-term portfolio debt flows, although com- 6 percent of GDP in only three quarters since the modity exporters also continue to enjoy large direct postcrisis trough; it took three years to reach a similar investment inflows. Despite significant accumula- magnitude in the surge that preceded the global crisis. tion of international reserves, real exchange rates in Portfolio inflows account, on average, for almost one- most cases have appreciated back to precrisis levels— half of inflows (Brazil and the Republic of Korea are although the degree of nominal appreciation has been the top two recipients), much more than in the previ- less pronounced and more varied across countries. ous wave. Direct investment and cross-border bank Surging portfolio inflows helped propel stock and lending are less predominant this time, reflecting lag- bond prices especially in countries with shallower cap- ging economic performance and impaired fi nancial ital markets. While there are limited signs of bubbles intermediation in advanced market economies. so far, cyclical pressures are emerging, with credit to The main pull and push factors behind the recent the private sector picking up strongly in some cases. acceleration of capital flows from advanced to emerg- Macroeconomic policy responses to the current ing economies are improved fundamentals and inflow episode have varied. On the monetary policy growth prospects in emerging market economies side, most countries have begun clawing back the and loose monetary policy in advanced economies. easier monetary policy stance adopted during the From a structural perspective, the global crisis and global crisis. That said, countries have refrained from the more recent jitters in Europe have exposed bal- tightening aggressively, despite emerging inflationary ance sheet vulnerabilities in advanced economies and pressures, out of fear that tightening would pull in appear to have triggered a gradual shift in the portfo- more capital. The fiscal stance has also varied widely lio allocation of institutional investors toward emerg- across countries, but most countries have yet to fully ing market economies, many of which are enjoying unwind the structural loosening adopted during the low debt, proven resilience to shocks, and improved crisis, despite closing output gaps, implying a procy- ratings. From a cyclical perspective, the two-speed clical stance in many cases. nature of the ongoing global recovery will likely keep The countries under review have generally comple- interest rate differentials between emerging market mented macroeconomic policy with other measures to and advanced economies wide for a prolonged period. manage capital inflows. Such measures include taxes Surging commodity prices are an additional cyclical on certain inflows, minimum holding periods, and force pushing capital toward commodity exporters, currency-specific reserve requirements. Recourse to such as Brazil and Peru. In relative terms, more-liquid these measures has been motivated by concerns about emerging markets are attracting larger inflows. All export competitiveness, fi nancial stability, steriliza- things considered, the stage seems set for the ongo- tion costs, and political constraints on fi scal policy. ing wave of inflows to be both large and persistent, Many measures were designed to address specific risks bringing important investment and growth benefits to associated with certain types of flows, such as their emerging market economies. However, inflows have impact on certain asset markets or their short-term tended to reverse suddenly and in a synchronized nature; and to guard against the risk of flow reversal. manner in the past, causing sharp currency deprecia- Evidence to date on their effectiveness in reducing tar- tion and severe balance sheet dislocations. Emerging geted inflows is mixed, though in most cases currency GLOBAL MONITORING REPORT 2011 ECONOMIC GROWTH IN DEVELOPING COUNTRIES 59 BOX 2.3 (continued) appreciation has slowed or halted around the time of system. Beyond this, when confronted with surg- the introduction of the measures. Market participants ing inflows, macroeconomic policies are appropriate have expressed concerns about policy and regulatory tools—namely allowing the currency to strengthen, uncertainty and distortions from measures that go accumulating reserves, and/or rebalancing the mon- beyond macroeconomic policies. Even so, they con- etary and fiscal policy mix. sider the measures implemented so far to be “at the Because they can potentially be used to avoid the margin” and are likely to continue investing in coun- necessary macroeconomic policy adjustments, mea- tries where the positive structural story dominates. sures designed to influence capital inflows (capital The variety of policy responses adopted—and flow management measures or CFMs) could be used their potential multilateral implications—suggests the when appropriate macroeconomic conditions are in importance of developing a broadly accepted frame- place—when the exchange rate is not undervalued, work for countries with open or partially open capi- reserves are more than adequate, and the economy is tal accounts on policies to deal with capital inflows. overheating so that lowering policy rates would not Choosing appropriate responses can be challenging be advisable. If these conditions exist but fiscal policy given the uncertainties associated with the causes and is pro-cyclical, CFMs could be used temporarily to effects of the inflows and with possible policy reac- complement fiscal tightening plans that are already in tions. Even so, a framework for considering appro- place, in view of the lags associated with the macro- priate macroeconomic and other policy responses to economic impact of fiscal consolidation. If CFMs are inflows can be beneficial. Primacy should be given adopted, residency-based measures generally should to structural measures that increase the capacity of be given lower priority, consistent with the general the economy to absorb capital inflows and prudential standard of fairness that countries expect from their measures that enhance the resilience of the financial participation in a multilateral framework. Note: This box presents the main messages from IMF (2011b). Capital Flows and Policy Responses in Selected Emerging Markets Magnitude of Composition of Currency Reserve Monetary Inflation Real credit Fiscal policy Capital flow net inflowsa net inflowsa appreciation increase policy growth Change in management Average in the last Orange = Portfolio flows, % change in Increase in Change in % year over Average % year over cyclically measuresb wave of inflows Gold = Other flows, the NEER from % of GDP policy rates year, average r during year, average of adjusted fiscal r (% of GDP) Green = FDI the trough since from the trough in the recent of last 6 months 2006–08 last 6 months stance between the crisis since the crisis wave 2009–10 Brazil 6.2 38.4 6.0 5.0 (4.5) 12.9 Y Yes Indonesia 2.6 19.4 7.4 6.2 (9.8) 9.2 Y Yes Korea, Rep. 1.9 17.5 10.7 3.3 (3.2) 0.4 Y Yes Peru 5.9 5.6 9.0 2.1 (3.2) 9.3 Y Yes South Africa 6.6 41.4 2.6 3.6 (6.4) –0.1 No Thailand 5.0 9.3 22.3 3.1 (4.1) 4.3 Y Yes Turkey 6.9 6.5 1.7 7.9 (9.6) 21.4 Yes Y Sources: International Financial Statistics; Haver Analytics; GDS; IMF staff calculations. a. Net inflows are defined as the sum of foreign direct investment, portfolio, and other investment balances. Calculations are made for the last wave of capital inflows (2009Q3–2010Q2). b. Capital flow management measures (CFMs) refer to certain administrative, tax, and prudential measures that are part of the policy tool kit to manage inflows. 60 ECONOMIC GROWTH IN DEVELOPING COUNTRIES GLOBAL MONITORING REPORT 2011 MAP 2.1 GDP: Economic growth in 2009 and 2010 has been uneven IBRD 38475 APRIL 2011 GDP growth 2009 Greenland (Den) Faeroe Norway Iceland Islands Finland (Den) Sweden The Netherlands Russian Federation Estonia Canada Isle of Man (UK) Denmark Russian Latvia Fed. Lithuania United Ireland Kingdom Germany Poland Belarus Channel Islands (UK) Belgium Ukraine Luxembourg Moldova Kazakhstan Liechtenstein Romania Mongolia Switzerland France Italy Andorra Bulgaria Georgia Azer- Uzbekistan Kyrgyz D.P.R. Spain Turkey Armenia baijan Turkmenistan Rep. of Korea United States Portugal Monaco Tajikistan Japan Greece Syrian Gibraltar (UK) Cyprus Rep. of Tunisia Malta Lebanon Arab Islamic Rep. Afghanistan China Korea Israel Rep. Iraq of Iran Bermuda Morocco Kuwait (UK) West Bank and Gaza Jordan Bahrain Pakistan Nepal Bhutan Former Algeria Libya Arab Rep. Qatar The Bahamas Spanish of Egypt Saudi Cayman Is. (UK) Sahara Arabia United Arab Bangladesh Cuba Turks and Caicos Islands (UK) India Mexico Emirates Myanmar Lao Haiti Mauritania Oman P.D.R. Belize Jamaica Cape Verde Mali N. Mariana Islands (US) Niger Rep. of Thailand Vietnam Guatemala Honduras Senegal Chad Eritrea Yemen Guam (US) El Salvador Nicaragua The Gambia Burkina Sudan Marshall Faso Djibouti Cambodia Philippines Islands Guinea-Bissau Guinea Benin Federated States of Costa Rica Panama Sri Micronesia R.B. de Guyana Sierra Leone Côte Ghana Nigeria Central Ethiopia Lanka Venezuela Suriname Liberia D’Ivoire African Rep. Brunei Cameroon Somalia Palau Colombia French Guiana (Fr) Togo Malaysia Equatorial Guinea Uganda Maldives Ecuador São Tomé and Príncipe Gabon Congo Rwanda Kenya Singapore Nauru Dem. Rep. of Burundi Seychelles Papua Solomon Congo Indonesia New Guinea Islands Tanzania Comoros Peru Brazil Timor-Leste Angola Malawi Mayotte Zambia (Fr) Vanuatu Bolivia Zimbabwe Mauritius Namibia Madagascar Paraguay Botswana Mozambique Réunion (Fr) New Caledonia Australia (Fr) Swaziland South Lesotho Africa Chile Uruguay Argentina New Dominican Puerto British Virgin Poland GDP growth 2009 Zealand Germany Republic Rico (US) Islands (UK) Ukraine Anguilla (UK) Czech Rep. Slovak Rep. Real GDP growth rate (%) U.S. Virgin Islands Antigua and Barbuda (US) St. Maarten (Neth) St. Kitts Guadeloupe (Fr) Austria Hungary < -6 and Nevis Dominica Slovenia Romania -6 ≤ <0 Curaçao Montserrat (UK) Martinique (Fr) Croatia (Neth) St. Lucia Bosnia and 0≤ <5 Aruba Bonaire St. Vincent and San Herz. Serbia Bulgaria (Neth) (Neth) The Grenadines Barbados Marino Kosovo 5≤ < 10 Montenegro Grenada FYR 10 ≤ Italy Macedonia Trinidad Vatican Albania R.B. de Venezuela and Tobago City Greece no data GDP growth 2010 Greenland (Den) Faeroe Norway Iceland Islands Finland (Den) Sweden The Netherlands Russian Federation Estonia Canada Isle of Man (UK) Denmark Russian Latvia Fed. Lithuania United Ireland Kingdom Germany Poland Belarus Channel Islands (UK) Belgium Ukraine Luxembourg Moldova Kazakhstan Liechtenstein Romania Mongolia Switzerland France Italy Andorra Bulgaria Georgia Azer- Uzbekistan Kyrgyz D.P.R. Armenia baijan Turkmenistan Rep. United States Portugal Spain Monaco Turkey Tajikistan of Korea Japan Greece Syrian Gibraltar (UK) Cyprus Rep. of Tunisia Malta Lebanon Arab Islamic Rep. Afghanistan China Korea Israel Rep. Iraq of Iran Bermuda Morocco Kuwait (UK) West Bank and Gaza Jordan Pakistan Bhutan Bahrain Nepal Former Algeria Libya Arab Rep. Qatar The Bahamas Spanish of Egypt Saudi Cayman Is. (UK) Sahara Arabia United Arab Bangladesh Cuba Turks and Caicos Islands (UK) India Mexico Emirates Myanmar Lao Haiti Mauritania Oman P.D.R. Belize Jamaica Cape Verde Mali N. Mariana Islands (US) Niger Rep. of Thailand Vietnam Guatemala Honduras Senegal Chad Eritrea Yemen Guam (US) El Salvador Nicaragua The Gambia Burkina Sudan Marshall Faso Djibouti Cambodia Philippines Islands Guinea-Bissau Guinea Benin Federated States of Costa Rica Panama Sri Micronesia R.B. de Guyana Sierra Leone Côte Ghana Nigeria Central Ethiopia Lanka Venezuela Suriname Liberia D’Ivoire African Rep. Brunei Cameroon Somalia Palau Colombia French Guiana (Fr) Togo Malaysia Equatorial Guinea Uganda Maldives Ecuador São Tomé and Príncipe Gabon Congo Rwanda Kenya Singapore Nauru Dem. Rep. of Burundi Seychelles Papua Solomon Congo Indonesia New Guinea Islands Tanzania Comoros Peru Brazil Timor-Leste Angola Zambia Malawi Mayotte (Fr) Vanuatu Bolivia Zimbabwe Mauritius Namibia Madagascar Paraguay Botswana Mozambique Réunion (Fr) New Caledonia Australia (Fr) Swaziland South Lesotho Africa Chile Uruguay Argentina New Dominican Puerto British Virgin Poland GDP growth 2010 Zealand Germany Republic Rico (US) Islands (UK) Ukraine Anguilla (UK) Czech Rep. Real GDP growth rate (%) U.S. Virgin Islands Antigua and Barbuda Slovak Rep. (US) St. Maarten (Neth) St. Kitts Guadeloupe (Fr) Austria Hungary < -6 and Nevis Slovenia -6 ≤ <0 Dominica Romania Curaçao Montserrat (UK) Martinique (Fr) Croatia (Neth) St. Lucia Bosnia and 0≤ <5 Aruba Bonaire St. Vincent and San Herz. Serbia Bulgaria This map was produced by the Map Design Unit of The World Bank. (Neth) (Neth) The Grenadines Barbados Marino Montenegro Kosovo 5≤ < 10 The boundaries, colors, denominations and any other information Grenada FYR Trinidad Italy Macedonia 10 ≤ shown on this map do not imply, on the part of The World Bank Vatican Albania Group, any judgment on the legal status of any territory, or any R.B. de Venezuela and Tobago City Greece no data endorsement or acceptance of such boundaries. Source: World Economic Outlook. GLOBAL MONITORING REPORT 2011 ECONOMIC GROWTH IN DEVELOPING COUNTRIES 61 MAP 2.2 Fiscal balance: Countries are diverging in their responses to fiscal challenges IBRD 38476 APRIL 2011 Fiscal balance 2009 Greenland (Den) Faeroe Norway Iceland Islands Finland (Den) Sweden The Netherlands Russian Federation Estonia Canada Isle of Man (UK) Denmark Russian Latvia Fed. Lithuania United Ireland Kingdom Germany Poland Belarus Channel Islands (UK) Belgium Ukraine Luxembourg Moldova Kazakhstan Liechtenstein Romania Mongolia Switzerland France Italy Andorra Bulgaria Georgia Azer- Uzbekistan Kyrgyz D.P.R. Spain Turkey Armenia baijan Turkmenistan Rep. of Korea United States Portugal Monaco Tajikistan Japan Greece Syrian Gibraltar (UK) Cyprus Rep. of Tunisia Malta Lebanon Arab Islamic Rep. Afghanistan China Korea Israel Rep. Iraq of Iran Bermuda Morocco Kuwait (UK) West Bank and Gaza Jordan Bahrain Pakistan Nepal Bhutan Former Algeria Libya Arab Rep. Qatar The Bahamas Spanish of Egypt Saudi Cayman Is. (UK) Sahara Arabia United Arab Bangladesh Cuba Turks and Caicos Islands (UK) India Mexico Emirates Myanmar Lao Haiti Mauritania Oman P.D.R. Belize Jamaica Cape Verde Mali N. Mariana Islands (US) Niger Rep. of Thailand Vietnam Guatemala Honduras Senegal Chad Eritrea Yemen Guam (US) El Salvador Nicaragua The Gambia Burkina Sudan Marshall Faso Djibouti Cambodia Philippines Islands Guinea-Bissau Guinea Benin Federated States of Costa Rica Panama Sri Micronesia R.B. de Guyana Sierra Leone Côte Ghana Nigeria Central Ethiopia Lanka Venezuela Suriname Liberia D’Ivoire African Rep. Brunei Cameroon Somalia Palau Colombia French Guiana (Fr) Togo Malaysia Equatorial Guinea Uganda Maldives Ecuador São Tomé and Príncipe Gabon Congo Rwanda Kenya Singapore Nauru Dem. Rep. of Burundi Seychelles Papua Solomon Congo Indonesia New Guinea Islands Tanzania Comoros Peru Brazil Timor-Leste Angola Malawi Mayotte Zambia (Fr) Vanuatu Bolivia Zimbabwe Mauritius Namibia Madagascar Paraguay Botswana Mozambique Réunion (Fr) New Caledonia Australia (Fr) Swaziland South Lesotho Africa Chile Uruguay Argentina New Dominican Puerto British Virgin Poland Fiscal balance 2009 Zealand Germany Republic Rico (US) Islands (UK) Ukraine Anguilla (UK) Czech Rep. Slovak Rep. (% GDP) U.S. Virgin Islands Antigua and Barbuda (US) St. Maarten (Neth) St. Kitts Guadeloupe (Fr) Austria Hungary < -10 and Nevis Slovenia -10 ≤ < -5 Dominica Romania Curaçao Montserrat (UK) Martinique (Fr) Croatia (Neth) St. Lucia Bosnia and -5 ≤ <0 Aruba Bonaire St. Vincent and San Herz. Serbia Bulgaria (Neth) (Neth) The Grenadines Barbados Marino 0≤ <5 Montenegro Kosovo Grenada FYR 5≤ Italy Macedonia Trinidad Vatican Albania R.B. de Venezuela and Tobago City Greece no data Fiscal balance 2010 Greenland (Den) Faeroe Norway Iceland Islands Finland (Den) Sweden The Netherlands Russian Federation Estonia Canada Isle of Man (UK) Denmark Russian Latvia Fed. Lithuania United Ireland Kingdom Germany Poland Belarus Channel Islands (UK) Belgium Ukraine Luxembourg Moldova Kazakhstan Liechtenstein Romania Mongolia Switzerland France Italy Andorra Bulgaria Georgia Azer- Uzbekistan Kyrgyz D.P.R. Turkey Armenia baijan Turkmenistan Rep. United States Portugal Spain Monaco Tajikistan of Korea Japan Greece Syrian Gibraltar (UK) Cyprus Rep. of Tunisia Malta Lebanon Arab Islamic Rep. Afghanistan China Korea Israel Rep. Iraq of Iran Bermuda Morocco Kuwait (UK) West Bank and Gaza Jordan Pakistan Bhutan Bahrain Nepal Former Algeria Libya Arab Rep. Qatar The Bahamas Spanish of Egypt Saudi Cayman Is. (UK) Sahara Arabia United Arab Bangladesh Cuba Turks and Caicos Islands (UK) India Mexico Emirates Myanmar Lao Haiti Mauritania Oman P.D.R. Belize Jamaica Cape Verde Mali N. Mariana Islands (US) Niger Rep. of Thailand Vietnam Guatemala Honduras Senegal Chad Eritrea Yemen Guam (US) El Salvador Nicaragua The Gambia Burkina Sudan Marshall Faso Djibouti Cambodia Philippines Islands Guinea-Bissau Guinea Benin Federated States of Costa Rica Panama Sri Micronesia R.B. de Guyana Sierra Leone Côte Ghana Nigeria Central Ethiopia Lanka Venezuela Suriname Liberia D’Ivoire African Rep. Brunei Cameroon Somalia Palau Colombia French Guiana (Fr) Togo Malaysia Equatorial Guinea Uganda Maldives Ecuador São Tomé and Príncipe Gabon Congo Rwanda Kenya Singapore Nauru Dem. Rep. of Burundi Seychelles Papua Solomon Congo Indonesia New Guinea Islands Tanzania Comoros Peru Brazil Timor-Leste Angola Zambia Malawi Mayotte (Fr) Vanuatu Bolivia Zimbabwe Mauritius Namibia Madagascar Paraguay Botswana Mozambique Réunion (Fr) New Caledonia Australia (Fr) Swaziland South Lesotho Africa Chile Uruguay Argentina New Dominican Puerto British Virgin Poland Fiscal balance 2010 Zealand Germany Republic Rico (US) Islands (UK) Ukraine Anguilla (UK) Czech Rep. (% GDP) U.S. Virgin Islands Antigua and Barbuda Slovak Rep. (US) St. Maarten (Neth) St. Kitts Guadeloupe (Fr) Austria Hungary < -10 and Nevis Slovenia -10 ≤ < -5 Dominica Romania Curaçao Montserrat (UK) Martinique (Fr) Croatia (Neth) St. Lucia Bosnia and -5 ≤ <0 Aruba Bonaire St. Vincent and San Herz. Serbia Bulgaria This map was produced by the Map Design Unit of The World Bank. (Neth) (Neth) The Grenadines Barbados Marino Montenegro Kosovo 0≤ <5 The boundaries, colors, denominations and any other information Grenada FYR Trinidad Italy Macedonia 5≤ shown on this map do not imply, on the part of The World Bank Vatican Albania Group, any judgment on the legal status of any territory, or any R.B. de Venezuela and Tobago City Greece no data endorsement or acceptance of such boundaries. Source: World Economic Outlook. 62 ECONOMIC GROWTH IN DEVELOPING COUNTRIES GLOBAL MONITORING REPORT 2011 in low-income countries must also be made chapter 5 for more discussion.) Given scarce more resilient so that progress is not derailed concessional financing, however, low-income by future shocks. countries with moderate debt vulnerabilities To achieve accelerated and sustained and effective public finance institutions would growth, low-income countries will require be justified in borrowing on market terms to much higher investment in infrastructure. By finance productive investments, within the raising productivity and encouraging private limits of a sound debt management strategy. investment, closing the present large infra- Stronger domestic revenue mobilization structure gap could substantially increase has a key role in sustainable infrastructure rates of per capita income growth.16 The cost financing. Tax revenue collections in low- of addressing Sub-Saharan Africa’s infra- income countries lag behind those of lower- structure needs is estimated at around $93 middle-income countries by 6–8 percentage billion a year, equivalent to 15 percent of the points of GDP. While low-income countries region’s GDP (or 22 percent of GDP for the have large shares of hard-to-tax informal region’s low-income countries).17 This raises and small-scale agricultural sectors, such the twin challenges of investing efficiently structural factors account for only part of the in infrastructure to get the biggest possible difference. Closing this gap will require long- growth dividend and fi nancing that invest- term efforts. Experience suggests that vigor- ment in a sustainable manner. ous reforms to tax policy and tax administra- Effective public investment management tion can raise revenue ratios by 1 percentage is critical to efficient investing. It begins with points in the first year and, if consistently establishing strategic guidance to anchor pursued, by up to 5 percentage points over government thinking and inform sector- 10 years. Individual country circumstances level decisions, and with independent review vary, but tax policy reforms for a typical low- and analysis of the feasibility of prospective income country would emphasize broaden- projects. In later stages, project selection ing tax bases (such as for value added and processes, budgeting, implementation, proj- corporate income taxes), increasing rates of ect evaluation, and audit are key.18 A new certain environmental and excise taxes, and index of the capacity of 71 developing coun- improving regional coordination on matters tries (including 40 low-income countries) such as corporate taxes.21 Tax policy reforms to appraise, select, implement, and evalu- should be complemented by efforts to secure ate infrastructure projects can guide public taxes from large and medium enterprises and investment management reforms to areas in wealthy individuals, and by other tax admin- which they are needed most.19 Although the istration reforms. index scores for many low-income countries Enhanced market access for low-income- indicate weak capacity throughout the pro- country exports would also support higher cess, this is not universal: Rwanda, for exam- growth and help meet the MDGs. Multilat- ple, scores better than many lower-middle- eral trade reform—especially by concluding income countries. the World Trade Organization Doha Round Financing more infrastructure spending (see chapter 5)—would spur global trade and presents its own challenges, beginning with growth and foster a development-friendly debt sustainability. Infrastructure competes global macroeconomic environment. Most with existing priorities (such as social spend- critically for low-income countries, which ing) and with the new demands of adapting need a stable global trading environment to to climate change.20 Ongoing needs for large- export and to attract investment, the Doha scale grants and highly concessional loans Round would bring the added security of for investment underscore the importance tightened World Trade Organization trade of donors meeting aid commitments, even rules. Further actions by advanced countries as they face fiscal constraints at home. (See and major emerging market economies to GLOBAL MONITORING REPORT 2011 ECONOMIC GROWTH IN DEVELOPING COUNTRIES 63 expand and improve trade preferences for the FIGURE 2.18 The overall debt outlook in low-income countries least-developed countries could further boost seems favorable as fiscal buffers are being rebuilt along with the recovery growth in the poorest countries.22 Fiscal balances and debt 10 70 Making growth resilient in low-income 8 60 countries 6 median, % of GDP 50 With policy buffers drawn down because of 4 % of GDP the global financial crisis, low-income coun- 40 2 tries now face the task of shoring up these 30 0 buffers to boost resilience to future adverse 20 –2 shocks. A gradual consolidation of low-in- –4 10 come countries’ macroeconomic positions over the medium term is projected, with –6 0 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 improving fiscal and current account bal- projected ances, moderate inflation, generally adequate public debt (right axis) primary balance (left axis) reserve positions, and declining debt paths. overall balance (left axis) Almost three-fourths of low-income coun- tries are seen improving their fi scal buffers Source: IMF 2010a. over the next four years. But doing so will require disciplined fiscal and monetary poli- cies, together with vigorous financial and real FIGURE 2.19 Low-income countries projected to cut real spending sector reforms. tend to have high deficits and debt One-third of the projected improvement in the primary balance would come from a 15 90 cyclical recovery in revenues. The balance 70 would come mainly from measures to boost 10 50 underlying revenue performance and trim 5 30 nonpriority spending, some of which are part % of GDP % of GDP of longer-term reform efforts initiated before 10 0 the crisis. Additional donor support to coun- –10 tries with larger projected adjustments could –5 –30 help ease the burden of rebuilding fiscal buf- –50 fers. With low-income countries’ revenue-to- –10 –70 GDP ratios below potential, substantial and –15 –90 sustained increases in fiscal revenues (as out- fiscal balance (left axis) gross debt (right axis) lined above) are critical to rebuilding buffers countries increasing real primary expenditure while making room for priority investments countries reducing real primary expenditure (figures 2.18 and 2.19). For most low-income countries, medium- Source: IMF 2010a. term debt dynamics are not a major concern Note: The average for 2010 is shown. as they rebuild fiscal buffers along the recov- ery path. For example, the median improve- ment in the primary balance is projected at almost half the low-income countries are 1.3 percentage points of GDP over five years. expected to have fiscal deficits of below 2 The share of countries with fiscal deficits in percent of GDP by 2014, compared with one- excess of 5 percent of GDP is projected to quarter in 2009. As growth recovers and drop from almost half in 2009 to one-tenth fiscal situations improve, the median public by 2014. At the other end of the spectrum, debt ratio would decline again: half the low- 64 ECONOMIC GROWTH IN DEVELOPING COUNTRIES GLOBAL MONITORING REPORT 2011 income countries would see public debt fall FIGURE 2.21 Reserve cushions would not below 40 percent of GDP by 2015, while one- improve much among low-income countries with relatively low reserves fi fth would have debt exceeding 65 percent of GDP. The global recovery is projected to boost Reserve coverage by initial coverage, 2009 6 demand for exports and strengthen current prospective imports of goods account balances, helping bolster reserve 5 median, months of cushions. The projected improvement in 4 and services the external environment would help main- tain median reserve coverage at around the 3 equivalent of four months of imports over the 2 medium term, although country experience 1 would differ: one in seven low-income coun- tries would have reserve coverage exceeding 0 2009 2010 2011 2012 2013 2014 six months of imports, while one in 20 would projected have less than two months. high 2009 coverage (38 countries) Exchange rate policies should continue to medium 2009 coverage (8 countries) be used to cushion the effect of future volatil- low 2009 coverage (18 countries) ity where possible. Many low-income coun- tries with fi xed exchange rate regimes could Source: IMF 2010a. benefit from somewhat faster accumulation Note: “Low” is defined as having less than three months of reserves cover- age; “medium” as having more than three but less than four months of of reserves. Conversely, some with floating reserves coverage. rates appear to have built more than adequate reserve cushions; they could afford to raise spending (figures 2.20 and 2.21). continue, but they must closely monitor the Most low-income countries can continue effects of commodity prices on their domestic using monetary policy to support the recov- inflation rates, given the risks associated with ery if the present moderate rates of inflation rising world prices for food and fuel. If these global shocks persist and feed through to local prices, monetary policy should accom- FIGURE 2.20 The current account balance is modate the direct impact; however, it may projected to improve gradually in low-income need to be tightened in some cases to counter countries as exports rebound second-round effects. The crisis has underscored the impor- 38 0 tance to low-income countries of adequate 36 –1 financial regulatory frameworks, effective 34 –2 supervision, and sound financial institutions. average, % of GDP 32 Supervisory authorities will need to ensure –3 % of GDP 30 that credit standards do not deteriorate dur- –4 28 ing times of strong credit growth. Regula- –5 26 tory frameworks should focus on the risks 24 –6 assumed by banks and the sources of their 22 –7 business growth to ensure that these are sus- 20 –8 2007 2008 2009 2010 2011 2012 tainable. Closer supervision is required to projected ensure that banks are complying with pru- exports of goods and services (left axis) dential regulations. In banks where financial current account balance, including FDI strains have been significant, the balance (right axis) sheet cleanup should proceed quickly— recognizing losses and having shareholders Source: IMF 2010a. inject needed capital. GLOBAL MONITORING REPORT 2011 ECONOMIC GROWTH IN DEVELOPING COUNTRIES 65 Developing domestic debt markets would which will require both low-income countries help mobilize national savings and increase and their trading partners to undertake fur- low-income country policy buffers, helping ther reforms of their trade regimes. Barriers cushion the impact of the crisis. Low-income in labor and product markets also impede countries also have enormous investment the emergence of new sectors and the entry needs that require financing. Although exter- of new firms; addressing these barriers would nal financing has to remain a significant part promote diversification. To the extent that of the financing mix, policies to mobilize transformation in economic structures leads domestic savings and develop domestic debt to social dislocation, it would be important markets would broaden the range of available to ensure that effective social safety nets are options. in place to protect vulnerable groups (see the Reforms that promote economic diversifi- discussion of indigenous peoples and socially cation also have an important role to play in excluded groups in chapter 4). Fragile states managing macroeconomic volatility and fos- face special challenges (box 2.4; also see fig- tering durable growth.23 Such diversification ure 1.7 and box 1.4 in chapter 1 and box 5.4 is likely to involve further trade integration, in chapter 5). BOX 2.4 Fragile states—experience and implicationsa Fragile states are generally off target to meet the characteristics contribute to a cycle of underdevelop- MDGs. They are typified by poor initial condi- ment that reinforces political instability and conflict, tions (left figure below), slow GDP growth (right with negative regional spillovers. figure below), and macroeconomic instability. These Growth of real GDP and consumption, Initial conditions fragile states and nonfragile states, 1970–2009 3.57 5 56 4.47 0.23 2.82 4 0.09 3 2.87 percent 1.78 38 2 1.24 1.52 1 0.28 0 CPIA 2008 PIMI % of coastal civil conflict, mean of real private mean of real GDP growth institutions countries average consumption growth geography conflict non-fragile states/low-income countries non-fragile states/low-income countries fragile states/low-income countries fragile states/low-income countries Source: IMF staff calculations. Source: IMF staff calculations. Note: CPIA = Country Policy and Institutional Assessment; PIMI = public investment management index. continued 66 ECONOMIC GROWTH IN DEVELOPING COUNTRIES GLOBAL MONITORING REPORT 2011 BOX 2.4 Fragile states—experience and implications (continued) The quality of institutions (measured by the Implications for global policy Country Policy and Institutional Assessment and International organizations and development part- by the index of public investment management men- ners have recognized that the problems of fragile tioned above) tend to be considerably weaker in frag- states present distinct challenges and warrant well- ile states than in other low-income countries.b Fragile targeted approaches. Concerned about preserving states are also characterized by lower levels of devel- peace and security in these countries and aware of opment and a much higher incidence of conflict.c the risk of spillovers to neighbors, they have made considerable efforts to develop effective paradigms Slow real GDP growth for engagement. • During 1970–2009, average real GDP growth in Their key challenge for effective engagement is to fragile states was about one-fi fth that of other low- devise strategies that recognize these states’ capacity income countries. The difference in the growth constraints, their large fi nancing needs, and the need performance is especially apparent since the early for prolonged engagement. Furthermore, given their 1990s. multidimensional needs, fragile states would benefit • While the number of growth upswings was broadly from a well-coordinated and comprehensive strategy similar, growth downswings are twice as frequent of assistance from the international community. This in fragile states. strategy should emphasize the paramount need to • Output losses that follow economic and political secure peace and security; the need to rebuild capac- shocks are larger and persist longer than in other ity and strengthen institutions through a common low-income countries. technical assistance strategy; and the design, scale, and timing of macroeconomic adjustment that can Macroeconomic instability secure a virtuous cycle of development. • Measured by high inflation, large debt-to-GDP a. The IMF does not maintain a formal list of fragile states, ratios, and low international reserves, macroeco- and the analysis here uses the World Bank list. nomic instability has been much higher in fragile b. The average Country Policy and Institutional Assessment states than in other low-income countries. score for fragile states in 2008 was 2.82, compared with 3.57 for nonfragile low-income countries. c. This is indicated by lower per capita GDP and investment ratios and by weaker health and education indicators. Boosting the impact of international growth domestically and to avoid possible support overheating. Second, actions are needed to help low- To regain momentum toward the MDGs, income countries achieve and sustain more international cooperation will be required rapid economic growth and restore their on three fronts. First, low-income countries policy buffers. Inadequate infrastructure is in particular will need a strong and stable one major challenge for growth. Address- global economic environment in which to ing the infrastructure gap requires stronger continue growing. To secure such an environ- public investment management systems and ment, the advanced market economies need enhanced domestic revenue mobilization— to repair and reform their fi nancial systems areas where international organizations can and address their fiscal imbalances, and offer policy advice and technical assistance. emerging market economies should adjust It also requires large-scale, highly conces- their macroeconomic policies to reorient sional finance, underscoring the need for GLOBAL MONITORING REPORT 2011 ECONOMIC GROWTH IN DEVELOPING COUNTRIES 67 donors to meet aid commitments and to educational achievement and social cohesion strengthen the overall effectiveness of aid. (ILO-IMF 2010). Policy makers can take Another major challenge is to expand trade the following steps to improve employment opportunities, particularly for the poorest opportunities (including where young adult countries. Key steps include completing the unemployment runs high): establishing more Doha Round and expanding and improving flexible labor market regulations together Aid for Trade.24 with effective social safety nets; implementing Third, fragile states lag farthest behind in programs to upgrade skills and better match reaching the MDGs, and they require addi- them with the demands of labor markets; and tional support to help build institutions and strengthening the business environment (IMF move toward a virtuous circle of develop- 2010d). ment, peace, and security. 9. IMF 2011c. 10. In addition to facilitating greater price stabil- ity, there is evidence that fiscal adjustment in Notes emerging economies has typically helped pro- 1. Except where otherwise noted, this chap- mote income equality in the longer term. See ter distinguishes “advanced” and “emerging IMF (2010b), appendix 3. and developing” economies according to the 11. More than half of the additional deficit was IMF World Economic Outlook country clas- financed domestically, including borrowing in sification. Within the latter group, “emerging domestic debt markets, central bank fi nanc- economies” are those not eligible for the IMF ing, or drawing down government deposits. Poverty Reduction and Growth Trust (PRGT) External borrowing accounted for much of (see http://www.imf.org) and “low-income the rest, with the IMF financing a significant countries” (or “developing economies”) are component. those eligible for the PRGT. The analysis in 12. IMF country desks were asked in January box 2.4 is based on the World Bank definition 2011 to assess the quality of macroeconomic of fragile states. Least-developed countries are policies in 2010 as unsatisfactory, adequate, a subset of low-income countries, as defi ned or good. by the United Nations. 13. Recovery has been very strong in emerging 2. The January 2011 Global Economic Pros- Asia and Latin America and more restrained pects (World Bank 2011a) examines the pov- in the Middle East and emerging Europe, erty implications of higher food prices (box where output in some countries remains well 6 of that volume) and explores links between below precrisis levels. financial markets and food prices, finding 14. Emerging and developing economies’ export little evidence that investment-based demand volumes grew 13 percent in 2010 and are has driven up prices (box 7). expected to remain buoyant in 2011. 3. In 2009, increases in the ratios of reserves 15. IMF 2010c. to imports reflected the IMF special draw- 16. Calderon (2009) estimates that by increasing ing rights (SDRs) allocation (see the factsheet the stock and quality of their infrastructure to on SDRs at http://www.imf.org) and 2009’s that of Mauritius, Sub-Saharan low-income sharp contraction in global trade. countries would raise their annual per capita 4. IMF 2010a. income growth by more than 2 percentage 5. World Bank 2011a. points. 17. Briceno-Garmendia, Smits, and Foster 2008. 6. World Bank 2011b, table 2. 18. Rajaram et al. 2010. 7. World Bank 2011b. 19. Dabla-Norris et al. 2010. 8. In addition to immediate income loss from not 20. Brumby and Verhoeven (2010) note that working—loss for society and for the individ- though government investment spending ual—the considerable human costs of unem- among low-income countries accelerated in ployment include long-lasting income loss, 2009–10, fiscal consolidation pressures may increased illness and mortality, and reduced reduce its growth over the next few years. 68 ECONOMIC GROWTH IN DEVELOPING COUNTRIES GLOBAL MONITORING REPORT 2011 21. IMF 2011a. (August). http://siteresources.worldbank.org/ 22. Elborgh-Woytek, Gregory, and McDonald INTPREMNET/Resources/EP25.pdf. 2010. ILO-IMF (International Labour Organization- 23. Haddad, Lim, and Saborowski 2010. International Monetary Fund). 2010. “The 24. Hoekman and Wilson 2010; see chapter 5 of Challenges of Growth, Employment and this volume. Social Cohesion.” Paper prepared for the Joint ILO-IMF Conference in cooperation with the References Office of the Prime Minister of Norway, Oslo, Norway, September 13. Briceno-Garmendia, C., K. Smits, and V. Fos- IMF (International Monetary Fund). 2010a. ter. 2008. “Financing Public Infrastructure “Emerging from the Global Crisis: Macroeco- in Sub-Saharan Africa: Patterns, Issues, and nomic Challenges Facing Low-Income Coun- Options.” Africa Infrastructure Country Diag- tries.” Washington, DC. http://www.imf.org/ nostic Background Paper No. 15, World Bank, external/np/pp/eng/2010/100510.pdf. Washington, DC. ——— . 2010b. Fiscal Monitor. Fiscal Exit: From Brumby, J., and M. Verhoeven. 2010. “Public Strategy to Implementation. Washington, DC. Expenditure after the Global Financial Crisis.” http://www.imf.org/external/pubs/ft/fm/2010/ In The Day after Tomorrow: A Handbook on fm1002.pdf. the Future of Economic Policy in the Devel- ——— . 2010c. “Reaching the MDGs: Macroeco- oping World, ed. O. Canuto and M. Giugale, nomic Prospects and Challenges in Low-Income 193–206. Washington, DC: World Bank. Countries.” Washington, DC. http://www.imf. Calderon, C. 2009. “Infrastructure and Growth org/external/np/exr/mdg/2010/091610.pdf. in Africa.” Policy Research Working Paper ——— . 2010d. Regional Economic Outlook: 4914, World Bank, Washington, DC. Middle East and Central Asia. October. Wash- Dabla-Norris, E., J. Brumby, A. Kyobe, Z. Mills, ington, DC. and C. Papageorgiou. 2010. “Investing in Pub- ——— . 2011a. “Revenue Mobilization in Devel- lic Investment: An Index of Public Investment oping Countries.” February. Efficiency.” Working Paper 11/37, Interna- ———. 2011b. “Recent Experiences in Managing tional Monetary Fund, Washington, DC. Capital Inflows—Cross-Cutting Themes and Elborgh-Woytek, K., R. Gregory, and B. McDon- Possible Policy Framework.” ald. 2010. “Reaching the MDGs: An Action ———. 2011c. “Strengthening Fiscal Credibility.” Plan for Trade.” Staff Position Note 10/14, Fiscal Monitor Update January. http://www. International Monetary Fund, Washington, imf.org/external/pubs/ft/fm/2011/01/update/ DC. fmindex.pdf. Haddad, M., J. J. Lim, and C. Saborowski. 2010. Rajaram, A., T. Minh Le, N. Biletska, and J. “Managing Openness and Volatility: The Role Brumby. 2010. “A Diagnostic Framework for of Export Diversification.” Poverty Reduction Assessing Public Investment Management.” and Economic Management Network, World Policy Research Working Paper 5397, World Bank. Economic Premise 6 (March). http:// Bank, Washington, DC. siteresources.worldbank.org/INTPREMNET/ World Bank. 2010. Global Monitoring Report: Resources/EP6.pdf. The MDGs after the Crisis. Washington, DC. Hoekman, B., and J. Wilson. 2010. “Aid for ———. 2011a. Global Economic Prospects: Nav- Trade: An Action Agenda Looking Forward.” igating Strong Currents. Washington, DC. Poverty Reduction and Economic Management ——— . 2011b. Migration and Remittances Fact- Network, World Bank. Economic Premise 25 book, 2011. Washington, DC. 3 Linking Spending and Outcomes: Some Lessons from Impact Evaluations in Education and Health E ducation and health are key dimen- difficult to improve the quality of education sions of many of the Millennium and health services, which remain low in Development Goals (MDGs). As 2015 many developing countries. Hence, the dis- approaches, calls for greater development connect between development spending and effectiveness are taking on increased urgency learning and health outcomes. The discon- for two reasons: development assistance for nect stems in part from the narrow focus on health and education has risen to unprece- financing inputs, while ignoring other parts dented amounts, but has not led to expected of the causal chain that links public spend- improvements in outcomes; and the global ing to changes in outcomes. The causal chain crisis has forced a reexamination of social shows that incentives facing both service pro- spending. viders and consumers are important mediat- This chapter highlights lessons about this ing variables in the link between development disconnect between spending and outcomes spending and outcomes, and policies have derived from recent impact evaluations in often failed to sufficiently account for them. health and education. Although designing Not surprisingly, therefore, one of the strong the right approach to improving outcomes is lessons emerging from impact evaluations is complex and context dependent, a system- that continuing to increase provision of tra- atic evidence base generated through impact ditional inputs will have only limited impact evaluations can provide useful guidance for if issues of service delivery and consumer policy makers. Furthermore, the body of behavior are not addressed. As highlighted in these evaluations has grown tremendously in the 2004 World Development Report: Mak- recent years. The base of impact evaluations ing Services Work for Poor People, improv- is still far from complete, and much more ing service delivery in health and education, work is needed; but some interesting findings in addition to increasing resources, is vital to have already begun to emerge. reach many MDGs. Existing impact evaluations show that Although this chapter points to the limited efforts aimed at improving access to educa- impact of input provision, it does not ques- tion and health services have met with some tion the need for additional financing because success. However, it has been much more improving quality also requires resources. It Girls from Lagos, Nigeria, are eager to get on the school bus. 71 72 LINKING SPENDING AND OUTCOMES GLOBAL MONITORING REPORT 2011 is important to ensure that institutions pro- Despite these huge funding gains, progress vide services efficiently and responsively— toward development targets has been uneven. and that their potential clients have the Lack of resources only partly explains many ability and desire to use services efficiently of the remaining gaps because links between and hold service providers accountable for spending and human development outcomes quality. And when quality and institutional are weak,1 and good policies and strong insti- arrangements improve, the returns to addi- tutions are central to improving the produc- tional investments are also greater. This tivity of education and health spending. focus on institutional arrangements is con- True, access to education and health ser- sistent with the findings presented in chapter vices has expanded sharply in recent years. 1, which highlights policies and institutions For example, basic immunization coverage in explaining the heterogeneity of MDG in low-income countries increased by 20 per- progress. centage points during 2000–08.2 Primary and secondary school enrollment rates have also improved, in some cases dramatically. But the Disconnect between expenditure quality (or outcomes) of education and health and outcomes remains a grave concern. In education, this is Budget allocations for health and educa- made evident by the poor learning outcomes of tion are at an all-time high in the developing school children.3 Often, countries that spend world. Development assistance for health has more on primary education (compared with quadrupled since 1990, peaking at $24 bil- the level predicted by per capita income) gen- lion in 2008 (figure 3.1a). Health spending erate on average only a small improvement in by developing-country governments has also test scores (compared with scores predicted by peaked (albeit at a much lower level), nearly per capita income) (figure 3.2). In India, even doubling to reach $240 million during 1995– though most children of primary-school age 2006. Similarly, development assistance for were enrolled in school, 35 percent of them education has doubled since 2002, reaching a could not read a simple paragraph and 41 per- high of $10.8 billion in 2007 (figure 3.1b). cent could not do a simple subtraction.4 FIGURE 3.1 Government and donor spending on health and education is unprecedented a. Spending on health b. Spending on education 30 1,000 12 900 25 800 10 US$ billions, 2006 US$ billions, 2006 US$ billions, 2007 20 700 8 600 15 500 6 400 10 300 4 5 200 2 100 0 0 0 1990 1993 1996 1999 2002 2005 2008 2002 2003 2004 2005 2006 2007 donor health assistance government health donor education assistance aid to basic education (left axis) expenditure (right axis) Sources: IHME 2010; OECD-DAC 2009, cited in UNESCO 2010. Note: Includes all on-budget and off-budget health-related disbursements from bilateral donor agencies, grant-making and loan-making institutions, United Nations agencies, and nongovernmental organizations. Development assistance to education comes not only as direct allocations to the education sector but also through general budget support. GLOBAL MONITORING REPORT 2011 LINKING SPENDING AND OUTCOMES 73 FIGURE 3.2 The relationship between FIGURE 3.3 There are wide differences in the share of public expenditure and learning outcomes is weak health spending going to the rich and the poor 20 35 share of public health expenditures, % (controlling for GDP per capita) normalized primary test score 15 30 10 25 5 20 0 –5 15 –10 10 –15 5 –20 0 –15 –10 –5 0 –5 –10 –15 Sub-Saharan South Europe and East Asia Latin America spending on primary education Africa Asia Central Asia and Pacific and the (controlling for GDP per capita) Caribbean poorest quintile richest quintile Source: Bruns, Filmer, and Patrinos 2011. Note: The graph shows the deviation of normalized test scores from that predicted by GDP per capita against the deviation of public spending Source: Authors’ calculations, based on data from Filmer 2003. on primary education per student (relative to GDP per capita) from that predicted by GDP per capita. In health, basic immunization rates have health spending than do poor people (figure expanded, but full immunization coverage— 3.3). In Sub-Saharan Africa, South Asia, and needed for the full impact on child mortal- Europe and Central Asia, the incidence of ity—is much lower. 5 Likewise, the global health spending among the richest fifth of average attendance at a minimum of one the populace was more than twice that of the prenatal visit is 78 percent, but attendance poorest fifth. at four visits—needed for the full impact on So, what are some of the factors that cause maternal mortality—is around 48 percent.6 the disconnect between public spending and Clinical competence may also be lacking (as changes in human development outcomes? in Senegal, where an assessment of service quality found that only 34 percent of cases were correctly diagnosed).7 Understanding the disconnect So, more resources—even when increasing Calls for greater development effectiveness coverage—may not be enough to have a big are implicitly calls to strengthen the links in effect on outcomes. the causal chain between spending on inputs The disconnect between spending and and human development outcomes. Each link outcomes partly reflects the failure of human of the chain, if weak, can prevent increased development spending to reach poor people. public fi nancial and physical resources from Public spending on health and education is being translated into improvements in human generally believed to be an effective means to development outcomes: the link between reach the poor, but empirical evidence fails resources allocated and services and goods to support this assumption. Since the 1990s, generated, between goods and services gen- numerous studies have found that public erated and consumption, and between con- health spending is not concentrated among sumption of goods and services and educa- the poor and that the rich benefit dispropor- tion outcomes (figure 3.4). The first link tionately from public health subsidies.8 Apart emphasizes supply-side factors, the third link from Latin America and the Caribbean, the demand-side factors. The middle link is the rich receive far greater benefits from public critical intermediate step where the demand- 74 LINKING SPENDING AND OUTCOMES GLOBAL MONITORING REPORT 2011 FIGURE 3.4 Links between public expenditure and human the challenges of service delivery are often development outcomes more pronounced in hard-to-reach areas and underserved populations. Public financial and The service delivery triangle in figure 3.4 physical resources outlines the relationships among different system actors. Citizens, especially poor peo- ple, who ultimately consume the education and health services generated by the public system are the clients. They have a direct relationship with frontline service providers, Goods and services such as teachers in public schools and health Service generated providers care workers in public health facilities—the short route of accountability. Crucially, how- Policy ever, the service providers generally have makers no direct accountability to the consumers, unlike in a market transaction. Instead, they Clients Consumption of are accountable only to the government that goods and services employs them. The accountability route from consumers to service providers is therefore through the government—the long route. To hold service providers accountable for the quantity and quality of services provided, Human citizens must act through the government— development a process that is difficult for poor people especially because they can seldom organize Sources: Adapted from Go 2010 and World Bank 2003. themselves and be heard by policy makers. Note: The service delivery triangle showing the interaction among clients, service providers, and Moreover, the government rarely has enough policy makers has been adapted from the 2004 World Development Report (World Bank 2003). This information—or indeed the mechanisms—to diagram is a simplification of very complex relationships among spending, supply, service delivery, demand, and outcomes. There are nuances that are not captured here. One such nuance is that improve service provider performance. services provision implies use, whereas the generation of goods does not. Another is that different Demand-side factors also mediate between service providers (government, nongovernment, faith-based) have different motivations and incen- tives, so their interaction with clients and policy makers and the outcomes will be different. See, for public spending and human development out- example, Reinikka and Svensson (2010). Finally, incentives created by different financing arrange- comes. The lack of uptake from consumers ments (including prepayment mechanisms) and their impact on provider and consumer behavior are not fully captured here. resulting from the lack of demand, informa- tion, or other factors can undermine the posi- side and supply-side factors intersect for the tive impacts of public spending in health and delivery of health and education services. education. On the supply side, inequitable resource All this suggests that merely increasing the allocation, lack of complementary inputs supply of human development inputs and ser- (both sectoral and cross-sectoral), and vices will not automatically improve human crowding out of private sector provision can development outcomes. Thus it is that the weaken the link between public spending pervasive approach in global resource mobili- and the amount of goods and services gen- zation efforts (that is, filling narrowly defined erated. And the consumption of these goods resource gaps through predominantly supply- and services will depend critically on the side interventions without strengthening ser- quality of service delivery, a crucial interme- vice delivery and consumer uptake) has lim- diate step where the results chain frequently ited impact. breaks down. This can often be traced to poor incentives for service providers and low accountability. The weak service delivery What do impact evaluations say? link can also reinforce existing inequalities By understanding the causal pathways, prac- in human development outcomes because titioners can be better placed to improve GLOBAL MONITORING REPORT 2011 LINKING SPENDING AND OUTCOMES 75 development effectiveness and thus accelerate settings, different forms of the intervention, progress toward the MDGs (and secure more and varying levels of rigor in the evaluation. value for money). Also, very few impact evaluations inform us Impact evaluations have emerged as an about the longer-term impact of programs or important tool to conduct such analyses. whether observed impacts persist after the They assess the changes in the well-being of program has ended. Another frequently heard individuals that can be directly attributed to critique of impact evaluations is that they do a particular project, program, or policy—and not focus enough on the reasons why certain thereby provide robust and credible evidence projects work and others do not.11 It has also on how a project has performed and what been argued that impact evaluations do not changes they have “caused.” At a broader lend themselves easily to distributional anal- level, the information they generate can con- ysis because they are focused on estimating tribute to a systematic evidence base on the average treatment effects.12 Finally, very few kinds of interventions that work. Rigorous, impact evaluations collect or report on pro- project-specific impact evaluations can also gram cost data; and this represents a missed shed light on the broader functioning of the opportunity to gain insights into efficiency, education and health systems and on general cost effectiveness, as well as longer-term household behavior. They often have forced a financing issues such as sustainability. rethinking of assumptions regarding human Many of these objections can be resolved behavior (as illustrated in box 3.1) or brought by expanding the base of rigorous impact to light unexpected impacts of certain types evaluations on certain topics. Researchers of development interventions (as illustrated in should replicate successful models in differ- box 3.2). ent contexts to assess the extent to which Because of the explicit focus on causal- programs work under varying circumstances. ity, the main methodological requirement for In addition, replicating interventions that an impact evaluation, and what sets it apart have been successful in small-scale settings from other types of monitoring and evalua- at a regional or national level will help con- tion, is the need for a credible counterfactual tribute to knowledge of what works in the (that is, what the outcome for project benefi- real world. In certain cases, such as incen- ciaries would have been without the project).9 tives and accountability, there is a need to Establishing a counterfactual retrospectively test various combinations and extensions of is hard. For this reason, it is best to design the existing approaches. Conducting a series of impact evaluation alongside the program and related evaluations in a comparable setting to build it into the program. with cross-cutting treatment designs gener- Some caveats regarding impact evaluations ates considerable cost savings in data col- need to be noted. Although extremely useful, lection and allows cost-effectiveness com- impact evaluations cannot answer all devel- parisons across program variants. It is also opment questions. Furthermore, the exist- important to collect cost and process data, ing knowledge base of impact evaluations in data on intermediate outcomes, and good education and health is far from complete. qualitative data to complement quantitative In many areas, the current evidence does data. The World Bank’s Development Impact not support as clear an answer as one would Evaluation (DIME) initiative is encouraging like. In others, it does not answer some key the use of impact evaluations and promoting questions. The results of many individual their integration into government programs impact evaluations cannot be generalized. (box 3.3). In fact, the question of external validity of Although this chapter focuses on impact impact evaluations has been an important evaluations, such evaluations are not the and often somewhat contentious one.10 For- only source of evidence concerning effective mulating general recommendations from the interventions to advance the education and conclusions of impact evaluations is often health MDGs. Other evaluation methods hampered by lack of comparability among also contribute to this debate—such as mixed 76 LINKING SPENDING AND OUTCOMES GLOBAL MONITORING REPORT 2011 BOX 3.1 Some impact evaluations have forced a rethinking of our assumptions about HIV/AIDS Development interventions are often designed on the agree with a person’s belief of HIV infection, there basis of certain assumptions, particularly for HIV/ is no statistically significant change in contracting a AIDS—an area where knowledge has grown by leaps sexually transmitted infection. Using the distribution and bounds, especially over the past decade. Impact of beliefs of HIV infection and prevalence from the evaluations have contributed to that learning, and study, the evaluation fi nds the overall number of HIV some evaluations have shown how erroneous these infections to increase by 25 percent when people are assumptions can be. Here are a few examples. tested, compared with when they are unaware of their status—an unintended consequence of testing. Assumption: Promoting knowledge of HIV/AIDS will reduce risky behavior Assumption: It is usually stigma that prevents people from picking up HIV test results Information and education campaigns have been the main focus of HIV/AIDS prevention campaigns. It is often argued that getting people to learn their Although they have improved knowledge and self- HIV status is crucial for fighting HIV/AIDS, but that reported behavior, their apparent impact on behaviors stigma and fear of obtaining a positive result create may actually be the result of a reporting bias fueled a major barrier preventing people from picking up by the people’s greater knowledge of what they would their results at testing centers. An impact evaluation need to do to reduce their risk, rather than a reflection in Malawi found, however, that small incentives and of substantial changes in actual behaviors. Two recent deadlines were enough to induce people to do this.c meta-analytic reviews show that only a small fraction Distance to the center was also a key determinant of behavior change interventions had an impact on of attendance. These fi ndings suggest that procras- HIV or other sexually transmitted infections.a Infor- tination and the inconvenience of travel, rather than mation needs to be better targeted and packaged to stigma, explain much of the problem. have an impact on actual behavior. Nor is more infor- mation necessarily better. Consider a program that sent text-message reminders concerning adherence to Assumption: People who test positive will antiretroviral treatment: weekly reminders improved reduce their risky behavior adherence, daily reminders did not. People got habitu- Although people can behave altruistically once they ated to them or perhaps considered them intrusive. know they are HIV-positive, they may also show disinhibition behavior because they have less to lose when engaging in risky behavior. A recent study in Assumption: More HIV testing will lead to Mozambique d found that risky sexual behaviors declines in risky behavior increase in response to the perceived changes in risk HIV testing is one of the key policy responses to the associated with greater access to antiretroviral ther- HIV/AIDS epidemic in Africa, but there is little rig- apy. So, scaling up access to antiretroviral therapy orous evidence on how individual sexual behavior without prevention programs may not be optimal if responds to testing. Impact evaluations from East the objective is to contain the disease because people Africa produce three findings.b First, individuals would adjust their sexual behavior in response to the surprised by an HIV-positive test are more than five perceived changes in risk. times more likely to contract a sexually transmit- ted infection than are members of a similar untested control group, indicating an increase in risky sexual behavior. Second, individuals who believed they were a. McCoy, Kangwende, and Padian 2009; Mavedzenge, at high risk for HIV have a 60 percent decrease in Doyle, and Ross 2010. their likelihood of contracting a sexually transmit- b. Weinhard et al. 1999; Gong 2010. ted infection following an HIV-negative test, indi- c. Thornton 2005. cating safer sexual behavior. Third, when HIV tests d. De Walque, Kazianga, and Over 2010. GLOBAL MONITORING REPORT 2011 LINKING SPENDING AND OUTCOMES 77 BOX 3.2 Spillovers and unintended effects can be important Impact evaluations have helped highlight positive with artemisinin-based combination therapy, the spillovers of certain types of human development only remaining effective class of first-line antima- interventions. In Kenya, providing deworming drugs larial drugs—may result in increased resistance of at treatment schools lowered the incidence among the malaria parasite to those drugs and overtreat- nontreated children as a result of a fall in transmis- ment of individuals not infected with malaria who sion rates. a Spillover effects are also shown to be are presumptively diagnosed (and the resultant nega- strong in the provision of AIDS antiretroviral treat- tive health outcomes because the true cause of illness ment. Also in Kenya, children’s weekly hours of remains untreated). There are also concerns over the school attendance increased more than 20 percent sustainability of the subsidy (as a result of inefficient within six months of initiating antiretroviral treat- targeting). Cohen, Dupas, and Schaner (2010) con- ment for any adult household member.b For boys in fi rmed the fears of overtreatment and showed that treatment households, these increases closely followed subsidized access to malaria rapid diagnostic tests their reduced supply of paid labor. Similarly, young substantially reduced these unintended consequences. children’s short-term nutritional status—measured as Another example of a negative spillover effect is that weight for height—also improved dramatically. some CCT programs show that, in some contexts, Although Mexico’s conditional cash transfer parents compensate for the reduction in labor mar- (CCT) program Opportunidades/Progresa targeted ket work of one child (the CCT beneficiary who is poor children, school participation also increased required to attend school) by increasing the hours among children above the poverty cutoff.c Scholar- worked by other siblings.e ship programs for girls have demonstrated positive a. Miguel and Kremer 2004. externalities for boys’ school participation and teach- b. Zivin, Thirumurthy, and Goldstein 2006. ers’ attendance.d c. Lalive and Cattaneo 2006; Bobonis and Finan 2009. Some projects, however, may have unintended d. Kremer, Miguel, and Thornton 2009. negative consequences. For example, increasing the e. Barerra-Osorio, Linden, and Urquiola 2007; Filmer and availability of affordable antimalarial drugs—as Schady 2009. methods, objectives-based approaches, quali- interventions and impacts—something that tative methods, and theories of change. traditional monitoring and evaluation does This chapter draws especially on the body not always accomplish. Acknowledging of knowledge emerging from impact evalu- that not all rigorous evidence comes from ations in the areas of health and education. impact evaluations alone, we focus only on Even though we have restricted the scope of them, partly because of limitations of space this discussion to health and education, it is and partly because a lot of evidence coming important to recognize that impact evalu- from impact evaluations is fairly recent and ations are also generating valuable policy- highlights how evidence-based decision mak- relevant evidence in the areas of other MDGs ing can be operationalized (for examples, see (such as water and sanitation)13 that are not boxes 3.1 and 3.2). As noted, the evidence discussed here. In synthesizing the evidence base from impact evaluations in health and on health and education, we have restricted education sectors is still evolving and, at this ourselves to the more rigorous empirical stage, remains very far from complete. In work that has relied on a viable identifica- light of that, the chapter refrains from mak- tion strategy (that is, the counterfactual is ing any explicit policy recommendations or reliable) and produced causal links between engaging in detailed policy discussions. 78 LINKING SPENDING AND OUTCOMES GLOBAL MONITORING REPORT 2011 BOX 3.3 The DIME initiative There are several initiatives in the World Bank to Second, DIME builds teams that combine both improve the quality of evidence to inform development operational and technical expertise. This ensures that practice. One example is the Development Impact researchers work in close collaboration with project Evaluation (DIME) initiative. Created in 2005 in the teams from the planning stage onward to sustain Office of the Chief Economist, DIME was relaunched a process of prospective and operationally driven in 2009 as a broad-based, decentralized effort to knowledge generation. In this model, the project and mainstream impact evaluation in the Bank. The objec- research teams work together to embed a learning tive is to improve the quality of the Bank’s operations, agenda in the project design. strengthen country institutions for evidence-based Third, DIME organizes impact evaluations around policy making, and generate knowledge in 15 develop- development areas and offers tailored programmatic ment areas. and strategic support. Cross-country events develop With 185 completed studies and 280 active stud- regional communities of practice around certain sub- ies in 85 countries, the World Bank is attempting to themes, and teams share knowledge and experience enhance and expand knowledge of the what and the going beyond the impact evaluation agenda. This how of economic development by delivering precise ensures that all studies share a coherent and strategic estimates of the cause-effect relationship between research agenda and that the results are policy relevant policy action and outcomes. and are disseminated to a wide audience of practitio- The DIME approach centers on three aspects, in ners. DIME sponsors 15 thematic impact evaluation a dramatic shift in the practice of evaluation. First, programs: access to infrastructure, active labor mar- DIME has been moving the evaluation practice from ket programs, agriculture adaptations, conditional external and ex post evaluations toward internal, cash transfers, early childhood development, educa- prospective, and operationally driven evaluations that tion service delivery, energy mitigation, health systems are externally validated through the use of state-of- and results-based fi nancing, HIV/AIDS, fi nance and the-art methods. This framework promotes a results- private sector growth, forestry adaptations, institu- based model and allows governments to own and set tional reform, local development, malaria, and water the research agenda, thus receiving just-in-time advice resource management adaptations. on their most-pressing operational questions and gen- uinely transforming the way they make decisions for their programs. Source: Legovini 2010. Efforts to help improve access have met impact on total years of schooling com- with some success, but some questions pleted.15 Similarly, school feeding interven- remain tions have increased school participation in a variety of settings.16 Reducing schooling costs has helped increase A rich evidence base also shows the posi- school enrollment tive role of financial incentives in promoting The driving focus of education policy mak- school participation in a variety of contexts. ing in developing countries has been a push This evidence includes the evaluation of Mex- to increase enrollments in primary and sec- ico’s conditional cash transfer (CCT) pro- ondary schools. Interventions that reduce the gram that documented the efficacy of CCTs out-of-pocket costs of schooling have been as a mechanism for encouraging enrollment particularly successful.14 The reduction and attendance.17 These results have been in schooling costs can take many direct or replicated by many other researchers in many indirect forms. For instance, providing free other countries.18 In fact, evidence from CCT uniforms and textbooks and building more programs in nine other countries (Bangla- classrooms in rural Kenya had a strong desh, Cambodia, Chile, Colombia, Ecuador, GLOBAL MONITORING REPORT 2011 LINKING SPENDING AND OUTCOMES 79 MAP 3.1 The maternal mortality rate is declining only slowly, even though the vast majority of deaths are avoidable IBRD 38450 APRIL 2011 Maternal mortality Greenland (Den) Faeroe Norway Iceland Islands Finland (Den) Sweden The Netherlands Russian Federation Estonia Canada Isle of Man (UK) Denmark Russian Latvia Fed. Lithuania United Ireland Kingdom Germany Poland Belarus Channel Islands (UK) Belgium Ukraine Luxembourg Moldova Kazakhstan Liechtenstein Romania Mongolia Switzerland France Italy Andorra Bulgaria Georgia Azer- Uzbekistan Kyrgyz D.P.R. Turkey Armenia baijan Turkmenistan Rep. United States Portugal Spain Tajikistan of Korea Japan Monaco Greece Gibraltar (UK) Cyprus Syrian Rep. of Tunisia Malta Lebanon Arab Islamic Rep. Afghanistan China Korea Israel Rep. Iraq of Iran Bermuda Morocco Kuwait (UK) West Bank and Gaza Jordan Bahrain Pakistan Nepal Bhutan Former Algeria Libya Arab Rep. Qatar The Bahamas Spanish of Egypt Saudi Cayman Is. (UK) Sahara Arabia United Arab Bangladesh Cuba Turks and Caicos Islands (UK) India Mexico Emirates Myanmar Lao Haiti Mauritania Oman P.D.R. Belize Jamaica Cape Verde Mali N. Mariana Islands (US) Niger Rep. of Thailand Vietnam Guatemala Honduras Senegal Chad Eritrea Yemen Guam (US) El Salvador Nicaragua The Gambia Burkina Sudan Marshall Faso Djibouti Cambodia Philippines Islands Guinea-Bissau Guinea Benin Federated States of Costa Rica Panama Sri Micronesia R.B. de Guyana Sierra Leone Côte Ghana Nigeria Central Ethiopia Lanka Venezuela Suriname Liberia D’Ivoire African Rep. Brunei Cameroon Somalia Palau Colombia French Guiana (Fr) Togo Malaysia Equatorial Guinea Uganda Maldives Ecuador São Tomé and Príncipe Gabon Congo Rwanda Kenya Singapore Nauru Dem. Rep. of Burundi Seychelles Papua Solomon Congo Indonesia New Guinea Islands Tanzania Comoros Peru Brazil Timor-Leste Angola Zambia Malawi Mayotte (Fr) Vanuatu Bolivia Zimbabwe Mauritius Namibia Madagascar Paraguay Botswana Mozambique Réunion (Fr) New Caledonia Australia (Fr) Swaziland South Lesotho Africa Chile Uruguay Argentina New Dominican Puerto British Virgin Poland Maternal mortality Zealand Germany Republic Rico (US) Islands (UK) Ukraine Anguilla (UK) Czech Rep. Slovak Rep. Maternal mortality ratio, per 100,000 live births U.S. Virgin Islands Antigua and Barbuda (US) St. Maarten (Neth) St. Kitts Guadeloupe (Fr) Austria Hungary ≥ 500 and Nevis Dominica Slovenia Romania 200–499 Curaçao Montserrat (UK) Martinique (Fr) Croatia (Neth) St. Lucia Bosnia and 100–199 Aruba Bonaire St. Vincent and San Herz. Serbia Bulgaria (Neth) (Neth) The Grenadines Barbados Marino 50–99 Montenegro Kosovo Grenada FYR ≤49 Italy Macedonia Trinidad Vatican Albania R.B. de Venezuela and Tobago City Greece no data MDG performance Greenland (Den) Faeroe Norway Iceland Islands Finland (Den) Sweden The Netherlands Russian Federation Estonia Canada Isle of Man (UK) Denmark Russian Latvia Fed. Lithuania United Ireland Kingdom Germany Poland Belarus Channel Islands (UK) Belgium Ukraine Luxembourg Moldova Kazakhstan Liechtenstein Romania Mongolia Switzerland France Italy Andorra Bulgaria Georgia Azer- Uzbekistan Kyrgyz D.P.R. Spain Turkey Armenia baijan Turkmenistan Rep. of Korea United States Portugal Tajikistan Japan Monaco Greece Gibraltar (UK) Cyprus Syrian Rep. of Tunisia Malta Lebanon Arab Islamic Rep. Afghanistan China Korea Israel Rep. Iraq of Iran Bermuda Morocco Kuwait (UK) West Bank and Gaza Jordan Bahrain Pakistan Nepal Bhutan Former Algeria Libya Arab Rep. Qatar The Bahamas Spanish of Egypt Saudi Cayman Is. (UK) Sahara Arabia United Arab Bangladesh Cuba Turks and Caicos Islands (UK) India Mexico Emirates Myanmar Lao Haiti Mauritania Oman P.D.R. Belize Jamaica Cape Verde Mali N. Mariana Islands (US) Niger Rep. of Thailand Vietnam Guatemala Honduras Senegal Chad Eritrea Yemen Guam (US) El Salvador Nicaragua The Gambia Burkina Sudan Marshall Faso Djibouti Cambodia Philippines Islands Guinea-Bissau Guinea Benin Federated States of Costa Rica Panama Sri Micronesia R.B. de Guyana Sierra Leone Côte Ghana Nigeria Central Ethiopia Lanka Venezuela Suriname Liberia D’Ivoire African Rep. Brunei Cameroon Somalia Palau Colombia French Guiana (Fr) Togo Malaysia Equatorial Guinea Uganda Maldives Ecuador São Tomé and Príncipe Gabon Congo Rwanda Kenya Singapore Nauru Dem. Rep. of Burundi Seychelles Papua Solomon Congo Indonesia New Guinea Islands Tanzania Comoros Peru Brazil Timor-Leste Angola Zambia Malawi Mayotte (Fr) Vanuatu Bolivia Zimbabwe Mauritius Namibia Madagascar Paraguay Botswana Mozambique Réunion (Fr) New Caledonia Australia (Fr) Swaziland South Lesotho Africa Chile Uruguay Argentina MDG performance New Dominican Puerto British Virgin Poland MDG 5.A Reduce by three-quarters, Zealand Germany Republic Rico (US) Islands (UK) between 1990 and 2015, the maternal Ukraine Anguilla (UK) Czech Rep. U.S. Virgin Islands Antigua and Barbuda Slovak Rep. mortality ratio (US) Austria St. Maarten (Neth) St. Kitts Guadeloupe (Fr) Hungary and Nevis Slovenia on target Dominica Romania Curaçao Montserrat (UK) Martinique (Fr) Croatia close to target (Neth) Bosnia and Bonaire St. Vincent and St. Lucia San Herz. Serbia Aruba far from target Bulgaria (Neth) (Neth) The Grenadines Barbados Marino This map was produced by the Map Design Unit of The World Bank. Montenegro Kosovo no data The boundaries, colors, denominations and any other information Grenada FYR Italy Macedonia shown on this map do not imply, on the part of The World Bank Trinidad Vatican Albania Group, any judgment on the legal status of any territory, or any and Tobago City Greece R.B. de Venezuela endorsement or acceptance of such boundaries. Source: World Bank staff calculations based on data from the World Development Indicators database. 80 LINKING SPENDING AND OUTCOMES GLOBAL MONITORING REPORT 2011 Honduras, Nicaragua, Pakistan, and Tur- students but also among untreated students at key) has consistently demonstrated signifi- treatment schools and among untreated stu- cant impacts of these programs on school dents at nearby nontreatment schools, thanks enrollment in the initial years of program to reduced disease transmission. 25 Also in operation. Reducing user fees can similarly Kenya, school-based intermittent preventive increase enrollment.19 However, few of these treatment helped reduce anemia prevalence.26 programs have been around long enough to The delivery of such treatment was estimated determine whether the impact is a short-term to cost $2 per child treated per year, and the effect of a novel project or is more lasting. estimated cost per anemia case averted was In addition, the higher enrollments induced $30. This type of intervention would be espe- by these programs were not accompanied by cially cost effective in areas with high rates of increased achievement.20 malaria transmission.27 Some impact evaluations have shown that Cheap telecommunication technologies can improving school quality does not seem to address uptake, especially in resource-limited be a major inducement for increasing school settings. An intervention in Kenya used text participation.21 Indeed, interventions improv- messages to remind antiretroviral therapy ing the quality of education have, on aver- patients to adhere to the treatment guide- age, generated no changes in participation.22 lines—an important goal, given concerns of But much more evidence on this question is widespread drug resistance when adherence needed before the results can be generalized. is poor. Weekly reminders increased the per- centage of participants achieving 90 percent Health input provision programs have to grapple adherence to the therapy by approximately with uptake issues 13–16 percent, compared with no reminder.28 All provision programs deal with the central They also reduced the frequency of treatment issue of how to get citizens to take up the interruptions. inputs and services they offer. In develop- Community-based interventions have also ing countries, health facilities are not always become more popular, particularly as comple- accessible to poor, rural, and remote com- ments to facility-based care, linking commu- munities. Other modes of delivery, such as nities with facilities-based services. Despite school-based and community-based health showing great promise, community health interventions, have been popular. But we are worker programs have had mixed results, as only beginning to learn how well such pro- reflected in the findings of one of the few rig- grams work and under what conditions. orous impact evaluations of such programs Impact evaluations have shown that in Ethiopia. The Ethiopian program signifi- schools can be an effective mode of delivery cantly increased the proportion of children for health input provision and can address fully immunized and the proportions of chil- some uptake challenges because the families dren and women using insecticide-treated do not have to be persuaded. Some (but not bednets for malaria protection. But the effect all) school feeding programs have improved on prenatal and postnatal care was rather lim- children’s nutritional intake23 and health. ited, as was the impact on the incidence and A recent study has shown improvements in duration of diarrhea among children under 5 anthropometric outcomes of the younger years of age.29 siblings of students who receive take-home Because community-based interven- rations.24 tions often come as a package comprising Other school-based health programs have input provision, information, advocacy, and also shown positive impacts. School-based other components, it is hard to untangle the mass treatment with inexpensive deworm- impacts of the different components. Impact ing drugs in Kenya (where the prevalence of evaluations may be useful in this; but, so far, intestinal worms among children is very high) their number is quite limited. A meta-analysis improved health not only among treated of health literature found that clean delivery GLOBAL MONITORING REPORT 2011 LINKING SPENDING AND OUTCOMES 81 had the most pronounced effect on reducing after copayments increased the transparency maternal mortality. In comparison, building and accountability of providers to poor cli- community support and advocacy groups, ents. 34 In contrast, an impact evaluation in involving other family members through Kenya fi nds that cost sharing does not seem community mobilization, or training com- to improve targeting of insecticide-treated munity health workers and traditional birth bednets for malaria prevention to those in attendants had little or no impact on reducing greatest need, and that women who pay for maternal mortality. But community interven- their bednets are no more likely to use them tions affected perinatal mortality through dif- than are those who receive them free.35 Cost ferent channels: community support groups, sharing also dampens demand considerably, advocacy through group sessions, and family with uptake dropping by 75 percent when the involvement in care were especially effective price of bednets increases from zero to $0.75 in reducing perinatal deaths. Family involve- (from a 100 percent subsidy to an 88 percent ment in care also showed a positive impact in subsidy). reducing stillbirths. Additional impact evalu- Another impact evaluation in Kenya ations are needed to confirm these findings. shows that introducing a small cost-sharing Other programs have tried to deal with component into a school-based deworming the uptake issue by combining demand-side program dramatically reduced the uptake of and supply-side interventions. An impact deworming medication and raised little rev- evaluation of an immunization program in enue, relative to administrative costs. Nor did India shows that whereas pure supply-side the user fees help target treatment to the sick- improvements were associated with increases est students.36 Interestingly, one study shows in immunization, uptake rates were much that uptake is particularly sensitive to price at higher when supply-side improvements were prices close to zero.37 In other words, uptake combined with modest demand-side incen- is found to be highly sensitive to having a tives in kind. 30 There are few comparative positive price, but there is less evidence that studies of the relative impact on demand-side it is sensitive to variation within the positive stimulation versus supply-side responses. For price range. one of the few CCTs that combined inter- These fi ndings suggest the need for more ventions on both sides—Nicaragua’s Red de research, particularly to establish optimal Proteccion Social—the impact evaluation levels of subsidization for different health was unable to untangle the demand-side and inputs.38 supply-side effects.31 CCT programs have often been effective in Cost sharing does not necessarily improve the reaching the poor efficiency of health input subsidies CCTs have traditionally been designed to tar- Many practitioners argue that for some types get poor people, so it is not surprising that of health inputs (particularly those requiring a number of CCT programs have reduced the recipient to follow a treatment regime to poverty. But the impact varies, and there is see health benefits), cost sharing can improve still not much clarity about the conditions the efficiency of the input subsidy. 32 In this under which CCTs will be most effective in long-standing debate, others contend that reducing poverty. Red de Proteccion Social cost sharing dampens demand, especially in Nicaragua and Familias en Accion in among the poor. Colombia show strong effects on poverty, In Zambia, households paying a higher whereas Programa de Asignacion Familiar price for a water treatment product were more in Honduras and Oportunidades/Progresa likely to report treating their drinking water in Mexico had significant but more modest two weeks later.33 Likewise, controlled stud- effects. Some of the larger CCTs from Bra- ies in several countries record improvements zil, Jamaica, and Mexico have also shown in the use of services among poor people an impact on poverty at the national level. 82 LINKING SPENDING AND OUTCOMES GLOBAL MONITORING REPORT 2011 But the CCTs in Cambodia and Ecuador did Whether cash transfers should be not affect median consumption or reduce “conditioned” on certain behaviors is national poverty.39 context specific In Ecuador, cash transfers improved cogni- tive outcomes for the poorest children but not Although CCTs have shown some positive for children somewhat better off.40 In poor impacts on schooling and other outcomes,44 households, child work is a major cause of unconditional cash transfers have also school-age children dropping out of school. improved child health and increased school Reduced child work by CCT beneficiaries participation.45 So, development practitioners has been found in Brazil, Cambodia, Ecua- and policy makers have become very inter- dor, Mexico, and Nicaragua. In Cambodia, ested in whether the programs should be con- for example, the average child receiving the ditioned on certain behaviors or not. transfer was 10 percentage points less likely A study shows that some households in to work for pay. Results on this have been Mexico and Ecuador did not realize that the mixed, however, and more research is needed cash transfers they were getting were condi- to understand work’s impact on siblings— tional on school attendance; and, for these especially if the number of CCTs targeting households, school enrollment was signifi- individual children increases. cantly lower than for those who thought that Despite being more likely to adopt the transfers were conditional.46 improved child care practices, the worst-off In Malawi, the effect of conditioning dif- households in Mexico had greater difficulties fered for different outcomes. For school par- translating these improvements into improved ticipation of adolescent girls, the impact of the nutritional outcomes—largely because of the transfers was significantly greater when they lack of complementary inputs among poorly were conditioned on school attendance than educated mothers, poor access to safe water, when not so conditioned. But they had little and poorer quality nutrients.41 effect on the likelihood of teenage pregnancies or marriages, while unconditional cash trans- fers were very effective in delaying marriage Incentives do not need to be large and childbearing.47 Therefore, the answer to Some impact evaluations have shown that the question of whether cash transfers should small benefits can often be enough to incen- be conditioned on certain behaviors appears tivize the desired change in behavior, and to be context specific—but, again, more increasing the total size of the transfer has research is needed on the subject. only small marginal effects on service uptake or behavior. In Cambodia, giving a child $45 Providing information can be a highly cost- quarterly to stay in school proved as effec- effective way to change behavior tive as giving $60.42 Similarly, in Malawi, the An impact evaluation from Madagascar school participation effects of giving a house- shows that providing information to youths hold $5 were not significantly different from about returns to schooling improved children’s giving $10. “[T]he key parameter in setting school performance and attendance in the benefit levels is the size of the elasticity of the first few months following the intervention.48 relevant outcomes to the benefit level.”43 Similarly, providing information on returns Elasticity also varies across the type of to schooling to boys in eighth grade increased behavior being incentivized. In southwest average years of schooling in the Dominican Tanzania, the group receiving $60 had reduc- Republic, although the program did not have tions of sexually transmitted infections, but an impact on the poorest students.49 the $30 group had the same infection rate as Providing information to adults has also the control group that received no payments. shown some promising results. In India, Not surprisingly, the program was more effec- informing households that their drink- tive for people from poorer and rural areas. ing water is contaminated increases the GLOBAL MONITORING REPORT 2011 LINKING SPENDING AND OUTCOMES 83 probability that they start purifying their of traditional inputs has not been very suc- water.50 In Bangladesh, informing households cessful for improving learning outcomes. Tra- that the water in their wells has an unsafe ditional inputs that have been tested on this concentration of arsenic raises the probability dimension include textbooks, school meals, that they will switch to another well. 51 But blackboards and other visual aids (like fl ip information provision alone does not change charts), teacher training, and even smaller consumer behavior in all contexts. A study class sizes. 57 Some of these studies are dis- from Kenya shows that providing consumers cussed below. information about the mortality and mor- Some example of such studies include one bidity from malaria or about the financial in Kenya that found that providing textbooks gains from avoiding malaria infection had to students did not raise average test scores; it no impact on the uptake of malaria control raised only the scores of the best students.58 devices.52 Similarly, evidence from the United Also in Kenya, introducing complementary States has shown that information alone is learning aids such as flip charts to schools usually ineffective in changing risky health- had no impact on student performance. 59 related behavior.53 Similarly, school feeding interventions have The experience with translating information had little or no impact on student perfor- and knowledge into HIV risk-reducing behav- mance. 60 The Bolivian Social Investment ior change has had mixed results, as the AIDS Fund had a significant impact on school literature suggests.54 There are, however, some infrastructure but not on education outcomes impact evaluations of behavioral interventions within the evaluation period. 61 An extra that have shown a reduction in risky sexual teacher in rural Indian nonformal educa- behavior. 55 In western Kenya, for example, tion centers did not significantly improve test female students were warned about the high scores (although this was not measured with HIV prevalence rates among older men and great precision).62 And, contrary to expecta- they responded by dramatically reducing the tions, some studies have shown that simply number of teenage childbirths with older men. reducing pupil–teacher ratios does little for Similarly, helping girls stay in school by giv- student performance.63 ing them a free school uniform reduced teen In recent years, education practitioners have pregnancy at a relatively low cost ($12 per girl) experimented with nontraditional inputs— when compared to the cost per teenage child- mostly those to fill gaps left by teachers—and birth averted ($750). Providing abstinence- technology-based inputs emerge as promis- only information about teenage sexual behav- ing. A Nicaraguan program of radio instruc- ior had no impact on the likelihood of risky tion for students finds strong positive impacts behavior. But providing detailed information on mathematics performance.64 Computer- on the risk of HIV led to a 28 percent decrease assisted learning in urban India increased test in the likelihood that girls started childbearing scores in mathematics.65 In another program in within a year—thus suggesting a reduction in Indian schools, an electronic machine or flash unprotected sex among those girls.56 card–based activities (to help teach English) increased test scores in English.66 Computers failed, however, to improve student perfor- Despite improvements in access, it has mance in a program in Colombia.67 been very difficult to improve learning Newer approaches toward remedial edu- and health outcomes cation for poor-performing students have also shown promise. A reading intervention Providing increased traditional schooling inputs in rural India that gave four days of training has often been ineffective in improving learning to community volunteers with grade-10 or outcomes grade-12 educations to enable them to teach Attempting to fill narrowly defi ned resource children how to read significantly improved gaps in schooling by increasing the provision reading.68 84 LINKING SPENDING AND OUTCOMES GLOBAL MONITORING REPORT 2011 CCTs help increase the uptake of services, One example is a small cash transfer program but their impact on health and learning in Malawi’s Zomba district, where the trans- outcomes is mixed fers improved school enrollment and atten- A positive impact on health care utilization dance but did not significantly improve learn- (such as children’s visits to health centers) ing variables (which included a self-report on was reported in Honduras, Jamaica, Mexico, English literacy and a teacher’s evaluation of Nicaragua, and Paraguay (but not in Chile).69 student progress).76 The lack of any discern- Out of seven studies reporting immunization ible effect on learning outcomes (despite large results, significant and sizable impacts on impacts on school enrollment) may partly be full immunization coverage (an intermediate caused by their drawing lower-ability stu- outcome) were found only in Nicaragua and dents back to school.77 Turkey.70 In Mexico, however, the impact was In a similar vein, CCTs have had a positive small and not significant—results ascribed to impact on the uptake of growth monitoring the already high immunization rate at base- services.78 But the impact on growth indica- line (above 90 percent). tors has been mixed, and the positive impacts The impact of CCTs on health and nutri- were generally modest.79 CCTs in Colombia, tional outcomes is also mixed. One of the Mexico, and Nicaragua were accompanied few studies that reported health impacts was by a substantial positive impact on linear Mexico’s Oportunidades/Progresa, where growth,80 although no impact was found in children under 3 who received CCTs were Brazil81 or Honduras.82 Nonetheless, despite 22 percent less likely to report an illness epi- the difficulty of cross-study comparisons sode in the previous four weeks than were the (because of differences in age groups, expo- children in the comparison group. Children sure, and program design), it can be concluded young enough to be exposed to the program that programs with larger cash transfers (such for 24 months were 40 percent less likely to as those in Colombia, Mexico, and Nicara- be reported ill—a fi nding that suggests the gua, where transfers represented 15–25 per- program generated cumulative health bene- cent of total household expenditures) tend to fits. Conversely, no effect on child height was have the largest impact. Younger age groups reported among children of any age in Brazil, are likely to benefit more (consistent with the Ecuador, Honduras, and Nicaragua;71 and, in nutrition literature), and the effects are larger Colombia, among children ages 3–7 years.72 for height than for weight indicators.83 Where studies have reported positive institu- Only three of the CCT program evalua- tional impacts, the impacts have been larger tions (Honduras, Mexico, and Nicaragua) among younger children (that is, younger at looked at the impact on micronutrient sta- the time of the baseline survey—in particu- tus. In Mexico, after one year, beneficiary lar, younger than 2 years at baseline). One of children had higher mean hemoglobin levels the few CCTs that addressed maternal health and significantly lower rates of anemia than outcomes, Janani Surksha Yojana in India, did nonbeneficiaries. 84 But nearly half the had a small impact on the uptake of prena- children under 1 year were still anemic after tal care; a large, positive impact on institu- one year of exposure to the intervention. tional deliveries; and a smaller effect on home No effect was found for the older children, delivery with a skilled birth attendant. There and no differences were observed for other was also some evidence of a lower probability micronutrients (such as iron, vitamin A, and of perinatal and neonatal deaths, but not of zinc). The modest impacts on micronutrient maternal deaths.73 status were ascribed to low use or uptake of Although the positive impact of CCTs the fortified foods (Mexico)85 and supple- in expanding access to education is well ments (Nicaragua), 86 sharing foods with documented,74 evidence on the impact of edu- other beneficiary household members, over- cational transfers (in-kind or cash transfers) dilution of the product, and program design on learning outcomes is not as encouraging.75 weaknesses.87 GLOBAL MONITORING REPORT 2011 LINKING SPENDING AND OUTCOMES 85 MAP 3.2 Each year of a girl’s education reduces, by as much as 10 percent, the risk of her children dying before age five IBRD 38447 APRIL 2011 Gender parity (primary) Greenland (Den) Faeroe Norway Iceland Islands Finland (Den) Sweden The Netherlands Russian Federation Estonia Canada Isle of Man (UK) Denmark Russian Latvia Fed. Lithuania United Ireland Kingdom Germany Poland Belarus Channel Islands (UK) Belgium Ukraine Luxembourg Moldova Kazakhstan Liechtenstein Romania Mongolia Switzerland France Italy Andorra Bulgaria Georgia Azer- Uzbekistan Kyrgyz D.P.R. Turkey Armenia baijan Turkmenistan Rep. United States Portugal Spain Tajikistan of Korea Japan Monaco Greece Gibraltar (UK) Cyprus Syrian Rep. of Tunisia Malta Lebanon Arab Islamic Rep. Afghanistan China Korea Israel Rep. Iraq of Iran Bermuda Morocco Kuwait (UK) West Bank and Gaza Jordan Bahrain Pakistan Nepal Bhutan Former Algeria Libya Arab Rep. Qatar The Bahamas Spanish of Egypt Saudi Cayman Is. (UK) Sahara Arabia United Arab Bangladesh Cuba Turks and Caicos Islands (UK) India Mexico Emirates Myanmar Lao Haiti Mauritania Oman P.D.R. Belize Jamaica Cape Verde Mali N. Mariana Islands (US) Niger Rep. of Thailand Vietnam Guatemala Honduras Senegal Chad Eritrea Yemen Guam (US) El Salvador Nicaragua The Gambia Burkina Sudan Marshall Faso Djibouti Cambodia Philippines Islands Guinea-Bissau Guinea Benin Federated States of Costa Rica Panama Sri Micronesia R.B. de Guyana Sierra Leone Côte Ghana Nigeria Central Ethiopia Lanka Venezuela Suriname Liberia D’Ivoire African Rep. Brunei Cameroon Somalia Palau Colombia French Guiana (Fr) Togo Malaysia Equatorial Guinea Uganda Maldives Ecuador São Tomé and Príncipe Gabon Congo Rwanda Kenya Singapore Nauru Dem. Rep. of Burundi Seychelles Papua Solomon Congo Indonesia New Guinea Islands Tanzania Comoros Peru Brazil Timor-Leste Angola Zambia Malawi Mayotte (Fr) Vanuatu Bolivia Zimbabwe Mauritius Namibia Madagascar Paraguay Botswana Mozambique Réunion (Fr) New Caledonia Australia (Fr) Swaziland South Lesotho Africa Chile Uruguay Argentina Gender parity (primary) New Dominican Puerto British Virgin Poland Percentage of girls to boys enrolled at Zealand Germany Republic Rico (US) Islands (UK) Ukraine Anguilla (UK) Czech Rep. primary level in public and private schools U.S. Virgin Islands Antigua and Barbuda Slovak Rep. (US) St. Maarten (Neth) St. Kitts Guadeloupe (Fr) Austria Hungary <79.9 and Nevis Dominica Slovenia Romania 80–89.9 Curaçao Montserrat (UK) Martinique (Fr) Croatia (Neth) St. Lucia Bosnia and 90–94.9 Aruba Bonaire St. Vincent and San Herz. Serbia Bulgaria (Neth) (Neth) The Grenadines Barbados Marino 95–99.9 Montenegro Kosovo Grenada FYR ≥100 Italy Macedonia Trinidad Vatican Albania R.B. de Venezuela and Tobago City Greece no data MDG performance Greenland (Den) Faeroe Norway Iceland Islands Finland (Den) Sweden The Netherlands Russian Federation Estonia Canada Isle of Man (UK) Denmark Russian Latvia Fed. Lithuania United Ireland Kingdom Germany Poland Belarus Channel Islands (UK) Belgium Ukraine Luxembourg Moldova Kazakhstan Liechtenstein Romania Mongolia Switzerland France Italy Andorra Bulgaria Georgia Azer- Uzbekistan Kyrgyz D.P.R. Turkey Armenia baijan Turkmenistan Rep. United States Portugal Spain Monaco Tajikistan of Korea Japan Greece Syrian Gibraltar (UK) Cyprus Rep. of Tunisia Malta Lebanon Arab Islamic Rep. Afghanistan China Korea Israel Rep. Iraq of Iran Bermuda Morocco Kuwait (UK) West Bank and Gaza Jordan Bahrain Pakistan Nepal Bhutan Former Algeria Libya Arab Rep. Qatar The Bahamas Spanish of Egypt Saudi Cayman Is. (UK) Sahara Arabia United Arab Bangladesh Cuba Turks and Caicos Islands (UK) India Mexico Emirates Myanmar Lao Haiti Mauritania Oman P.D.R. Belize Jamaica Cape Verde Mali N. Mariana Islands (US) Niger Rep. of Thailand Vietnam Guatemala Honduras Senegal Chad Eritrea Yemen Guam (US) El Salvador Nicaragua The Gambia Burkina Sudan Marshall Faso Djibouti Cambodia Philippines Islands Guinea-Bissau Guinea Benin Federated States of Costa Rica Panama Sri Micronesia R.B. de Guyana Sierra Leone Côte Ghana Nigeria Central Ethiopia Lanka Venezuela Suriname Liberia D’Ivoire African Rep. Brunei Cameroon Somalia Palau Colombia French Guiana (Fr) Togo Malaysia Equatorial Guinea Uganda Maldives Ecuador São Tomé and Príncipe Gabon Congo Rwanda Kenya Singapore Nauru Dem. Rep. of Burundi Seychelles Papua Solomon Congo Indonesia New Guinea Islands Tanzania Comoros Peru Brazil Timor-Leste Angola Zambia Malawi Mayotte (Fr) Vanuatu Bolivia Zimbabwe Mauritius Namibia Madagascar Paraguay Botswana Mozambique Réunion (Fr) New Caledonia Australia (Fr) Swaziland South Lesotho Africa Chile Uruguay Argentina MDG performance New Dominican Puerto British Virgin Poland MDG 3.A Eliminate gender disparity in Zealand Germany Republic Rico (US) Islands (UK) primary education, preferably by 2005, Ukraine Anguilla (UK) Czech Rep. U.S. Virgin Islands Antigua and Barbuda Slovak Rep. and at all levels of education no later than 2015 (US) Austria St. Maarten (Neth) St. Kitts Guadeloupe (Fr) Hungary and Nevis Slovenia on target Dominica Romania Curaçao Montserrat (UK) Martinique (Fr) Croatia close to target (Neth) Bosnia and Bonaire St. Vincent and St. Lucia San Herz. Serbia Aruba far from target Bulgaria (Neth) (Neth) The Grenadines Barbados Marino This map was produced by the Map Design Unit of The World Bank. Montenegro Kosovo no data The boundaries, colors, denominations and any other information Grenada FYR Italy Macedonia shown on this map do not imply, on the part of The World Bank Trinidad Vatican Albania Group, any judgment on the legal status of any territory, or any and Tobago City Greece R.B. de Venezuela endorsement or acceptance of such boundaries. Source: World Bank staff calculations based on data from the World Development Indicators database. 86 LINKING SPENDING AND OUTCOMES GLOBAL MONITORING REPORT 2011 Nutritional improvements are expected to micronutrient supplementation before 3 years flow from the increased consumption arising of age showed beneficial effects on school- from the income effect of the cash transfer ing, reading, and intelligence tests in adult- and the better practices (such as breastfeed- hood (25– 42 years). 91 Some evaluations ing, proper weaning, growth monitoring, and show significant improvements in cognitive consumption of nutritious foods) that arise and learning outcomes among children who from health education during growth moni- benefited from CCT programs before they toring. The relative contribution of health entered school.92 At every level of income, education to each of these practices has not households with young children receiving the been assessed. The majority of impact evalu- cash transfers were more likely to read, tell ations have found that CCTs may dispropor- stories, and sing to their children and to have tionately affect the consumption of particular books, paper, and pencils for them to use at items, such as food. It has been hypothesized home.93 This is an important result because that CCTs increase the bargaining power of early cognitive development was a strong pre- women within the household, and that this dictor of school attainment in Brazil, Gua- drives the greater food expenditure.88 There temala, Jamaica, the Philippines, and South is also some evidence of changes in diet com- Africa, even after controlling for wealth and position. Not only do households increase mother’s education.94 dietary diversity, but they also shift toward Children who qualify for CCTs benefit higher-quality sources of calories. In Nicara- most from schooling at an early age. In Ecua- gua, households that receive transfers from dor, most children at age 3 in the sample were the Atencion a Crisis program spend sig- only modestly behind the reference popula- nificantly less on staples and more on animal tion on a test of cognitive development.95 By protein, fruits, and vegetables.89 age 6, however, when they entered first grade, Identifying the reasons for increased use of children in the two poorest deciles of the health or education services is complicated. national distribution of wealth were almost Sometimes it rises because services are physi- three standard deviations behind where they cally more accessible. This is likely the case should be—and have almost no chance of for such low-cost health services as bednets catching up. Remedial investments targeting and treatment for sexually transmitted infec- them therefore need to make equity–efficiency tions where the nonservice costs of use— tradeoffs; investments in early childhood are particularly transport and perhaps facility more likely to avoid such tradeoffs.96 Simi- search costs—are a larger share of cost to the larly, the results reported above on the impact consumer than is the service cost. In educa- of nutrition programs on growth and micro- tion, a crucial element is always likely to be nutrient status suggest larger impacts among the opportunity cost to households (the loss small children. Despite this evidence, there in earnings from sending children). A fur- is some debate over which age group CCTs ther reason for increased uptake is that use should target to ensure that school-age chil- is linked to finance, and thus it improves the dren also benefit.97 providers’ incentive to increase the quantity of services.90 Service delivery often fails the poor A recent study reports results from surveys in Intervening early in a child’s life can produce which enumerators made unannounced visits strong impacts to primary schools and health clinics in Ban- Evidence suggests that education and health gladesh, Ecuador, India, Indonesia, Peru, and programs targeting young children can have Uganda. They recorded whether they found strong impacts on overall human develop- teachers and health workers in the facilities.98 ment outcomes. Some of the most compel- On average, about 19 percent of teachers and ling evidence for this comes from Guate- 35 percent of health workers were absent, and mala, where nutritional intervention and many teachers and health workers in their GLOBAL MONITORING REPORT 2011 LINKING SPENDING AND OUTCOMES 87 facilities were not working. Across Indian public clinics have less political power than government-run schools, only 45 percent of do middle-class teachers and health workers. teachers assigned to a school were engaged In many countries, even those moderately in teaching at any given time. In Ghana, well off send their children to private schools Morocco, Tunisia, and the Brazilian state of and use private clinics. Even when part of the Pernambuco99 the instructional time teachers population chooses private providers instead allocated to learning tasks ranged from 78 of public ones, public money keeps flowing percent in Tunisia to 39 percent in Ghana. into public schools and hospitals. Hence, the Observational data for a sample of doc- cycle of weak accountability and weak incen- tors in New Delhi, India, showed differences tives continues. between the levels of expertise and quality of care provided among private doctors who New approaches are being designed serve the rich and the poor. Poorer patients to improve the critical link of service receive low-quality advice and spend a fair delivery amount of money on unnecessary (and often substandard) drugs. Wealthier patients get As seen in figure 3.4, service delivery is the better advice, both because they see more link between generation and consumption of competent providers and because their pro- goods and services. Empirical evidence from viders put in more effort.100 microeconomic studies has shown that this Why are incentives weak and accountabil- link is often weak in developing countries— ity low among service providers in develop- an important reason why public spending on ing countries? One reason is the highly cen- inputs and demand-side interventions does tralized and bureaucratic education and not translate into commensurate human health systems. Hiring, salaries, and pro- development outcomes. What explains the motion decisions are made at the center, service delivery weaknesses? determined largely by educational qualifica- The following issues stand out: inequita- tions and seniority, with almost no scope for bly low allocations to services reaching low- performance-based pay. Governments fi nd income groups, leaks of funding between it difficult to monitor the performance of central ministries and frontline providers, and health workers, especially those providing failures of frontline providers such as teach- highly discretionary services such as clini- ers and doctors to perform effectively. Impact cal care. Furthermore, teachers and health evaluations have highlighted and explored workers are typically organized interest solutions on the last of those three points.101 It groups that can be politically influential. has been argued that frontline service provid- These factors imply that disciplinary action ers’ weak incentives and low accountability in for poor performance is rare and that teach- developing countries often translates into inef- ers and health workers are almost never fective service delivery.102 New service delivery fired. In addition, asymmetric informa- approaches attempt to solve the incentive and tion between client and provider (especially accountability issues while reducing the need strong for health services) makes account- for (onerous) monitoring by the government ability difficult. and the citizenry. These approaches can be These problems are almost never addressed classified under three headings: information through the long route of accountability— for accountability, greater citizen accountabil- through political pressure that consumers ity and autonomy in service provision, and exert on the government. One possible rea- pay for performance.103 son poor performance of service providers is not on the political agenda: providers are Information to increase accountability can be an organized interest group, and clients, par- effective in some contexts ticularly in health, are diffuse. Those who To hold their service providers accountable, are poor enough to use public schools and poor people need information on their rights 88 LINKING SPENDING AND OUTCOMES GLOBAL MONITORING REPORT 2011 and on how well their service providers are yielded mixed results. In both studies, villag- performing. Furthermore, availability of ers were provided with information about information on service delivery performance their rights and responsibilities for educa- provides the means to monitor progress. In tion provision and oversight. One part of the recent years, programs have been designed to country showed no impact from providing provide citizens with this information in the information alone,109 and a different part of hope that it will enable them to hold service the country showed some impact on student providers accountable. This has been easier learning through improved service provision to do in education than in health because of by teachers.110 the nature of the services.104 It seems clear that, although providing Citizen report cards provide clients with information for accountability to citizens simple information on how well local ser- may be important, it is ineffective to sim- vice providers are doing. In Pakistan, pro- ply give communities information on school viding parents with report cards containing quality without also increasing their ability information about the relative performance to take action. of children and schools in the village (includ- ing private schools) improved student perfor- Greater citizen accountability and autonomy in mance in public schools and lower-quality service provision show promise private schools and reduced fees at higher- Parents and community leaders have informa- quality private schools.105 tion about local service providers and local Elsewhere, impact evaluations of similar service delivery that centralized systems tend programs have shown less direct results. For to ignore. Involving them in decision making example, a Liberian program that publicized with system actors (teachers, head teachers, to parents the results of an early-grade reading doctors, and nurses) can better align spend- assessment and showed teachers how to pre- ing with local needs. Greater autonomy and pare quarterly report cards for parents showed accountability of local service delivery can very little impact on learning after two years. increase incentives to adopt proven successful But when the information on student reading practices; evaluate the effectiveness of home- skills was combined with intensive teacher grown initiatives; and build a sense of com- training, student performance improved dra- mitment, ownership, and pride. But there are matically.106 Similarly in Chile, publicizing also risks, including limited information, nar- rankings of school quality by region did not row interests, elite capture, and inadequate appear to affect school performance.107 representation of the disadvantaged. Increas- A large information campaign in Uganda ingly, development practitioners are experi- used school display boards and newspapers menting with ways to increase the autonomy to provide information to parents and com- of local stakeholders in service provision. munities on the exact amount of operational Again, this model has been evaluated more in funding each school could expect and the education than in health. date when the funding transfer to the school One way to increase accountability is to would be made. The campaign was associ- give local communities resources and train- ated with a sharp reduction in the leak of ing to manage service provision, an approach school grants expenditures.108 The research- tried in school-based management programs ers also documented that schools in districts that give local communities decision-making with higher media exposure experienced power and resources for school management. greater increases in enrollments and higher Studies show a positive impact on how schools test scores on the primary-school exit exam are managed and how teachers, head-teachers, than did schools in districts with lower media and parents interact, but they reveal no consis- exposure. tent improvement in student performance. Another dimension of information for For example, school-based management accountability, tested in two studies in India, in Mexico has a positive impact on teacher GLOBAL MONITORING REPORT 2011 LINKING SPENDING AND OUTCOMES 89 behavior and school management,111 but lit- outcomes. In some cases (such as in India),115 tle impact on student performance.112 A simi- individual bonuses led to larger improve- lar program in Nepal showed a significant ments in student outcomes, whereas group impact on access and equity; the impact on bonuses had lower implementation costs. school governance measures was mixed, with That suggests some cost–benefit trade-offs regional variation in outcomes; and there was between the two. In Chile’s National System no evidence that the changes were associated of School Performance, group bonuses also with improvements in learning outcomes.113 yielded quite modest impacts. In more detail, There has been considerable work on hos- one evaluation of the system found a modest pital autonomy and the impacts of decentral- cumulative positive effect on student achieve- ized management on hospital performance, ment for schools that had a higher probability but the analyses have rarely had any controlled of winning the bonus,116 and a positive effect or quasi-experimental design to control for a on student test scores.117 counterfactual. This is partly because hos- Some evaluations suggest uncertainty pital autonomy is usually implemented with about the duration of benefits. In Kenya, many other management reforms, making it group-based incentives for teachers improved difficult to isolate autonomy from manage- student performance but the gains were ment capacity and system improvements. short-lived.118 Other settings suggest that the Another approach is to give clients more targets are important. In the Brazilian state control over service providers. In western of Pernambuco, during the first year, schools Kenya, giving parents stronger oversight with more ambitious targets made signifi- over teachers improved student performance cantly larger test score gains than did simi- when combined with other reforms. And the larly performing comparison schools assigned effect on student performance of hiring extra lower targets. In the second year—with con- teachers can be greater when school commit- trols for schools’ 2008 test results and other tees are given training on how to supervise school characteristics—schools that barely and manage teachers; the effect was largest missed the threshold for getting the bonus in among the students assigned to a contract 2008 improved more than schools that barely teacher (as opposed to a civil servant). In achieved it. It appears that both higher targets India, local communities’ hiring of contract and barely missed bonus achievement created teachers (who were younger and less quali- incentives that had a positive effect on school fied) at lower salaries than those received by motivation and performance.119 regular teachers by improved student perfor- Two other evaluated programs rewarded mance.114 But very little is known about the teachers for attendance. One allowed school long-term effects of contract teachers, and heads in rural Kenya to give individual teach- such a system is unlikely to be sustainable ers bonus pay for regular attendance (mea- over a long period. sured by unannounced random visits).120 Again, the evidence on school-based But they simply distributed the full bonus to management is inconclusive. More impact all teachers, regardless of attendance. The evaluations are needed, particularly on the inability or unwillingness of the head teacher pathways that translate this kind of decen- to enforce the performance bonus to reward tralization into better school quality. regular attendance explained the lack of impact on attendance as well as the absence Pay for performance holds great promise, but of change in teacher pedagogy, pupil atten- more evidence is needed dance, or pupil test scores. Performance-based bonuses that reward A program in rural India, however, pro- schools and teachers for improving student duced very different results.121 In randomly outcomes have been introduced in diverse set- selected rural schools run by nongovern- tings (such as India, Israel, and Kenya), and mental organizations, a schedule of monthly all have positive impacts on student learning teacher bonuses and fines based on attendance 90 LINKING SPENDING AND OUTCOMES GLOBAL MONITORING REPORT 2011 was monitored with daily date-stamped and effect was isolated from the resource effect by time-stamped photographs. The maximum increasing comparison facilities’ budgets by bonus for a teacher with no days absent was the average payments made to the treatment about 25 percent of a month’s salary. The facilities. The evaluation shows the greatest program had a dramatic effect on teacher impact on those services that had the high- absenteeism over three years. Although there est payment rates (institutional deliveries) were no observed changes in teachers’ class- and that needed the lowest provider effort room behavior and pedagogy (other than (preventive care visits by young children).126 greater presence), student test scores and Financial performance incentives improved rates of graduation to the next level of educa- both the use and quality of health services. tion rose significantly. The program was also quite cost effective. In health care, paying for performance Future directions is designed to provide fi nancial rewards for Development practitioners need to do much providers who increase service delivery and more rigorous empirical work—particu- improve the quality of care. To date, the larly, impact evaluations—to come closer to experience has been promising, but there is answering critical questions of development little evidence from evaluations using cred- effectiveness. But the start has been promis- ible comparison groups—in fact, a recent ing. Indeed, a recent review of CCTs concludes review of the literature finds methodologi- that no new research is needed on whether cal limitations that constrain the ability to households—particularly, poor households— interpret the results as causal impacts of pay respond to conditions or incentives.127 for performance.122 Several studies without Other questions remain unresolved or comparison groups found increases in service underaddressed—for example, the size of the delivery indicators, such as immunization incentive and its influence on the size of the and attended deliveries (as in Haiti),123 and impact,128 complementary supply-side inter- in the number of curative consultations and ventions or investments, payment frequency, institutional deliveries (as in early pilot pro- means of payment, and other features of pro- grams in Rwanda).124 gram design (such as how frequently condi- A quasi-experimental study in Cambo- tions should be monitored, the gender of the dia found service delivery improvements cash transfer recipient, and whether to penal- among contracted providers, but the provid- ize noncomplying households). ers received significantly more resources than A systematic base of impact evaluations did the control groups—suggesting that it of different interventions, in different con- was potentially the additional resources (the texts, will go a long way in moving toward resource effect) rather than pay for perfor- evidence-based decision making and in mak- mance (the incentive effect) that explained ing public spending more effective to close these findings.125 This raises an important the remaining gaps in education and health policy issue: if pay for performance achieves outcomes in developing countries. Some top- its results from increased financial resources ics, such as maternal mortality, are under- rather than incentives, the same results could represented in the impact evaluation litera- be achieved from an increase in traditional ture—a critical shortcoming given that this inputs, and there would be no reason to incur coincides with the MDG that is least likely the administrative costs associated with pay to be achieved. Also needed is more work on for performance. designing impact evaluations that have a basis The only impact evaluation in a low- in theory and provide useful information on resource setting that has been able to separate whether and why programs work. the incentive effect from the resource effect is The evidence suggests that service delivery a recent one of the Rwanda health center pay- for education and health is weak in develop- for-performance program targeting maternal ing countries. But very little is known about and child health outcomes. The incentive the best ways to measure and strengthen this GLOBAL MONITORING REPORT 2011 LINKING SPENDING AND OUTCOMES 91 critical link in the results chain. Exploring have improved the performance of service new and different approaches to measure and delivery units in education and health. improve service delivery is one of the most Policy makers are realizing that continu- promising areas for future work—especially, ing to spend ever-greater funds on health comparative evaluations that assess alterna- and education without quality improvements tive models of implementation. will neither bring their countries closer to the Cost and cost effectiveness should be an MDGs nor greatly improve human develop- explicit part of evaluation design. There rarely ment outcomes among poor people. More is discussion of the costs to achieve particular supply-side subsidies for health and edu- impacts. What magnitude of impact justifies cation services are not always the answer. the cost, or is it impact at any cost? Informa- The problems of supply-side dominance are, tion to implement at scale is often missing, among others, weak targeting of the poor, with no indication of the institutional invest- lack of consumer choice, absence of links ments to feasibly implement smaller experi- between provider payment and performance, ments at scale. This lack of information on weak incentives for service quality, and program costs and investments hampers deci- weak accountability to consumers. Supply- sion making and implementing at scale. side effects have dominated planning—and More work is also needed on the politi- too many multilateral development banks, cal economy surrounding the use of empiri- including the World Bank, are guilty of this. cal evidence to inform policy. Should donors Resource-needs projection models too often invest more in impact evaluation and inter- ignore the evidence suggesting a tenuous rela- vention design to improve political incen- tionship between increased health expendi- tives and government interventions? How ture and outcomes. can governments be encouraged to pursue The stronger and more explicit focus on policies that strengthen service delivery (by incentives in the health and education lit- improving the incentives of providers, among erature is encouraging. The modest human other things)? Some evidence on these issues development results do not stem from a short- is emerging. For example, enabling less- age of technical solutions. But a principal educated people in Brazil to use electronic finding of this review is that simply increasing voting technologies increased their political the size of interventions may not be enough participation, shifted public spending toward to make adequate progress. Even where public health care, improved service use by access and use have been expanded, human less-educated mothers, and thus reduced low- development outcomes have not necessarily weight births.129 Public spending on primary improved (as the experience with CCTs sug- education rose in Africa when countries gests). Establishing appropriate incentives for moved to greater multiparty competition.130 A suppliers and consumers, ensuring adequate similar association between electoral compe- provision of services to meet demand, ensur- tition and education spending has been found ing that suppliers are accountable to consum- in Indian states.131 But mobilizing communi- ers for the quality of service, and providing ties may have little impact on public provider sufficient resources to poor consumers are all accountability if communities are severely necessary to have an appreciable impact on constrained in taking public action.132 human development outcomes. Interest is growing in demand-side financ- ing approaches and more explicit use of incentives for service providers to improve Notes performance. On the demand side, the mas- 1. See, for example, Filmer, Hammer, and sive increase in CCTs has improved human Pritchett (1998) and Burnside and Dollar development outcomes. Similarly, vouchers (1998). Baldacci, Guin-Siu, and de Mello received increased prominence in health and (2002) conclude that the links between education. On the supply side, performance- outcomes and public spending are stron- based financing and pay for performance ger for education than for health. Gupta, 92 LINKING SPENDING AND OUTCOMES GLOBAL MONITORING REPORT 2011 Verhoeven, and Tiongson (2003) disaggre- 26. Clarke et al. 2008. gated the impact on the outcomes of the 27. Temperley et al. 2008. poor, and found (as did Bidani and Rival- 28. Pop-Eleches et al. 2011. lion [1997]), that the poor have worse health 29. Admassie, Abebaw, and Woldemichael outcomes than the nonpoor and that public 2009. spending has a larger impact on the health 30. Banerjee et al. 2010. outcomes of the poor than the nonpoor. 31. Regalía and Castro 2009. 2. “Basic immunization” here refers to diph- 32. For a detailed review of this debate, see theria, pertussis, and tetanus vaccination. Kremer and Holla (2008). Median coverage in low-income countries 33. Ashraf, Berry, and Shapiro 2007. increased from 62 percent to 81 percent 34. World Bank 2003. between 2000 and 2008 (WHO 2010). 35. Cohen and Dupas 2010. 3. Hanushek and Woessmann 2008. 36. Kremer and Miguel 2007. 4 Pratham Organization 2005. 37. Ibid. 5. WHO 2008. Full immunization refers to 38. In this regard, the vast body of health financ- basic immunization plus measles and polio ing and, specifically, the health insurance vaccination. literature (although not explicitly captured 6. WHO 2010. in this chapter) is an important source of 7. World Bank/AERC 2010. information. 8. See, for example, Van de Walle (1995), 39. Fiszbein and Schady 2009. Castro-Leal et al. (1999), Mahal et al. 40. Paxson and Schady 2007. (2000), Sahn and Younger (2000), and 41. Barber and Gertler 2010. O’Donnell et al. (2007). 42. Filmer and Schady 2009. 9. A complete discussion of impact evaluation 43. World Bank 2009, p. 182. methodologies can be found in Gertler et al. 44. Schultz 2004; de Janvry, Final, and Sadoulet (2010). 2006; Filmer and Schady 2010. 10. See, for example, Rodrik (2008), Ravallion 45. Edmonds and Schady 2009; Edmonds 2006; (2009), Deaton (2009), Sims (2010), and Case, Hosegood, and Lund 2005; Duflo Angrist and Pischke (2010). 2003. 11. Deaton 2009. 46. de Brauw and Hoddinott 2008; Schady and 12. Ibid. Araujo 2008. 13. See Kremer and Zwane (2007) for a review of 47. Baird, McIntosh, and Özler 2011. water and sanitation literature on preventing 48. Nguyen 2008. diarrheal diseases. 49. Jensen 2010. 14. Kremer and Holla 2009. 50. Jalan and Somanathan 2008. 15. Kremer, Moulin, and Namunyu 2003. 51. Madajewicz et al. 2007. 16. Vermeersch and Kremer 2004; Kazianga, 52. Dupas 2009. de Walque, and Alderman 2009; Meng and 53. Downs, Loewenstein, and Wisdom 2009; Ryan 2010; Powell et al. 1998. Davis et al. 2009. 17. Schultz 2004. 54. Bertrand et al. 2006; Medley et al. 2009. 18. See, for example, Cardoso and Souza (2004), 55. Duflo et al. 2006. Levy and Ohls (2007), Barrera-Osorio et al. 56. Dupas 2006, 2009. (2008), and Filmer and Schady (2008). 57. For a full review, see Kremer and Holla 19. Barrera-Osorio, Linden, and Urquiola 2007; (2009). Borkum 2009. 58. Glewwe, Kremer, and Moulin 2007. The 20. Hanushek 2008. authors argue that problems with the edu- 21. See the review in Kremer and Holla (2009). cation system as a whole constrained the 22. It should be noted that a recent study (not an impact of increasing inputs. For example, impact evaluation) from Egypt (Hanushek, textbooks are written in English (most stu- Lavy, and Hitomi 2008) has shown that dents’ third language) and many students school dropout decisions are very responsive could not use them effectively. More gener- to the quality of the school. ally, the curriculum in Kenya and in many 23. Afridi 2007; Jacoby, Cueto, and Pollitt 1996 other developing countries tends to be ori- 24. Kazianga, de Walque, and Alderman 2009. ented toward academically strong students. 25. Vermeersch and Kremer 2004. 59. Glewwe et al. 2004. GLOBAL MONITORING REPORT 2011 LINKING SPENDING AND OUTCOMES 93 60. Vermeersch and Kremer 2004; Kazianga, de the impact of growth monitoring on nutri- Walque, and Alderman 2009. tional outcomes is weak. This may explain 61. Newman et al. 2002. the mixed fi ndings. “Available evidence on 62. Banerjee et al. 2005. growth monitoring was not sufficient to sup- 63. Duflo, Dupas, and Kremer 2007; Kremer, port its use alone (without adequate nutri- Moulin, and Namunyu 2003; Banerjee et al. tion counseling and referrals) as an essential 2005; Banerjee et al. 2007. nutrition support” (Bhutta et al. 2008, p. 64. Jamison, Searle, Galda, and Heyneman 1981. 429). See also Roberfroid et al. (2005); Gar- 65. Banerjee et al. 2007. ner et al. (2000); and Ashworth, Shrimpton, 66. He, Linden, and MacLeod 2007. and Jamil (2008). 67. Barrera-Osorio and Linden 2009. 84. Rivera et al. 2004; Gertler 2004. 68. Banerjee et al. 2010. 85. Behrman and Hoddinott 2001. 69. Chile (Galasso 2006); Honduras (Morris et 86. Adato and Roopnaraine 2004. al. 2004); Jamaica (Levy and Ohls 2007); 87. Leroy, Marie, and Verhofstadt 2009. Mexico (Gertler and Boyce 2001); Nicara- 88. Schady and Rosero 2008. gua (Maluccio and Flores 2005); Paraguay 89. Macours, Schady, and Vakis 2008. (Soares, Ribas, and Hirata 2008). 90. Other ways are available to increase incen- 70. Nicaragua (Barham and Maluccio 2008); tives facing providers by changing the way Turkey (Ahmed et al. 2007). in which budgets or provider payments are 71. Nicaragua (Macours, Schady, and Vakis structured. Thus, although improving incen- 2008); Honduras (Hoddinott 2008); Brazil tives may be a useful side effect of voucher (Morris et al. 2004); Ecuador (Paxson and systems, it may be quite a costly option. Schady 2008). 91. Behrman and Hoddinott 2001. 72. Attanasio et al. 2005. 92. Ecuador (Paxson and Schady 2008); Nica- 73. The increase in institutional deliveries and ragua (Macours, Schady, and Vakis 2008). a smaller increase in home delivery with a 93. Macours, Schady, and Vakis 2008. skilled birth attendant suggest that part of 94. Grantham-McGregor et al. 2007. the increase in the number of births in facili- 95. Paxson and Schady 2008. ties resulted from shifting births that other- 96. This is consistent with the conclusion reached wise would have occurred at home with a by Cunha and Heckman (2010), as well as skilled attendant (Lim et al. 2010). a recent World Bank publication on early 74. Among others, see Schultz (2004), Cardoso childhood development (Alderman 2011). and Souza (2004), Levy and Ohls (2007), 97. Worldwide, a worrying dropout occurs in the Barrera-Osorio et al. (2008), and Filmer and transition to secondary school. Attanasio et Schady (2008). al. (2005) argue that eliminating the funding 75. Miguel and Kremer 2004; Glewwe, Kremer, transfers to children in grade six and below and Moulin 2007. and using those resources to increase the size 76. Baird et al. 2010. of the transfer to children in grade seven and 77. Filmer and Schady 2009. beyond would double school participation 78. Growth monitoring services include weighing among the older children, with no effect on children at regular intervals during the fi rst school participation by younger ones. few years after birth and tracking growth 98. Chaudhury et al. 2006. against expected growth at a given age. 99. Abadzi 2009. 79. Mexico (Behrman and Hoddinott 2001; 100. Das and Hammer 2005. Gertler 2004; and Rivera et al. 2004); 101. A detailed discussion is in World Develop- Colombia (Mesnard 2005); Nicaragua ment Report 2004 (World Bank 2003). (Maluccio and Flores 2005); Colombia 102. World Bank 2003. (Attanasio et al. 2005). 103. This classification comes from the recent 80. Rivera et al. 2004; Attanasio et al. 2005; work on school accountability done by Maluccio and Flores 2005. Bruns, Filmer, and Patrinos (2011). 81. Morris et al. (2004). 104. The most important differences between 82. Caldés, Coady, and Maluccio 2006. education and health services are because 83. It is curious that weighing and growth of information asymmetry and the relative monitoring featured so prominently in the infrequency of contact between health pro- CCTs because the epidemiologic evidence of viders and their patients. 94 LINKING SPENDING AND OUTCOMES GLOBAL MONITORING REPORT 2011 105. Andrabi, Das, and Khwaja 2009. Afridi, F. 2007. “Child Welfare Programs and 106. Piper and Korda 2010. Child Nutrition: Evidence from a Mandated 107. Mizala, Romaguera, and Urquiola 2007. School Meal Program in India.” Journal of 108. Reinikka and Svensson 2005, 2006; Björk- Development Economics 92 (2): 152–65. man 2006; Björkman and Svensson 2009. Ahmed, A., M. Adato, A. Kudat, D. Gilligan, and 109. Banerjee et al. 2010. R. Colasan. 2006. “Interim Impact Evaluation 110. Pandey et al. 2010. of the Conditional Cash Transfer Program in 111. Banerjee et al. 2010. Turkey. Final Report.” International Food Pol- 112. Skoufi as and Shapiro 2006; Gertler, Patri- icy Research Institute, Washington, DC. nos, and Rubio-Codina 2006. Alderman, H., ed. 2011. No Small Matter: The 113. Chaudhury and Parajuli 2010. Impact of Poverty, Shocks, and Human Capi- 114. Muralidharan and Sundararaman 2009; tal Investments in Early Childhood Develop- Goyal and Pandey 2009. ment. Washington, DC: World Bank. 115. Muralidharan and Sundararaman 2009. Andrabi, T., J. Das, and A. Khwaja. 2009. “Report 116. Mizala, Romaguera, and Urquiola 2007. Cards: The Impact of Providing School and 117. Gallego 2008. Child Test Scores on Educational Markets.” 118. Glewwe, Ilas, and Kremer 2003. Unpublished manuscript, World Bank, Wash- 119. Ferraz and Bruns forthcoming. ington, DC. 120. Kremer and Chen 2001. Angrist, J. D., and J. Pischke. 2010. “The Cred- 121. Duflo, Hanna, and Ryan 2010. ibility Revolution in Empirical Economics: 122. Eldridge and Palmer 2009. How Better Research Design Is Taking the 123. Eichler et al. 2007. Con Out of Econometrics” Journal of Eco- 124. Meessen, Kashala, and Musango 2007; nomic Perspectives 24 (2): 3–30. Meessen et al. 2006; Soeters, Habineza, and Ashraf, N., J. Berry, and J. M. Shapiro. 2007. Peerenboom 2006. “Can Higher Prices Stimulate Product Use? 125. Schwartz and Bhushan 2004. Evidence from a Field Experiment in Zambia.” 126. Basinga et al. 2010. Working Paper 13247, National Bureau of 127. Gaarder, Glassman, and Todd 2010. Economic Research, Cambridge, MA. 128. Filmer and Schady (2009), in the context Ashworth, A., R. Shrimpton, and K. Jamil. 2008. of education enrollment, conclude that “Growth Monitoring and Promotion: Review there are diminishing returns to the size of of Evidence of Impact.” Maternal and Child incentives. Nutrition 4 (Suppl 1): 86–117. 129. Fujiwara 2010. Attanasio, O., E. Battistin, E. Fitzsimmons, A. 130. Stasavage 2005. Mesnard, and M. Vera-Hernández. 2005. 131. Khemani 2010. “How Effective Are Conditional Cash Trans- 132. Banerjee et al. 2010. fers? Evidence from Colombia.” Briefi ng Note 54, Institute for Fiscal Studies, London. Baird, S., E. Chirwa, C. McIntosh, and B. Özler. References 2010. “The Short-Term Impacts of a School- Abadzi, H. 2009. “Instructional Time Loss in ing Conditional Cash Transfer Program on the Developing Countries: Concepts, Measure- Sexual Behavior of Young Women.” Health ment, and Implications.” World Bank Research Economics 19: 55–68. Observer 24 (2): 267–90. Baird, S., C. McIntosh, and B. Özler. 2011. “Cash Adato, M., and T. Roopnaraine. 2004. “Sistema or Condition? Evidence from a Cash Transfer de evaluación de la Red de Protección Social Experiment.” Policy Research Working Paper de Nicaragua: Un análisis social de la ‘Red de 5259, World Bank, Washington, DC. Protección Social’ (RPS) en Nicaragua.” Inter- Baldacci, E., M. Guin-Siu, and L. de Mello. 2002. national Food Policy Research Institute, Wash- “More on the Effectiveness of Public Spending ington, DC. on Health Care and Education: A Covariance Admassie, A., D. Abebaw, and A. Woldemi- Structure Model.” Working Paper WP/02/90, chael. 2009. “Impact Evaluation of the Ethio- International Monetary Fund, Washington pian Health Services Extension Programme.” DC. Journal of Development Effectiveness 1 (4): Banerjee, A. V., R. Banerji, E. Duflo, R. Glenner- 430–49. ster, and S. Khemani. 2010. “Pitfalls of GLOBAL MONITORING REPORT 2011 LINKING SPENDING AND OUTCOMES 95 Participatory Programs: Evidence from a Ran- Behrman, J., and J. Hoddinott. 2001. “An Evalu- domized Evaluation in Education in India.” ation of the Impact of Progresa on Pre-school American Economic Journal: Economic Pol- child Height.” Discussion Paper 104, Interna- icy 2 (1): 1–30. tional Food Policy Research Institute, Wash- Banerjee, A., E. Duflo, S. Cole, and L. Linden. ington, DC. 2007. “Remedying Education: Evidence from Bertrand, J. T., K. O’Reilly, J. Denison, R. Two Randomized Experiments in India.” Anhang, and M. Sweat. 2006. “Systematic Quarterly Journal of Economics 122 (3): Review of the Effectiveness of Mass Communi- 1235–64. cation Programs to Change HIV/AIDS-related Banerjee, A., E. Duflo, R. Glennerster, and D. Behaviors in Developing Countries.” Health Kothari. 2010. “Improving Immunisation Cov- Education Research 21 (4): 567–97. erage in Rural India: Clustered Randomised Bhutta, Z., T. Ahmed, R. Black, S. Cousens, K. Controlled Evaluation of Immunisation Cam- Dewey, E. Giugliani, B. Haider, B. Kirkwood, paigns With and Without Incentives.” British S. Morris, H. P. S. Sachdev, and M. Shekar. Medical Journal 340 (1): c2220. 2008. “What Works? Interventions for Mater- Banerjee, A., S. Jacob, and M. Kremer, with J. nal and Child Undernutrition and Survival.” Lanjouw, and P. Lanjouw. 2005. “Promot- Lancet 371: 417–40. ing School Participation in Rural Rajasthan: Bidani, B., and M. Ravallion. 1997. “Decomposing Results from Some Prospective Trials.” Unpub- Social Indicators Using Distributional Data.” lished manuscript, Massachusetts Institute of Journal of Econometrics 77 (1): 125–39. Technology, Cambridge, MA. Björkman, M. 2006. “Does Money Matter for Barber, S. L., and P. J. Gertler. 2010. “Empower- Student Performance? Evidence from a Grant ing Women: How Mexico’s Conditional Cash Program in Uganda.” Working Paper 326, Transfer Program Raised Prenatal Care Qual- Institute for International Economic Studies, ity and Birth Weight.” Journal of Development Stockholm University, Sweden. Effectiveness 2 (1): 51–73. Björkman, M., and J. Svensson. 2009. “Power Barham, T., and J. Maluccio. 2008. “The Effect to the People: Evidence from a Randomized of Conditional Cash Transfers on Vaccination Experiment of a Community-based Monitor- Coverage in Nicaragua.” Health and Society ing Project in Uganda.” Quarterly Journal of Working Paper HS2008-01, Institute of Behav- Economics 124: 734–69. ioral Science, University of Colorado, Boulder, Bobonis, G., and F. Finan. 2009. “Neighborhood CO. Peer Effects in Secondary School Enrollment Barrera-Osorio, F., M. Bertrand, L. L. Linden, Decisions.” Review of Economics and Statis- and F. Perez-Calle. 2008. “Conditional Cash tics 91 (4): 695–716. Transfers in Education: Design Features, Peer Borkum, E. 2009. “Can Eliminating School Fees and Sibling Effects Evidence from a Ran- in Poor Districts Boost Enrollment? Evidence domized Experiment in Colombia.” Policy from South Africa.” Discussion Paper 0910- Research Working Paper 4580, World Bank, 06, Columbia University, New York. Washington, DC. Bruns, B., D. Filmer, and H. A. Patrinos. 2011. Barrera-Osorio, F., and L. L Linden. 2009. “The Making Schools Work: New Evidence on Use and Misuse of Computers in Education: Accountability Reforms. Washington, DC: Evidence from a Randomized Experiment in World Bank. Colombia.” Policy Research Working Paper Burnside, C., and D. Dollar. 1998. “Aid, the 4836, World Bank, Washington, DC. Incentive Regime, and Poverty Reduction.” Barrera-Osorio, F., L. L Linden, and, M. Urquiola. Policy Research Working Paper 1937, World 2007. “The Effects of User Fee Reductions Bank, Washington, DC. on Enrollment: Evidence from a Quasi- Caldés, N., D. Coady, and J. Maluccio. 2006. Experiment.” Working Paper, Columbia Uni- “The Cost of Poverty Alleviation Transfer Pro- versity, New York. grams: A Comparative Analysis of Three Pro- Basinga P., P. J. Gertler, A. Binagwaho, A. Soucat, grams in Latin America.” World Development J. Sturdy, and C. Vermeersch. 2010. “Paying 34 (5): 818–37. Primary Health Care Centers for Performance Cardoso, E., and A. Souza. 2004. “The Impact in Rwanda.” Policy Research Working Paper of Cash Transfers on Child Labor and School 5190, World Bank, Washington, DC. Attendance in Brazil.” Working Paper 04-W07, 96 LINKING SPENDING AND OUTCOMES GLOBAL MONITORING REPORT 2011 Department of Economics, Vanderbilt Univer- for the Elusive Keys to Economic Develop- sity, Nashville, TN. ment.” Working Paper 14690, National Bureau Case, A., V. Hosegood, and F. Lund. 2005. “The of Economic Research, Cambridge, MA. Reach and Impact of Child Support Grants: de Brauw, A., and J. Hoddinott. 2008. “Must Evidence from KwaZulu-Natal.” Development Conditional Cash Transfer Programs Be Con- Southern Africa 22: 467–82 ditioned to Be Effective? The Impact of Con- Castro-Leal, F., J. Dayton, L. Demery, and K. ditioning Transfers on School Enrollment in Mehra. 1999. “Public Social Spending in Mexico.” Discussion Paper 757, International Africa: Do the Poor Benefit?” World Bank Food Policy Research Institute, Washington, Research Observer 14 (1): 49–72. DC. Chaudhury, N., J. Hammer, M. Kremer, K. de Janvry, A., F. Finan, and E. Sadoulet. 2006. Muralidharan, and F. H. Rogers. 2006. “Miss- “Can Conditional Cash Transfer Programs ing in Action: Teacher and Health Worker Serve as Safety Nets in Keeping Children in Absence in Developing Countries.” Journal of School and from Working When Exposed to Economic Perspectives 20 (1): 91–116. Shocks?” Journal of Development Economics Chaudhury, N., and D. Parajuli. 2010. “Giving 79: 349–73. It Back: Evaluating the Impact of Devolution de Walque, D., H. Kazianga, and M. Over. 2010. of School Management to Communities in “Antiretroviral Therapy Awareness and Risky Nepal.” Unpublished manuscript, World Bank, Sexual Behaviors: Evidence from Mozam- Washington, DC. bique.” Policy Research Working Paper 5486, Clarke, S. E., M. C. H. Jukes, K. Njagi, L. Kha- World Bank, Washington, DC. sakhala, B. Cundill, J. Otido, C. Crudder, Downs, J. S., G. Loewenstein, and J. Wisdom. B. Estambale, and S. Brooker. 2008. “Effect 2009. “Strategies for Promoting Healthier of Intermittent Preventive Treatment on Food Choices.” American Economic Review Health and Education in Schoolchildren: A 99 (2): 159–64. Cluster-Randomised, Double-Blind, Placebo- Duflo, E. 2003. “Grandmothers and Grand- Controlled Trial.” Lancet 372: 127–38. daughters: Old-Age Pension and Intrahouse- Cohen, J., and P. Dupas. 2010. “Free Distribution hold Allocation in South Africa.” World Bank or Cost-Sharing? Evidence from a Randomized Economic Review 17: 1–25. Malaria Prevention Experiment.” Quarterly Duflo, E., P. Dupas, and M. Kremer. 2007. “Peer Journal of Economics 125 (1): 1–45. Effects, Pupil-teacher Ratios, and Teacher Cohen, J., P. Dupas, and S. Schaner. 2010. “Prices, Diagnostic Tests and the Demand for Malaria Incentives.” Unpublished manuscript, Harvard Treatment: Evidence from a Randomized University, Cambridge, MA. Trial.” Working Paper, Harvard University, Duflo, E., P. Dupas, M. Kremer, and S. Sinei. Cambridge, MA. 2006. “Education and HIV/AIDS Prevention: Cunha, F., and J. Heckman. 2010. “Investing in Evidence from a Randomized Evaluation in Our Young People.” Working Paper 16201, Western Kenya.” Policy Research Working National Bureau of Economic Research, Cam- Paper 4024, World Bank, Washington, DC. bridge, MA. Duflo, E., R. Hanna, and S. Ryan. 2010. “Incen- Das, J., and J. Hammer. 2005. “Money for Noth- tives Work: Getting Teachers to Come to ing: The Dire Straits of Medical Practice in School.” Unpublished manuscript, University Delhi, India.” Policy Research Paper 3669, of Chicago, Chicago, IL. World Bank, Washington, DC. Dupas, P. 2006. “Relative Risks and the Market Davis, K. C., M. C. Farrelly, P. Messeri, and J. for Sex: Teenagers, Sugar Daddies, and HIV in Duke. 2009. “The Impact of National Smok- Kenya.” Unpublished manuscript, Dartmouth ing Prevention Campaigns on Tobacco-Related College, Hanover, NH. Beliefs, Intentions to Smoke and Smoking Ini- ——— . 2009. “Do Teenagers Respond to HIV tiation: Results from a Longitudinal Survey Risk Information? Evidence from a Field of Youth in the United States.” International Experiment in Kenya.” Working Paper 14707, Journal of Environmental Research and Pub- National Bureau of Economic Research, Cam- lic Health 6 (2): 722–40. bridge, MA. Deaton, A. 2009. “Instruments of Development: Edmonds, E. V. 2006. “Child Labor and School- Randomization in the Tropics, and the Search ing Responses to Anticipated Income Shock in GLOBAL MONITORING REPORT 2011 LINKING SPENDING AND OUTCOMES 97 South Africa.” Journal of Development Eco- Working Paper 4424, World Bank, Washing- nomics 81: 386–414. ton, DC. Edmonds, E. V., and N. Schady. 2009. “Pov- Garner, P., R. Panpanich, S. Logan, and D. Davis. erty Alleviation and Child Labor.” Working 2000. “Is Routine Growth Monitoring Effec- Paper 15345, National Bureau of Economic tive? A Systematic Review of Trials.” Archives Research, Cambridge, MA. of Diseases in Childhood 82: 197–201. Eichler, R., P. Auxila, A. Uder, and B. Desman- Gertler, P. J. 2004. “Do Conditional Cash Trans- gles. 2007. “Performance-Based Incentives for fers Improve Child Health? Evidence from Health: Six Years of Results from Supply-Side Progresa’s Control Randomized Experiment.” Programs in Haiti.” Working Paper 121, Cen- American Economic Review 94 (2): 336–41. ter for Global Development, Washington, DC. Gertler, P. J., and S. Boyce. 2001. “An Experiment Eldridge, C., and N. Palmer. 2009. “Performance- in Incentive-Based Welfare: The Impact of Pro- Based Payment: Some Reflections on the Dis- gresa on Health in Mexico.” Royal Economic course, Evidence and Unanswered Questions.” Society Annual Conference 2003, No. 85. Health Policy and Planning 24 (3): 160–66. Gertler, P. J., S. Martinez, P. Premand, L. B. Raw- Ferraz, C., and B. Bruns. Forthcoming. “Incen- lings, and C. M. J. Vermeersch. 2010. Impact tives to Teach: The Effects of Performance Pay Evaluation in Practice. Washington, DC: in Brazilian Schools.” World Bank, Washing- World Bank. ton, DC. Gertler, P. J., H. A. Patrinos, and M. Rubio- Filmer, D. 2003. “The Incidence of Public Expen- Codina. 2006. “Empowering Parents to diture on Health and Education.” Background Improve Education: Evidence from Rural Mex- Note for World Development Report 2004, ico.” Policy Research Working Paper 3935, World Bank, Washington, DC. World Bank, Washington, DC. Filmer, D., J. Hammer, and L. Pritchett. 1998. Glewwe, P., N. Ilas, and M. Kremer. 2003. “Health Policy in Poor Countries: Weak Links “Teacher Incentives.” Working Paper 9671, in the Chain?” Policy Research Working Paper National Bureau of Economic Research, Cam- 1874, World Bank, Washington, DC. bridge, MA. Filmer, D., and N. Schady. 2008. “Getting Girls Glewwe, P., M. Kremer, and S. Moulin. 2007. into School: Evidence from a Scholarship Pro- “Many Children Left Behind? Textbooks and gram in Cambodia.” Economic Development Test Scores in Kenya.” Working Paper 13300, and Cultural Change 56: 581–617. National Bureau of Economic Research, Cam- ———. 2009. “Are There Diminishing Returns to Transfer Size in Conditional Cash Transfers?” bridge, MA. Policy Research Working Paper 4999, World Glewwe, P., M. Kremer, S. Moulin, and E. Zitze- Bank, Washington, DC. witz. 2004. “Retrospective vs. Prospective ——— . 2010. “Does More Cash in Conditional Analyses of School Inputs: The Case of Flip Cash Transfer Programs Always Lead to Charts in Kenya.” Journal of Development Larger Impacts on School Attendance?” Jour- Economics, 74: 251–68. nal of Development Economics. Go, D. 2010. “Global Monitoring Report 2011 Fiszbein, A., and N. Schady. 2009. Conditional Approach Paper.” Unpublished manuscript, Cash Transfers: Reducing Present and Future World Bank, Washington, DC. Poverty. Washington DC: World Bank. Gong, R. 2010. “HIV Testing and Risky Sexual Fujiwara, T. 2010. “Voting Technology, Political Behavior.” Job Market Paper, PhD Candidate, Responsiveness, and Infant Health: Evidence University of California-Berkeley, Berkeley, from Brazil.” Department of Economics, Uni- CA. versity of British Columbia, Vancouver. http:// Goyal, S., and P. Pandey. 2009. “How Do Gov- grad.econ.ubc.ca/fujiwara/jmp.pdf. ernment and Private Schools Differ? Findings Gaarder, M., A. Glassman, and J. Todd. 2010. from Two Large Indian States.” South Asia “Conditional Cash Transfers and Health: Human Development Sector Report No. 30, Unpacking the Causal Chain.” Journal of World Bank, Washington, DC. Development Effectiveness 2 (1): 6–50. Grantham-McGregor, S., Y. B. Cheung, S. Cueto, Galasso, E., and N. Umapathi. 2007. “Improving P. Glewwe, L. Richer, B. Trupp, and the Inter- Nutritional Status through Behavioral Change: national Child Development Steering Group. Lessons from Madagascar.” Policy Research 2007. “Developmental Potential in the First 5 98 LINKING SPENDING AND OUTCOMES GLOBAL MONITORING REPORT 2011 Years for Children in Developing Countries.” Khemani, S. 2010. “Political Economy of Infra- Lancet 369: 60–70. structure Spending in India.” Policy Research Gupta, S., M. Verhoeven, and E. Tiongson. 2003. Working Paper 5423, World Bank, Washing- “Public Spending on Health Care and the ton, DC. Poor.” Health Economics 12: 685–96. Kremer, M. E., and D. Chen. 2001. “Interim Hanushek, E. A. 2008. “Incentives for Efficiency Report on a Teacher Incentive Program in and Equity in the School System.” Perspektiven Kenya.” Unpublished manuscript, Harvard der Wirtschaftspolitik 9 (Special Issue): 5–27. University, Cambridge, MA. Hanushek, E. A., V. Lavy, and K. Hitomi. 2008. Kremer, M., and A. Holla. 2008. “Pricing and “Do Students Care about School Quality? Access: Lessons from Randomized Evaluations Determinants of Dropout Behavior in Devel- in Education and Health.” Unpublished manu- oping Countries.” Journal of Human Capital script, Harvard University, Cambridge, MA. 1 (2): 69–105. ———. 2009. “Improving Education in the Devel- Hanushek, E. A., and L. Woessmann. 2008. “The oping World: What Have We Learned from Role of Cognitive Skills in Economic Develop- Randomized Evaluations?” Annual Review of ment.” Journal of Economic Literature 46 (3): Economics 1: 513–45. 607–68. Kremer, M., and E. Miguel. 2007. “The Illusion He, F., L. Linden, and M. MacLeod. 2007. of Sustainability.” Quarterly Journal of Eco- “Teaching What Teachers Don’t Know: An nomics 112 (3): 1007–65. Assessment of the Pratham English Language Kremer, M., E. Miguel, and R. Thornton. 2009. Program.” Unpublished manuscript, Columbia “Incentives to Learn.” Review of Economics University, New York. and Statistics 91 (3): 437–56. Hoddinott, J. 2008. “Nutrition and Conditional Kremer, M., S. Moulin, and R. Namunyu. 2003. Cash Transfer (CCT) Programs.” Unpub- “Decentralization: A Cautionary Tale.” lished manuscript, International Food Policy Unpublished manuscript, Harvard University, Research Institute, Washington, DC. Cambridge, MA. IHME (Institute for Health Metrics and Evalua- Kremer, M., and A. P. Zwane. 2007. “What tion). 2010. “Financing Global Health 2009: Works in Fighting Diarrheal Diseases in Devel- Tracking Development Assistance for Health.” oping Countries? A Critical Review.” World University of Washington, Seattle, WA. Bank Research Observer 22:1–24. Jacoby, E., S. Cueto, and E. Pollitt. 1996. “Ben- Lalive, R., and A. Cattaneo. 2006. “Social Inter- efits of a School Breakfast Program among actions and Schooling Decisions.” Discussion Andean Children in Huaraz, Peru.” Food and Paper 2250, Institute for the Study of Labor Nutrition Bulletin 17 (1): 54–64. (IZA), Bonn, Germany. Jalan, J., and E. Somanathan. 2008. “The Impor- Legovini, A. 2010. “Development Impact Evalu- tance of Being Informed: Experimental Evi- ation Initiative: A World Bank–Wide Strategic dence on Demand for Environmental Qual- Approach to Enhancing Development Effec- ity.” Journal of Development Economics 87 tiveness” Unpublished manuscript, World (1) 14–28. Bank, Washington, DC. Jamison, D., B. Searle, K. Galda, and S. Leroy, J. L., R. Marie, and E. Verhofstadt. 2009. Heyneman. 1981. “Improving Elementary “The Impact of Conditional Cash Transfer Mathematics Education in Nicaragua: An Programmes on Child Nutrition: A Review of Experimental Study of the Impact of Text- Evidence Using a Programme Theory Frame- books and Radio on Achievement.” Journal of work.” Journal of Development Effectiveness Educational Psychology 73 (4): 556–67. 1: (2) 103–29. Jensen, R. 2010. “The (Perceived) Returns to Edu- Levy, D., and J. Ohls. 2007. “Evaluation of Ja- cation and Demand for Schooling.” Quarterly maica’s PATH Program: Final Report.” Math- Journal of Economics 125 (2): 515–48. ematica Policy Research, Washington, DC. Kazianga, H., D. de Walque, and H. Alderman. Lim, S., L. Dandona, J. Hoisington, S. James, M. 2009. “Educational and Health Impacts of Hogan, and E. Gakidou. 2010. “India’s Janani Two School Feeding Schemes: Evidence from Suraksha Yojana, a Conditional Cash Trans- a Randomized Trial in Rural Burkina Faso.” fer Programme to Increase Births in Health Policy Research Working Paper 4976, World Facilities: An Impact Evaluation.” Lancet 375: Bank, Washington, DC. 2009–23. GLOBAL MONITORING REPORT 2011 LINKING SPENDING AND OUTCOMES 99 Macours, K., N. Schady, and R. Vakis. 2008. Mesnard, A. 2005. “Evaluation of the Familias “Cash Transfers, Behavioral Changes, and the en Acción Program in Colombia: Do Condi- Cognitive Development of Young Children: tional Subsidies Improve Education, Health Evidence from a Randomized Experiment.” and Nutritional Outcomes?” Institute for Fis- Policy Research Working Paper 4759, World cal Studies, London. Bank, Washington, DC. Miguel, E., and M. Kremer. 2004. “Worms: Iden- Madajewicz, M., A. Pfaff, A. van Geen, J. Gra- tifying Impacts on Education and Health in the ziano, I. Hussein, H. Monotaj, R. Sylvi, and H. Presence of Treatment Externalities.” Econo- Ahsan. 2007. “Can Information Alone Change metrica 72 (1): 159–217. Behavior? Response to Arsenic Contamination Mizala, A., P. Romaguera, and M. Urquiola. of Groundwater in Bangladesh.” Journal of 2007. “Socioeconomic Status or Noise? Trade- Development Economics 84 (2) 731–54. offs in the Generation of School Quality Infor- Mahal A., J. Sing, F. Afridi, V. Lamba, A. Gumber, mation.” Journal of Development Economics and V. Selvaraju. 2000. “Who Benefits from 84: 61–75. Public Health Spending in India?” National Morris, S., P. Olinto, R. Flores, E.A.F. Nilson, Council for Applied Economic Research, New and A. C. Figueiró. 2004. “Conditional Cash Delhi. Transfers Are Associated with a Small Reduc- Maluccio, J., and R. Flores. 2005. “Impact Evalu- tion in the Weight Gain of Preschool Children ation of a Conditional Cash Transfer Program: in Northeast Brazil.” Journal of Nutrition 134: The Nicaraguan Red de Protección Social.” 2336–41. Research Report 141, International Food Pol- Muralidharan, K., and V. Sundararaman. 2009. icy Research Institute, Washington, DC. “Teacher Performance Pay: Experimental Mavedzenge, S., A. Doyle, and D. Ross. 2010. Evidence from India.” Working Paper 15323, “HIV Prevention in Young People in Sub- National Bureau of Economic Research, Cam- Saharan Africa: A Systematic Review.” Infec- bridge, MA. tious Disease Epidemiology Unit, London Newman, J. L., M. Pradhan, L. B. Rawlings, G. School of Hygiene and Tropical Medicine, Ridder, R. Coa, and J. L. Evia. 2002. “An London. Impact Evaluation of Education, Health, and McCoy, S., R. Kangwende, and N. Padian. 2009. Water Supply Investments by the Bolivian “Behavior Change Interventions to Prevent Social Investment Fund.” World Bank Eco- HIV among Women Living in Low and Mid- nomic Review 16 (2): 241–74. dle Income Countries.” Synthetic Review 008, International Initiative for Impact Evaluation, Nguyen, T. 2008. “Information, Role Models and New Delhi, India. Perceived Returns to Education: Experimental Medley, A., C. Kennedy, K. O’Reilly, and M. Evidence from Madagascar.” Working Paper, Sweat. 2009. “Effectiveness of Peer Education Massachusetts Institute of Technology, Cam- Interventions for HIV Prevention in Develop- bridge, MA. ing Countries: A Systematic Review and Meta- O’Donnell, O., E. van Doorslaer, R. Rannan- Analysis.” AIDS Education and Prevention 21 Eliya, A. Somanathan, S. Adhikari, D. Harbi- (3): 181–206. anto, C. Garg, P. Hanvoravongchai, M. Huq, Meessen, B., J. P. Kashala, and L. Musango. 2007. A. Karan, G. Leung, C. W. Ng, B. R. Pande, “Output-Based Payment to Boost Staff Produc- K. Tin, K. Tisayaticom, L. Trisnantoro, Y. tivity in Public Health Centres: Contracting in Zhang, and Y. Zhao. 2007. “The Incidence of Kabutare District, Rwanda.” Bulletin of the Public Spending on Healthcare: Comparative World Health Organization 85 (2): 108–15. Evidence from Asia.” World Bank Economic Meessen, B., L. Musango, J. P. Kashala, and J. Review 21 (1): 93–123. Lemlin. 2006. “Reviewing Institutions of OECD-DAC (Organisation for Economic Co- Rural Health Centres: The Performance Initia- operation and Development-Development tive in Butare, Rwanda.” Tropical Medicine & Assistance Committee). 2009. International International Health 11 (8): 1303–17. Development Statistics: Online Databases on Meng, X., and J. Ryan. 2010. “Does a Food for Aid and Other Resource Flows. Paris, France. Education Program Affect School Outcomes?” http://www.oecd.org/dac/stats/idsonline. The Bangladesh Case. Journal of Population Paxson, C., and N. Schady. 2007. “Cognitive Devel- Economics 23 (2): 415–47. opment among Young Children in Ecuador: 100 LINKING SPENDING AND OUTCOMES GLOBAL MONITORING REPORT 2011 The Roles of Wealth, Health, and Parenting.” and Nutrition (Progresa) on Rates of Growth Journal of Human Resources 42 (1): 49–84. and Anemia in Infants and Young Children: A ——— . 2008. “Does Money Matter? The Effects Randomized Effectiveness Study.” JAMA 291 of Cash Transfers on Child Health and Devel- (21): 2563–70. opment in Rural Ecuador.” Unpublished man- Roberfroid, D., P. Kolsteren, T. Hoerée, and B. uscript, World Bank, Washington, DC. Maire. 2005. “Do Growth Monitoring and Piper, B., and M. Korda. 2010. “EGRA Plus: Libe- Promotion Programs Answer the Performance ria. Program Evaluation Report. Draft.” RTI Criteria of a Screening Program? A Criti- International, Research Triangle Park, NC. cal Analysis Based on a Systematic Review.” Pop-Eleches C., H. Thirumurthy, J. P. Habya- Tropical Medicine & International Health 10: rimana, J. G. Zivin, M. P. Goldstein, D. de 1121–33. Walque, L. Mackeen, J. Haberer, S. Kimaiyo, Rodrik, D. 2008. “The New Development Eco- J. Sidle, D. Ngare, and D. R. Bangsberg. 2011. nomics: We Shall Experiment, But How Shall “Mobile Phone Technologies Improve Adher- We Learn?” Unpublished manuscript, John F. ence to Antiretroviral Treatment in a Resource- Kennedy School of Government, Harvard Uni- Limited Setting: A Randomized Controlled versity, Cambridge, MA. Trial of Text Message Reminders.” AIDS 25 Sahn, D., and S. Younger. 2000. “Expenditure (6): 825–34. Incidence in Africa: Microeconomic Evidence.” Powell, C. A., S. P. Walker, S. M. Chang, and S. Fiscal Studies 21 (3): 321–48. M. Grantham-McGregor. 1998. “Nutrition Schady, N., and M. C. Araujo. 2008. “Cash and Education: A Randomized Trial of the Transfers, Conditions, and School Enrollment Effects of Breakfast in Rural Primary School in Ecuador.” Economía 8 (2): 43–70. Schady, N., and J. Rosero. 2008. “Are Cash Children.” American Journal of Clinical Transfers Made to Women Spent Like Other Nutrition 68: 873–79. Sources of Income?” Economics Letters 101 Pratham Organization. 2005. “Annual Status of (3): 246–48. Education Report.” Pratham Resource Center, Schultz, T. P. 2004. “School Subsidies for the Mumbai, India. Poor: Evaluating the Mexican Progresa Pov- Ravallion, M. 2009. “Evaluation in the Prac- erty Program.” Journal of Development Eco- tice of Development.” World Bank Research nomics 74 (1): 199–250. Observer 24 (1): 29–53. Schwartz, J. B., and I. Bhushan. 2004. “Improv- Regalía, F., and L. Castro. 2009. “Nicaragua: ing Equity in Immunization through a Public- Combining Demand- and Supply-Side Incen- Private Partnership in Cambodia.” Bulletin of tives.” In Performance Incentives for Global the World Health Organization 82: 661–67. Health—Potential and Pitfalls, ed. R. Eichler, Sims, C.A. 2010. “But Economics Is Not an R. Levine, and the Performance-Based Incen- Experimental Science.” Journal of Economic tives Working Group, 215–35. Washington, Perspectives 24 (2): 59–68. DC: Center for Global Development. Skoufi as, E., and J. Shapiro. 2006. “Evaluating Reinikka, R., and J. Svensson. 2005. “Fighting the Impact of Mexico’s Quality Schools Pro- Corruption to Improve Schooling: Evidence gram: The Pitfalls of Using Nonexperimental from a Newspaper Campaign in Uganda.” Data.” Policy Research Working Paper 4036, Journal of the European Economic Associa- World Bank, Washington, DC. tion 3 (2-3): 259–67. Soares, F. V., R. Ribas, and G. I. Hirata. 2008. ——— . 2006. “The Power of Information: Evi- “Achievements and Shortfalls of Conditional dence from a Newspaper Campaign to Reduce Cash Transfers: Impact Evaluation of Para- Capture of Public Funds.” World Bank, Wash- guay’s Tekopora Program.” International Pov- ington, DC; Institute for International Eco- erty Center, Brasilia, Brazil. nomic Studies, Stockholm University, Sweden. Soeters, R., C. Habineza, and P. B. Peerenboom. ———. 2010. “Working for God? Evidence from a 2006. “Performance-Based Financing and Change in Financing of Nonprofit Health Care Changing the District Health System: Expe- Providers in Uganda.” Journal of the European rience from Rwanda.” Bulletin of the World Economic Association 8 (6): 1159–78. Health Organization 84 (11): 884–89. Rivera, J. A., D. Sotres-Álvarez, J.-P. Habicht, T. Stasavage, D. 2005. “Democracy and Education Shamah, and S. Villalpando. 2004. “Impact of Spending in Africa.” American Journal of the Mexican Program for Education, Health, Political Science 49 (2): 343–58. GLOBAL MONITORING REPORT 2011 LINKING SPENDING AND OUTCOMES 101 Temperley, M., D. H. Mueller, J. Kiambo Njagi, Weinhard, L. S., M. P. Carey, B. T. Johnston, and W. Akhwale, S. E. Clarke, M. C. H. Jukes, N. L. Bickham. 1999. “Effects of HIV Coun- B. B. A. Estambale, and S. Brooker. 2008. seling and Testing on Sexual Risk Behavior: A “Costs and Cost-Effectiveness of Delivering Meta-Analytic Review of Published Research, Intermittent Preventive Treatment through 1985–1997.” American Journal of Public Schools in Western Kenya.” Malaria Journal Health 89 (9): 1152–63. 7: 196–207. WHO (World Health Organization). 2008. WHO Thornton, R. 2005. “The Demand for and Impact Recommendations for Routine Immuniza- of HIV Testing: Evidence from a Field Experi- tion. Geneva, Switzerland. http://www.who. ment.” Unpublished manuscript, Harvard Uni- int/immunization/policy/WHO_EPI_Sum_ versity, Cambridge, MA. tables_Def_200713.pdf. UNESCO (United Nations Educational, Sci- ——— . 2010. World Health Report. Health Sys- entific, and Cultural Organization). 2010. tems Financing: The Path to Universal Cover- Education for All Global Monitoring Report age. Geneva, Switzerland. 2010: Reaching the Marginalized. Oxford, World Bank. 2003. World Development Report UK: Oxford University Press. 2004: Making Services Work for Poor People. Van de Walle, D. 1995. “The Distribution of Sub- Washington, DC. sidies through Public Health Services in Indo- World Bank/AERC (African Economic Research nesia, 1978–87.” In Public Spending and the Consortium). 2010. “Service Delivery Indica- Poor: Theory and Evidence, ed. D. Van de tors: Pilot in Education and Health in Africa.” Walle and K. Neads, 226–58. Washington, World Bank, Washington, DC. DC: World Bank. Zivin J., H. Thirumurthy, and M. Goldstein. Vermeersch, C., and M. Kremer. 2004. “School 2006. “AIDS Treatment and Intrahousehold Meals, Educational Achievement, and School Resource Allocations: Children’s Nutrition and Competition: Evidence from a Randomized Schooling in Kenya.” Working Paper 12689, Evaluation.” Policy Research Working Paper National Bureau of Economic Research, Cam- 3523, World Bank, Washington, DC. bridge, MA. 4 Assisting Indigenous and Socially Excluded Populations A s we approach 2015 and come closer populations and countries on track that have to meeting the Millennium Devel- off-track population groups. It also touches opment Goals (M DGs) targets, on the impact of the modern world on small, the remaining challenges take on increased largely isolated groups, highlighting the need urgency. One of these is reaching out to the for sensitivity in dealing with them. last and most difficult groups to reach— vulnerable populations with tenuous connec- tions to the modern state and its economy. Last in heading down the home Indigenous groups tend to have deeper stretch poverty, less access to education, and worse The last groups to be helped in countries are health outcomes than the general popula- invariably in areas difficult to reach, are usu- tion. In some rapidly growing economies, ally vulnerable, and have a tenuous connec- particularly in Asia, indigenous groups have tion to the modern state and its economy. enjoyed reductions in poverty at a pace com- Coming from minority ethnic groups and parable to general society. In other countries, indigenous populations (box 4.1), they tend particularly in Latin America, slow growth to speak a language different from the major- has slowed the pace of poverty reduction for ity.1 Whereas data coverage is a perennial indigenous groups, and targeted programs challenge, MDG indicators for ethnic minor- have made only a limited contribution. There ities and indigenous peoples are (with some is an urgent need for research to improve the exceptions) worse than population averages. availability and reliability of data on human development among indigenous groups—and Poverty is higher among indigenous to determine what kinds of programs have peoples than in the general population improved their welfare and in what contexts. This chapter documents the poverty, edu- As the global development community cation, and health of vulnerable and indige- looks for ways to meet the MDG of halv- nous groups, highlighting their progress—and ing (from its 1990 level) the share of people the remaining gaps. It also reviews countries in poverty by 2015, it cannot ignore the that are off track in meeting the MDGs plight of the world’s indigenous peoples (box mainly because they have large vulnerable 4.1). They make up 4 percent of the global From helmet to turban, hats worn by women and children give distinctive touches to each group of Hmong. 103 104 A S S I S T I N G I N D I G E N O U S A N D S O C I A L LY E X C L U D E D P O P U L AT I O N S GLOBAL MONITORING REPORT 2011 BOX 4.1 Who are indigenous peoples? “Indigenous peoples” has no widely accepted defi- the national level. In India, however, infant mortal- nition. The World Bank’s position is that “because ity among the scheduled tribes is uniformly higher of the varied and changing contexts in which indig- (worse) than the national average, and water depri- enous peoples live and because there is no universally vation rates both exceed and fall short of their accepted definition of indigenous peoples, World national levels. Among the Hill Tribes in Thailand; Bank policy does not define the term. Indigenous the Kammu and Leu in the Lao People’s Demo- peoples may be referred to in different countries by cratic Republic; and the Hmong, Muong, and BaNa such terms as ‘indigenous ethnic minorities,’ ‘aborigi- in Vietnam, these rates are the worst in the region. nals,’ ‘hill tribes,’ ‘minority nationalities,’ ‘scheduled The lowest female literacy rates are found among the tribes,’ or ‘tribal groups’” (World Bank Operational Hmong in Lao PDR and Vietnam. And the existence Directive 4.10). of vulnerable groups who fare much worse than the The United Nations has not adopted a defi nition general population is hardly unique to developing but has developed a modern understanding of the countries. In Australia and New Zealand, all indi- term based on a variety of characteristics: self-iden- cators for Aborigines and Maori are worse than tification at the individual level and accepted by the national averages. community as their member; historical continuity Data coverage is far more limited in Africa, mak- with precolonial or presettler societies; a strong link ing overarching conclusions difficult. In many cases, to territories and surrounding natural resources; a dis- available data do not cover core groups widely con- tinct social, economic, or political system; a distinct sidered to be indigenous because of their small size language, culture, and belief system; individuals who (for example, the Ogiek in Kenya). The existing data form nondominant groups within society; and those show that under-five mortality rates tend to be high- who resolve to maintain and reproduce their ancestral est among West African groups (such as the Fulani environments and systems as distinctive peoples and and Tuareg) and lowest among the Masai and Ethio- communities.a pian groups. These last two groups also experience Moreover, evidence is growing that joining the highest rates of water deprivation. Education together under a common identity as indigenous indicators are uniformly worse; even in countries peoples is fairly new and has accompanied a process with higher literacy (such as Namibia), the male lit- among some groups of “reclaiming” identity. eracy rate for San males is less than half that of the With few exceptions, MDG indicators for indig- national sample; and for females, it is less than one enous groups across Asia are worse than all-popu- third. lation indicators. Under-five mortality rates for the Nepalese Janajati are distributed around the national Source: Hall and Patrinos 2011. level, but as a whole are below (that is, better than) a. UN 1981. population—nearly 300 million—but they One-third of indigenous people are poor.2 account for about 10 percent of the world’s Among the majority of countries with disag- poor; and nearly 80 percent of them are in gregated data, poverty rates exceed 50 percent Asia. (table 4.1). Although the majority of indig- Turning the situation around will require enous people come from China and India, widespread and sustainable economic growth the indigenous poor are slightly more evenly and poverty reduction, along with strategies distributed, largely owing to their strikingly to address multiple sources of disadvantage low poverty rate in China. In other countries, to reach those who need—and are willing to indigenous peoples have disproportionately accept—special assistance. high poverty rates. GLOBAL MONITORING REPORT 2011 A S S I S T I N G I N D I G E N O U S A N D S O C I A L LY E X C L U D E D P O P U L AT I O N S 105 TABLE 4.1 Poverty rates, latest year available Similar trends were seen in Bolivia, Ecuador, Poverty head count (% poor) and Peru. Indeed, for most countries in Latin Country/data year Indigenous Nonindigenous America with a sizable indigenous popula- tion, there is almost no progress in poverty Democratic Republic reduction for the indigenous population— of Congo/2005 84.8 71.7 Mexico/2008 80.6 45.3 even in countries where the nonindigenous Ecuador/2006 78.0 46.6 population is making progress. Guatemala/2006 74.8 36.2 By contrast, poverty rates, even among Gabon/2003 70.1 32.7 indigenous people, have declined sharply in Bolivia/2006 69.3 46.0 emerging Asia—notably, China, India, and Peru/2005 62.3 35.0 Vietnam/2006 52.3 10.3 Vietnam (figure 4.1). China’s progress is Lao PDR 2002 50.6 25.0 remarkable: in 1981, it had one of the world’s Brazil/2002 48.0a 23.0b highest proportions of people living in pov- India/2004 43.8 22.7 erty—84 percent. By 2005, that had fallen Chile/2006 15.2 9.1 to 16 percent. 5 Even better, China’s ethnic China/2002 5.4 3.5 minorities’ poverty rates declined faster than Source: Hall and Patrinos 2011. Note: Head count poverty rates are national. did the nonminority rates. a. Refers to whites and “black/brown” (African origin). India and Vietnam also showed significant b. Refers to whites (Telles 2007). Head count poverty rates are national. declines: although not as dramatic as China’s, overall, the declines were much more robust for India’s minorities, scheduled castes, and tribes and for Vietnam’s ethnic minorities. A sizable poverty gap remains for indig- But despite the rapid decline in poverty among enous people in global terms, ranging from indigenous people in Vietnam, their poverty small in China to significant in Vietnam. 3 rates remain much higher than those for the The gap has even been increasing over time. nonindigenous population; and the gap has Research from Latin American countries widened slightly since the early 1990s. with large indigenous populations—Bolivia, Ecuador, Guatemala, and Peru—shows a Vulnerable groups receive less sticky persistence of (nonimproving) pov- schooling and underperform in school erty rates for indigenous peoples over time. Progress used to be limited in Mexico as well, As with poverty, indigenous people are also but poverty rates there have been falling over disadvantaged in education, as in access to the last decade, especially between 2004 and schooling (table 4.2). To varying degrees in 2008.4 This may be the first period in history nearly all countries, educational progress has that the indigenous population in Mexico lagged because of indigenous, low-income, has seen an improvement. More important, gender, or disability status. Leaving such analysis in Mexico shows that the difference groups at the fringes of the education system in poverty rates between indigenous and non- will eventually impede a country’s ability to indigenous peoples has been declining mostly develop. Even in India, despite its success as a result of explained or observable factors: in rapidly reducing poverty for the whole education and access to services, the policy- population (including significantly for the driven variables, are important. scheduled tribes, or Adivasi [literally “origi- Chile has also seen reductions in indige- nal inhabitants”]), this minority population nous poverty in recent years. But other coun- accounts for a fourth of the poorest wealth tries in Latin America have not been so suc- decile. Education indicators tell a similar cessful. Guatemala reduced poverty rapidly story, with improvements but large and per- until recently; but even when poverty declined sistent differences. Scheduled tribe children for the population as a whole, progress lag far behind when it comes to educational among indigenous people was much slower. attainment above the primary level.6 106 A S S I S T I N G I N D I G E N O U S A N D S O C I A L LY E X C L U D E D P O P U L AT I O N S GLOBAL MONITORING REPORT 2011 FIGURE 4.1 Progress is mixed in reducing the poverty of indigenous groups poverty rate over time a. Mexico b. Guatemala 100 100 90 90 80 80 70 percent percent 70 60 60 50 50 40 40 30 30 20 1992 1994 1996 1998 2000 2002 2004 2006 2008 1992 1994 1996 1998 2000 2002 2004 2006 indigenous nonindigenous indigenous nonindigenous c. Ecuador d. Peru 95 75 70 85 65 75 60 55 percent percent 65 50 45 55 40 45 35 30 35 25 1994 1996 1998 2000 2002 2004 2006 1994 1996 1998 2000 2002 2004 2005 indigenous nonindigenous indigenous nonindigenous e. Chile f. Bolivia 25 80 75 20 70 percent percent 65 15 60 55 10 50 45 5 40 1996 1998 2000 2002 2004 2006 1997 1999 2001 2003 2005 2007 indigenous nonindigenous indigenous nonindigenous GLOBAL MONITORING REPORT 2011 A S S I S T I N G I N D I G E N O U S A N D S O C I A L LY E X C L U D E D P O P U L AT I O N S 107 FIGURE 4.1 (continued) g. India h. Vietnam 70 100 65 90 60 80 55 70 50 60 percent percent 45 50 40 40 35 30 30 20 25 10 20 0 1992 1994 1996 1998 2000 2002 2004 1993 1995 1997 1999 2001 2003 2005 2006 scheduled caste others scheduled tribes ethnic minority ethnic majority i. China 14 12 10 percent 8 6 4 2 0 1998 1999 2000 2001 2002 minority Han Source: Hall and Patrinos 2011. TABLE 4.2 Years of schooling, latest year teaching, vulnerable groups are likely to available experience discrimination that manifests in Country Indigenous Nonindigenous different ways (including self-perceptions China/2002 8.2 8.9 and stigma). Peru/2001 6.4 8.7 Stigma is sometimes internalized, as an Mexico/2008 6.3 8.6 experimental investigation into the impact of Bolivia/2002 5.9 9.6 caste on test scores in India demonstrated.7 Brazil/2002 4.6a 6.6b Children ages 11 and 12 were chosen at ran- Ecuador/2006 4.3 6.9 Guatemala/2006 2.5 5.7 dom from a low caste and three high castes and were given a series of puzzles to solve. Source: Hall and Patrinos 2011. Note: See box 1.5 for more information about Brazil. When caste was not announced to the chil- a. Refers to whites and “black/brown” (African origin). dren, it had no bearing on the initial score b. Refers to whites (Telles 2007). or on the improvement in score registered in subsequent test rounds. But when caste was Not only do vulnerable groups receive announced before the test, the scores for low- less schooling; they also tend to achieve caste children fell dramatically, by 14–39 less. In addition to poor starting condi- percent (figure 4.2). These fi ndings confi rm tions, insufficient inputs, and ineffective how much social identities are a product of 108 A S S I S T I N G I N D I G E N O U S A N D S O C I A L LY E X C L U D E D P O P U L AT I O N GLOBAL MONITORING REPORT 2011 FIGURE 4.2 Test results show how stigma can be fractionalization on the quality of institutions internalized and economic growth. The data show just how heterogeneous many developing coun- 4 tries are, but they do not show that a high degree of heterogeneity produces negative 3 outcomes. Fractionalization has been shown to slow the progress on many of the MDGs.9 test scores 2 Although fractionalization is not correlated with schooling progress, there is an indica- tion of a negative effect of heterogeneity on 1 test scores (figure 4.3). Both the global and Indian examples 0 caste not announced caste announced underscore the need to be careful in develop- high caste low caste ing appropriate learning materials and envi- ronments for minorities. Although some of the self-perceptions and stigma will take time Source: Hoff and Pandey 2004. to address, these more immediate lessons for promoting rights and accountabilities can be used now. history, culture, and personal experience, creating pronounced educational disadvan- Indigenous peoples live shorter lives tages through their effects on individual everywhere expectations. Global evidence also shows how identity The World Health Report 2010 states, can limit development. A fractionalization “income is not the only factor influencing data set compiled by Alesina and colleagues [health] service coverage. In many settings, measures the degree of ethnic, linguistic, migrants, ethnic minorities and indigenous and religious heterogeneity in various coun- people use [health] services less than other tries. 8 It can be used to test the effects of population groups, even though their needs may be greater.”10 Indigenous peoples therefore deserve spe- FIGURE 4.3 Heterogeneity has a negative effect on test scores cial attention. They live shorter lives and are in worse health than their nonindigenous 60 compatriots almost everywhere. In seven Latin American countries, the proportion of 55 indigenous women receiving antenatal care or giving birth at health facilities was much 50 lower than for nonindigenous women.11 Such inequality is one of the causes of the dispar- test scores 45 ity in maternal health outcomes between the two populations. African-American 40 women in the same countries also gave birth at health facilities less frequently and had 35 poorer maternal health outcomes than other women. 30 Such health inequalities are not unique to 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 low-income countries. Australia, Canada, ethnolinguistic fractionalization index and New Zealand show variations in access to health services between the two popu- Source: World Bank staff calculations using data from Alesina et al. 2003 and Schlotter 2010. lations, frequently linked to distance and GLOBAL MONITORING REPORT 2011 A S S I S T I N G I N D I G E N O U S A N D S O C I A L LY E X C L U D E D P O P U L AT I O N 109 MAP 4.1 Almost nine million children still die each year before they reach their fifth birthday IBRD 38449 APRIL 2011 Child mortality Greenland (Den) Faeroe Norway Iceland Islands Finland (Den) Sweden The Netherlands Russian Federation Estonia Canada Isle of Man (UK) Denmark Russian Latvia Fed. Lithuania United Ireland Kingdom Germany Poland Belarus Channel Islands (UK) Belgium Ukraine Luxembourg Moldova Kazakhstan Liechtenstein Romania Mongolia Switzerland France Italy Andorra Bulgaria Georgia Azer- Uzbekistan Kyrgyz D.P.R. Turkey Armenia baijan Turkmenistan Rep. United States Portugal Spain Monaco Tajikistan of Korea Japan Greece Syrian Gibraltar (UK) Cyprus Rep. of Tunisia Malta Lebanon Arab Islamic Rep. Afghanistan China Korea Israel Rep. Iraq of Iran Bermuda Morocco Kuwait (UK) West Bank and Gaza Jordan Bahrain Pakistan Nepal Bhutan Former Algeria Libya Arab Rep. Qatar The Bahamas Spanish of Egypt Saudi Cayman Is. (UK) Sahara Arabia United Arab Bangladesh Cuba Turks and Caicos Islands (UK) India Mexico Emirates Myanmar Lao Haiti Mauritania Oman P.D.R. Belize Jamaica Cape Verde Mali N. Mariana Islands (US) Niger Rep. of Thailand Vietnam Guatemala Honduras Senegal Chad Eritrea Yemen Guam (US) El Salvador Nicaragua The Gambia Burkina Sudan Marshall Faso Djibouti Cambodia Philippines Islands Guinea-Bissau Guinea Benin Federated States of Costa Rica Panama Ghana Nigeria Ethiopia Sri Micronesia R.B. de Guyana Sierra Leone Côte Central Lanka Venezuela Suriname Liberia D’Ivoire African Rep. Brunei Cameroon Somalia Palau Colombia French Guiana (Fr) Togo Malaysia Equatorial Guinea Uganda Maldives Ecuador São Tomé and Príncipe Gabon Congo Rwanda Kenya Singapore Nauru Dem. Rep. of Burundi Seychelles Papua Solomon Congo Indonesia New Guinea Islands Tanzania Comoros Peru Brazil Timor-Leste Angola Zambia Malawi Mayotte (Fr) Vanuatu Bolivia Zimbabwe Mauritius Namibia Madagascar Paraguay Botswana Mozambique Réunion (Fr) New Caledonia Australia (Fr) Swaziland South Lesotho Africa Chile Uruguay Argentina New Dominican Puerto British Virgin Poland Child mortality Zealand Germany Republic Rico (US) Islands (UK) Ukraine Anguilla (UK) Czech Rep. Slovak Rep. Mortality rate, under-five, per 1,000 U.S. Virgin Islands Antigua and Barbuda (US) St. Maarten (Neth) St. Kitts Guadeloupe (Fr) Austria Hungary ≥ 100 and Nevis Dominica Slovenia Romania 50–99.9 Curaçao Montserrat (UK) Martinique (Fr) Croatia (Neth) St. Lucia Bosnia and 30–49.9 Aruba Bonaire St. Vincent and San Herz. Serbia Bulgaria (Neth) (Neth) The Grenadines Barbados Marino 15–29.9 Montenegro Kosovo Grenada FYR ≤14.9 Italy Macedonia Trinidad Vatican Albania R.B. de Venezuela and Tobago City Greece no data MDG performance Greenland (Den) Faeroe Norway Iceland Islands Finland (Den) Sweden The Netherlands Russian Federation Estonia Canada Isle of Man (UK) Denmark Russian Latvia Fed. Lithuania United Ireland Kingdom Germany Poland Belarus Channel Islands (UK) Belgium Ukraine Luxembourg Moldova Kazakhstan Liechtenstein Romania Mongolia Switzerland France Italy Andorra Bulgaria Georgia Azer- Uzbekistan Kyrgyz D.P.R. Turkey Armenia baijan Turkmenistan Rep. United States Portugal Spain Monaco Tajikistan of Korea Japan Greece Syrian Gibraltar (UK) Cyprus Rep. of Tunisia Malta Lebanon Arab Islamic Rep. Afghanistan China Korea Israel Rep. Iraq of Iran Bermuda Morocco Kuwait (UK) West Bank and Gaza Jordan Pakistan Bhutan Bahrain Nepal Former Algeria Libya Arab Rep. Qatar The Bahamas Spanish of Egypt Saudi Cayman Is. (UK) Sahara Arabia United Arab Bangladesh Cuba Turks and Caicos Islands (UK) India Mexico Emirates Myanmar Lao Haiti Mauritania Oman P.D.R. Belize Jamaica Cape Verde Mali N. Mariana Islands (US) Niger Rep. of Thailand Vietnam Guatemala Honduras Senegal Chad Eritrea Yemen Guam (US) El Salvador Nicaragua The Gambia Burkina Sudan Marshall Faso Djibouti Cambodia Philippines Islands Guinea-Bissau Guinea Benin Federated States of Costa Rica Panama Sri Micronesia R.B. de Guyana Sierra Leone Côte Ghana Nigeria Central Ethiopia Lanka Venezuela Suriname Liberia D’Ivoire African Rep. Brunei Cameroon Somalia Palau Colombia French Guiana (Fr) Togo Malaysia Equatorial Guinea Uganda Maldives Ecuador São Tomé and Príncipe Gabon Congo Rwanda Kenya Singapore Nauru Dem. Rep. of Burundi Seychelles Papua Solomon Congo Indonesia New Guinea Islands Tanzania Comoros Peru Brazil Timor-Leste Angola Malawi Mayotte Zambia (Fr) Vanuatu Bolivia Zimbabwe Mauritius Namibia Madagascar Paraguay Botswana Mozambique Réunion (Fr) New Caledonia Australia (Fr) Swaziland South Lesotho Africa Chile Uruguay Argentina MDG performance New Dominican Puerto British Virgin Poland MDG 4.A Reduce by two-thirds, Zealand Germany Republic Rico (US) Islands (UK) between 1990 and 2015, the under-five Ukraine Anguilla (UK) Czech Rep. U.S. Virgin Islands Antigua and Barbuda Slovak Rep. mortality rate (US) Austria St. Maarten (Neth) St. Kitts Guadeloupe (Fr) Hungary and Nevis Slovenia on target Dominica Romania Curaçao Montserrat (UK) Martinique (Fr) Croatia close to target (Neth) Bosnia and Bonaire St. Vincent and St. Lucia San Herz. Serbia Aruba far from target Bulgaria (Neth) (Neth) The Grenadines Barbados Marino This map was produced by the Map Design Unit of The World Bank. Montenegro Kosovo Grenada Italy FYR no data The boundaries, colors, denominations and any other information shown on this map do not imply, on the part of The World Bank Trinidad Macedonia and Tobago Vatican Albania Group, any judgment on the legal status of any territory, or any R.B. de Venezuela City Greece endorsement or acceptance of such boundaries. Source: World Bank staff calculations based on data from the World Development Indicators database. 110 A S S I S T I N G I N D I G E N O U S A N D S O C I A L LY E X C L U D E D P O P U L AT I O N S GLOBAL MONITORING REPORT 2011 FIGURE 4.4 Both poor and rich countries show transportation costs. Some rich countries also gaps in life expectancy show gaps in life expectancy (figure 4.4). Further evidence of insufficient attention Australia to the needs of minority groups is seen in Europe, where the largest and most disadvan- Nepal taged group, Roma, lags far behind the conti- nent’s majority population (box 4.2). Guatemala New Zealand Gender is a key dimension Panama Gender is a key dimension in the social indi- Canada cators of indigenous or otherwise excluded groups. Although both boys and girls in these Mexico groups are significantly underrepresented in 0 2 4 6 8 10 12 14 16 18 20 school and lag on other important indicators, girls are particularly disadvantaged. Disturb- years ingly, out-of-school girls from indigenous populations make up a large portion of total Source: UN 2009, p. 159. Note: The figure shows the difference in years of life expectancy between population of out-of-school girls in develop- indigenous and nonindigenous groups in each country. ing countries.12 BOX 4.2 The Roma of Europe The Roma, or “Gypsies,” are a Romani subgroup of European governments and others have made a about 7 million to 9 million people living primarily political commitment to improve Romani socioeco- in Central and Eastern Europe. They are poorer than nomic status and social inclusion under the aegis of the general population: in some countries, Roma pov- the Decade of Roma Inclusion 2005–15. The initia- erty rates are more than 10 times those of non-Roma.a tive brings together governments and intergovern- And they are more likely to fall into poverty and stay mental and nongovernmental organizations as well as poor. They also form a large proportion of youth and Romani civil society to help ensure progress toward the potential labor force in these countries. But their improving the welfare of Roma and to review such low level of education means that European Union progress transparently and quantifiably. It commits economies are losing hundreds of millions of euros governments to focus on education, employment, annually in production and fiscal contributions. health, and housing while taking into account pov- Lower-bound annual estimates of productivity erty, discrimination, and gender bias. losses for Bulgaria, the Czech Republic, Romania, and The Open Society Foundation documents the prog- Serbia total more than €2 billion, and there are similar ress that countries are making at the halfway point estimates of €860 million in fiscal losses.b Using other for the 4.5 million Roma in participating countries.d Roma population estimates, the economic losses for Unfortunately, a lack of data remains the biggest the four countries combined amount to €5.7 billion obstacle in any assessment, as shown in the table on annually, and the fiscal losses are €2 billion annually.c the next page. But even these patchy data allow one to The annual fiscal gains from bridging the employment see, for example, that the infant mortality rate among gap are much larger than the total cost of investing in Roma is two or three times as high as that among the public education for all Roma children—by factors of general population. 7.7 in Bulgaria, 7.4 in the Czech Republic, 3.3 in Ser- bia, and 2.4 in Romania. GLOBAL MONITORING REPORT 2011 A S S I S T I N G I N D I G E N O U S A N D S O C I A L LY E X C L U D E D P O P U L AT I O N S 111 BOX 4.2 The Roma of Europe (continued) Roma progress on selected indicators Primary education Infant mortality rate completion rate (per 1,000 live births) Country Year Poverty (%)a General (%) Among Roma (%) General Among Roma 2005 Insufficient data 14.0 16.0 No data Albania Most recent 78 Insufficient data Insufficient data 13.0 No data Insufficient Bosnia and 2005 Insufficient data No data data No data Herzegovina Most recent 27 Insufficient data No data 13.0 No data 2005 Insufficient data 28.3 10.4 25.0 Bulgaria Most recent 46 Insufficient data 31.6 8.6 25.0 2005 95.0 No data 5.7 25.1 Croatia Most recent 10 101.0 No data 4.5 11.5 2005 102.0 No data 3.4 No data Czech Republic Most recent 45 93.2 No data 2.7 No data 2005 95.0 76.4 6.2 No data Hungary Most recent 9 96.7 No newer data 5.6 No data 2005 96.0 50.8 12.8 Insufficient data Macedonia, FYR Most recent 33 92.0 No data 10.8 Insufficient data 2005 91.1 9.2 9.5 No data Montenegro Most recent 27 No newer data 20.0 7.5 No data 2005 No data 31.8 15.0 Insufficient data Romania Most recent 66 No data 19.8 11.0 No data 2005 95.0 22.7 8.0 25.0 Serbia Most recent 57 99.5 No data 6.7 No newer data 2005 94.0 No data 7.2 No data Slovak Republic Most recent n.a. No newer data No data 5.9 No data Source: OSI 2010; poverty estimates from the World Bank. Note: n.a. = not available. a. Percent of Roma living in households below $4.30 per day (purchasing power parity), expenditure based. What works for Roma inclusion Such interventions can be for supply (such as explicit desegregation efforts, teacher training, and school Increasing the number of high school graduates grants) and for demand (such as Roma mediators and would give the current generation of young Roma a conditional cash transfer programs). better chance in the labor market and would improve school readiness to boost the chances of the next gen- eration entering the classroom with the same skills a. Ringold, Orenstein, and Wilkens 2005. as non-Roma. International experience suggests that b. De Laat 2010; World Bank 2010. investments in early childhood development and c. De Laat 2010; World Bank 2010. improvements in school attendance and completion d. OSI 2010. are the most promising interventions to break the e. See, for example, Heckman (2006); Kremer, Miguel, and intergenerational transmission of social exclusion.e Thornton (2009); and Patrinos (2007). 112 A S S I S T I N G I N D I G E N O U S A N D S O C I A L LY E X C L U D E D P O P U L AT I O N S GLOBAL MONITORING REPORT 2011 Sub-Saharan Africa has about 23 million populations, although their presence may not girls out of school, 75 percent from excluded be the only reason for this shortfall. Even groups; South Asia has 23 million and 67 some on-track countries have these popula- percent; the Middle East and North Africa tions; thus, when data are disaggregated, have almost 5 million and 90 percent; East- they are off track in the sense that although ern Europe and Central Asia have 1.5 million the country may reach all (or most) of the and 90 percent (including rural populations in MDGs, but has significant segments of the Turkey); and Latin America has 1.5 million— population that are off track (or seriously off almost all from excluded groups. In aggre- track). Indigenous people get little attention gate, 71 percent of out-of-school girls in devel- in country reports designed to monitor prog- oping countries are from excluded groups.13 ress in meeting the goals (box 4.3). In India, tribal women fare the worst: even among younger age cohorts (ages 15–21 in Off-track countries—minority groups 2005), they attain an average of just four years are lagging severely of education, three years less than nontribal women.14 In Lao PDR, non-Lao-Tai women Many of the off-track countries are in receive significantly less education than non- Africa.16 And some of them are off track Lao-Tai men or Lao-Tai of both sexes. For because of the severe lack of progress of their example, 34 percent of rural non-Lao-Tai minority groups. Nigeria, for example, has girls had never attended school in 2002–03, three major tribes—the Igbo, Hausa, and compared with 6 percent of Lao-Tai girls.15 Yoruba—more than 200 ethnic groups, and more than 500 indigenous languages and dialects. It would be important to use disag- Indigenous groups and the gregated data on these subgroups to investi- MDGS gate the link between minority groups and Some countries that are off track to reach one the failure to achieve the MDGs—not just for or more of the MDGs have large indigenous Nigeria, but more generally. BOX 4.3 Vulnerable groups deserve more attention in MDG country reports A review of 50 MDG country reports found that eth- Attention to indigenous peoples, in general, is nic or linguistic minorities are mentioned in only 19, significantly higher than attention to nonindigenous and the inequalities experienced by religious minori- marginalized minorities. Although attention to gen- ties are mentioned in only 2. An additional 10 reports der issues in many MDG country reports is positive, mention only indigenous peoples, without identifying there is very little consideration of discrimination any other minority groups. experienced by minority women or of targeted poli- The mentions of minorities vary widely, with some cies for them. reports providing a good range of information and Minorities are virtually absent from the MDG disaggregated data under several MDGs. In other country reports from donor countries. None of those cases, minorities are mentioned only in the back- reports considers minorities under any of the eight ground section describing the national population goals, and only the reports for Nepal and Vietnam and without particular attention to their situation in give sufficient attention to indigenous peoples. Nor do relation to the MDGs. Minorities are mentioned most any reports indicate any consultation with indigenous frequently in connection with Goal 2 on universal peoples or consistently provide disaggregated data for primary education. them. Sources: UN 2010a; Hartley 2008. GLOBAL MONITORING REPORT 2011 A S S I S T I N G I N D I G E N O U S A N D S O C I A L LY E X C L U D E D P O P U L AT I O N S 113 TABLE 4.3 Status of indigenous groups relative to the MDGs, selected African countries Nutrition Net primary Under-five mortality Water deprivation deprivation Literacy rate enrollment (% of (% of (% of (% of Households in Country/data year/group (per 1,000 live births) (% of households) (% of children) males) females) boys) girls) sample (n) Cameroon/2004 148 29 13 80 65 81 78 10,462 Peulh 170 21 11 43 19 60 55 387 Pygmy 160 54 54 17 10 65 100 18 Fulfulde 192 26 18 27 13 57 49 1,632 Niger/2006 218 28 29 28 12 44 29 7,660 Peul 204 38 33 5 5 17 19 350 Tuareg 204 52 30 17 5 30 29 635 Mali/2006 215 6 16 37 17 46 39 12,998 Peulh 229 4 16 31 16 38 35 1,400 Tanachek 194 20 17 28 11 41 28 815 Guinea/2005 188 59 15 44 16 58 49 6,282 Peulh 171 61 12 40 14 56 45 2,228 Senegal/2005 135 11 5 54 35 58 58 7,407 Poular 137 13 8 44 29 56 56 1,990 Source: Computed from demographic and health surveys in MacDonald forthcoming. Demographic and health surveys show the lowest index in Mexico, the Metlatonoc, is the frequent considerable disadvantage in at about the same level as Angola and Malawi. belonging to an indigenous group in some Perhaps it is not surprising that 98 percent of low-income African countries (table 4.3). In the population there is indigenous.17 Cameroon, the Peulh and the Fulfulde (a large India is also on track to meet the MDGs, minority) are doing poorly on adult literacy but the scheduled tribes are largely off track. and net primary enrollment. Niger’s Peul Although educational enrollment rates are and Tuareg face inequalities, even though the fairly high for all groups, scheduled tribes national population is far off track. Mali is suffer from high under-five mortality, poor nationally so far off track that there is seeming access to water and sanitation, nutrition equality, with all groups performing poorly. deprivation, and low levels of adult literacy.18 Ecuador’s under-five mortality is almost twice as high among indigenous people, at Groups left behind in countries 138 per 1,000 live births versus 77 per 1,000 otherwise performing well for nonindigenous groups.19 Vietnam, a star Indigenous people in on-track middle-income performer in economic growth and poverty countries are also being left behind. A few reduction, does very well on the MDGs over- examples will suffice. Mexico, a middle- all and its ethnic minorities are generally income country and an Organisation for Eco- progressing, although some are seriously off nomic Co-operation and Development mem- track for adult literacy.20 Although indigenous ber, is on track to meet most of the MDGs. But groups in Vietnam have seen rapid declines in under-five mortality for indigenous groups is poverty, their poverty rates remain some 40 at 52 per 1,000 live births—almost twice the percentage points higher than those for the 27 for nonindigenous groups. In addition, the rest of the population. poorest municipalities, heavily populated by indigenous peoples, have poor indicators for poverty and access to education and health. Lessons from progress—and The worst-performing municipalities have a from stagnation Human Development Index similar to those in As we approach 2015, the daunting challenge poorer African countries: the community with of reaching the last of the excluded groups 114 A S S I S T I N G I N D I G E N O U S A N D S O C I A L LY E X C L U D E D P O P U L AT I O N GLOBAL MONITORING REPORT 2011 becomes apparent. They have a tenuous con- direct redistributive interventions have not nection to the modern economy, likely speak been prominent in China’s efforts to reduce a language different from the majority, and poverty.23 typically are an ethnic minority. Where such India’s strong growth—focused since the groups make up a sizable proportion of the 1980s on moving from a state-controlled, population, the solution requires rapid, inward-looking economy to an outward- broad-based economic growth and poverty oriented, market-led economy—has not been reduction. But even that may not be enough damped by its poorly performing targeted to address multiple sources of disadvantage programs.24 Similarly, Vietnam’s remarkable from centuries of political and social exclu- gains in economic growth and poverty reduc- sion. Instead, a direct effort at empowerment tion were the product of market reforms and a is required, using targeted approaches and loosening of state control.25 It was only after enforcing political and social rights. growth was established that policy makers turned their attention to poverty reduction strategies and ethnic minority development;26 Growth is essential for poverty and, even then, such programs focused on reduction developing local economic activities rather In some Asian countries, rapid poverty reduc- than targeting the poor, as in Latin America. tion and the relative gains of minority groups Latin American countries with large indig- were generated by rapid growth and fairly enous populations also have had some suc- low inequality in access to productive inputs cess in equalizing human capital and expand- such as land and human capital.21 ing access to basic services, starting from From the 1970s’ reforms to 1985, China’s very unequal conditions. Indigenous people main strategy to reduce poverty was to focus often have political representation, and their on rural economic growth. After 1985, how- rights are generally well established. All Latin ever, the government, recognizing that growth American countries but one have national was necessary but insufficient to reduce pov- bilingual education programs, 27 some of erty, adopted a development-oriented strat- which may benefit indigenous populations egy, providing assistance through various (box 4.4). And 17 have conditional cash programs in poor regions. 22 Nonetheless, transfer (CCT) programs.28 Despite these efforts, slow growth has hampered progress in overall poverty reduc- BOX 4.4 Bilingual education in Guatemala tion and more equitable income distribu- tion. 29 The Latin American countries with the largest indigenous populations expe- Bilingual education is an efficient public investment. rienced per capita growth rates of no more According to a crude cost–benefit exercise, a shift to than 1 percent in the 1990s and not much bilingual schooling for indigenous people in Guate- more since then. High inequality, slow mala would result in considerable cost savings because growth, and little participation by indigenous of reduced grade repetition. The higher quality of edu- peoples in the economic progress that did cation would probably help students complete the pri- occur meant that indigenous peoples made mary education cycle and substantially increase total little progress in poverty reduction. Most education levels at lower cost. distributive programs, as currently designed The cost savings from bilingual education are esti- mated at $5 million, equal to the cost of providing pri- and implemented, have not reduced poverty mary schooling to 100,000 students a year. A reduction among indigenous people. One significant in the number of dropouts and the effect on personal exception is Mexico’s CCT program Opor- earnings would thus be significant.a tunidades, an important part of the country’s antipoverty program that has benefited indig- a. Patrinos and Velez 2009. enous people disproportionately (box 4.5). It is characterized as well targeted (to the poor, GLOBAL MONITORING REPORT 2011 A S S I S T I N G I N D I G E N O U S A N D S O C I A L LY E X C L U D E D P O P U L AT I O N 115 MAP 4.2 For girls in some African countries, secondary education remains elusive IBRD 38448 APRIL 2011 Gender parity (secondary) Greenland (Den) Faeroe Norway Iceland Islands Finland (Den) Sweden The Netherlands Russian Federation Estonia Canada Isle of Man (UK) Denmark Russian Latvia Fed. Lithuania United Ireland Kingdom Germany Poland Belarus Channel Islands (UK) Belgium Ukraine Luxembourg Moldova Kazakhstan Liechtenstein Romania Mongolia Switzerland France Italy Andorra Bulgaria Georgia Azer- Uzbekistan Kyrgyz D.P.R. Turkey Armenia baijan Turkmenistan Rep. United States Portugal Spain Monaco Tajikistan of Korea Japan Greece Syrian Gibraltar (UK) Cyprus Rep. of Tunisia Malta Lebanon Arab Islamic Rep. Afghanistan China Korea Israel Rep. Iraq of Iran Bermuda Morocco Kuwait (UK) West Bank and Gaza Jordan Bahrain Pakistan Nepal Bhutan Former Algeria Libya Arab Rep. Qatar The Bahamas Spanish of Egypt Saudi Cayman Is. (UK) Sahara Arabia United Arab Bangladesh Cuba Turks and Caicos Islands (UK) India Mexico Emirates Myanmar Lao Haiti Mauritania Oman P.D.R. Belize Jamaica Cape Verde Mali N. Mariana Islands (US) Niger Rep. of Thailand Vietnam Guatemala Honduras Senegal Chad Eritrea Yemen Guam (US) El Salvador Nicaragua The Gambia Burkina Sudan Marshall Faso Djibouti Cambodia Philippines Islands Guinea-Bissau Guinea Benin Federated States of Costa Rica Panama Ghana Nigeria Ethiopia Sri Micronesia R.B. de Guyana Sierra Leone Côte Central Lanka Venezuela Suriname Liberia D’Ivoire African Rep. Brunei Cameroon Somalia Palau Colombia French Guiana (Fr) Togo Malaysia Equatorial Guinea Uganda Maldives Ecuador São Tomé and Príncipe Gabon Congo Rwanda Kenya Singapore Nauru Dem. Rep. of Burundi Seychelles Papua Solomon Congo Indonesia New Guinea Islands Tanzania Comoros Peru Brazil Timor-Leste Angola Zambia Malawi Mayotte (Fr) Vanuatu Bolivia Zimbabwe Mauritius Namibia Madagascar Paraguay Botswana Mozambique Réunion (Fr) New Caledonia Australia (Fr) Swaziland South Lesotho Africa Chile Uruguay Argentina Gender parity (secondary) New Dominican Puerto British Virgin Poland Percentage of girls to boys enrolled at Zealand Germany Republic Rico (US) Islands (UK) Ukraine Anguilla (UK) Czech Rep. secondary level in public and private schools U.S. Virgin Islands Antigua and Barbuda Slovak Rep. (US) St. Maarten (Neth) St. Kitts Guadeloupe (Fr) Austria Hungary <79.9 and Nevis Dominica Slovenia Romania 80–89.9 Curaçao Montserrat (UK) Martinique (Fr) Croatia (Neth) St. Lucia Bosnia and 90–94.9 Aruba Bonaire St. Vincent and San Herz. Serbia Bulgaria (Neth) (Neth) The Grenadines Barbados Marino 95–99.9 Montenegro Kosovo Grenada FYR ≥100 Italy Macedonia Trinidad Vatican Albania R.B. de Venezuela and Tobago City Greece no data MDG performance Greenland (Den) Faeroe Norway Iceland Islands Finland (Den) Sweden The Netherlands Russian Federation Estonia Canada Isle of Man (UK) Denmark Russian Latvia Fed. Lithuania United Ireland Kingdom Germany Poland Belarus Channel Islands (UK) Belgium Ukraine Luxembourg Moldova Kazakhstan Liechtenstein Romania Mongolia Switzerland France Italy Andorra Bulgaria Georgia Azer- Uzbekistan Kyrgyz D.P.R. Turkey Armenia baijan Turkmenistan Rep. United States Portugal Spain Monaco Tajikistan of Korea Japan Greece Syrian Gibraltar (UK) Cyprus Rep. of Tunisia Malta Lebanon Arab Islamic Rep. Afghanistan China Korea Israel Rep. Iraq of Iran Bermuda Morocco Kuwait (UK) West Bank and Gaza Jordan Pakistan Bhutan Bahrain Nepal Former Algeria Libya Arab Rep. Qatar The Bahamas Spanish of Egypt Saudi Cayman Is. (UK) Sahara Arabia United Arab Bangladesh Cuba Turks and Caicos Islands (UK) India Mexico Emirates Myanmar Lao Haiti Mauritania Oman P.D.R. Belize Jamaica Cape Verde Mali N. Mariana Islands (US) Niger Rep. of Thailand Vietnam Guatemala Honduras Senegal Chad Eritrea Yemen Guam (US) El Salvador Nicaragua The Gambia Burkina Sudan Marshall Faso Djibouti Cambodia Philippines Islands Guinea-Bissau Guinea Benin Federated States of Costa Rica Panama Sri Micronesia R.B. de Guyana Sierra Leone Côte Ghana Nigeria Central Ethiopia Lanka Venezuela Suriname Liberia D’Ivoire African Rep. Brunei Cameroon Somalia Palau Colombia French Guiana (Fr) Togo Malaysia Equatorial Guinea Uganda Maldives Ecuador São Tomé and Príncipe Gabon Congo Rwanda Kenya Singapore Nauru Dem. Rep. of Burundi Seychelles Papua Solomon Congo Indonesia New Guinea Islands Tanzania Comoros Peru Brazil Timor-Leste Angola Malawi Mayotte Zambia (Fr) Vanuatu Bolivia Zimbabwe Mauritius Namibia Madagascar Paraguay Botswana Mozambique Réunion (Fr) New Caledonia Australia (Fr) Swaziland South Lesotho Africa Chile Uruguay Argentina MDG performance New Dominican Puerto British Virgin Poland MDG 3.A Eliminate gender disparity in Zealand Germany Republic Rico (US) Islands (UK) Secondary education, preferably by 2005, Ukraine Anguilla (UK) Czech Rep. U.S. Virgin Islands Antigua and Barbuda Slovak Rep. and at all levels of education no later than 2015 (US) Austria St. Maarten (Neth) St. Kitts Guadeloupe (Fr) Hungary and Nevis Slovenia on target Dominica Romania Curaçao Montserrat (UK) Martinique (Fr) Croatia close to target (Neth) Bosnia and Bonaire St. Vincent and St. Lucia San Herz. Serbia Aruba far from target Bulgaria (Neth) (Neth) The Grenadines Barbados Marino This map was produced by the Map Design Unit of The World Bank. Montenegro Kosovo Grenada Italy FYR no data The boundaries, colors, denominations and any other information shown on this map do not imply, on the part of The World Bank Trinidad Macedonia and Tobago Vatican Albania Group, any judgment on the legal status of any territory, or any R.B. de Venezuela City Greece endorsement or acceptance of such boundaries. Source: World Bank staff calculations based on data from the World Development Indicators database. 116 A S S I S T I N G I N D I G E N O U S A N D S O C I A L LY E X C L U D E D P O P U L AT I O N S GLOBAL MONITORING REPORT 2011 BOX 4.5 Examples of successful targeted programs CCT programs have become very popular for attack- average years of schooling, 8- to 12-year-olds ing poverty. They provide assistance conditioned on 4 the beneficiary’s actions—that is, the government provides money only to beneficiaries who fulfi ll cer- 3 tain criteria, such as enrolling children in school, ensuring that they attend regularly and complete 2 grades, and receiving regular medical attention. CCTs have become the largest social assistance program in 1 some countries, covering millions of households—for example, in Brazil (Bolsa Família) and Mexico.a 0 The program in Mexico—PROGRESA (now indigenous bilingual Spanish Oportunidades)—started early (1997) and evolved (monolingual) (monolingual) thoughtfully. The program produced successive 1997 1999 waves of data to evaluate its impact and put the data Sources: Bando, Lopez-Calva, and Patrinos 2005; Godoy et al. 2005, 2006. in the public domain. Hundreds of studies attest to its success in confronting poverty and disadvantage. Although targeted generally to the poor, the program recorded success in reaching indigenous populations.b has improved indigenous children’s schooling attain- Cambodia’s Education Sector Support Project has a ment relative to Spanish-speaking or bilingual chil- CCT that targets ethnic minorities. Exported from dren (see figure at right). Before the program, indig- the developing world to the developed, CCTs have enous children had lower school attainment than did also been used in New York City and for indigenous Spanish-speaking or bilingual children. After three communities in Australia. years of implementation, school attainment among indigenous children increased, reducing the gap. a. Fiszbein and Schady 2009. Ecuador’s CCT—Bono de Desarrollo Humano—also b. Ibid. among whom the indigenous in Mexico are Empowering indigenous people can overrepresented), transparent, and evidence improve their development prospects based—with the results of impact evaluations To properly empower vulnerable groups so used to adjust and improve it. that broad-based growth and targeted pov- The differing experiences in China and Latin America underline the importance of erty reduction programs work in their favor, growth for poverty reduction. 30 Consider- the groups need greater political representa- able reductions in poverty can be achieved tion and greater voice. Political representation through rapid growth, at least at first. But in legislatures or parliaments, for instance, given the typical widening of inequality with could significantly increase their say in the rapid development (as in China and India), design of programs and delivery of services. there is a desperate need to design effective It could also ensure that they are referred to programs that target underprivileged groups. in plans to achieve the MDGs and that they This is one area where Asia can learn from are consulted in the development and design the successes of such programs in Latin of national programs. America (including Oportunidades and a Although beneficial, effecting change similar program in Brazil)31 that combine through the political process takes time, as careful identification of beneficiary groups, made evident in Latin America. Empowering transparent program delivery, and learning stakeholders directly—for example, enabling from experience. parents to influence school policies or patients GLOBAL MONITORING REPORT 2011 A S S I S T I N G I N D I G E N O U S A N D S O C I A L LY E X C L U D E D P O P U L AT I O N S 117 to choose health providers—holds more those who are least exposed to a modern immediate promise. In Mexico, parents are national economy and allow them to main- given the right to exercise voice and decision- tain their identity and choice of livelihood. making authority over the use of small sums The prime consideration of such programs is of money for improving schooling in rural the need to tread lightly and be well informed areas. These programs target disadvantaged by on-the-ground research. and poorly performing schools, including a Indigenous peoples have dealt with the substantial number that provide services to expansion of western influence and market indigenous groups. A randomized evalua- systems in at least two broad ways: some tion of one such program shows that parental have confronted the expansion and either pressure produced efficiency gains, increasing adapted to it or been deracinated, and oth- access and improving test scores.32 ers have retreated farther into the backlands To work, such empowerment must be to avoid exposure and deal with the modern accompanied by real efforts to generate, world at arm’s length. 34 We know a great use, and disseminate disaggregated data and deal about the vulnerability and well-being of information on vulnerable groups—some- the fi rst group, but little about the second.35 thing that can be done well only to the extent Nor is it clear what role, if any, the govern- that those groups play a role in conceptualiz- ment should play in enhancing the well-being ing and implementing the data gathering, and of people who eschew the modern economy. in formulating policy. Two examples illuminate the issues. Although much of the economic disad- vantage of indigenous and excluded groups Persistent poverty in Central Africa is the result of lower human capital, there is also evidence of labor market discrimi- Among the oldest inhabitants of Africa, the nation. 33 In decompositions of the earnings Pygmy peoples lead a nomadic or semino- gap, the portion of the difference between madic lifestyle (depending on the country) indigenous and nonindigenous peoples that is that has persisted largely unchanged for “unexplained”—perhaps because of discrim- thousands of years, with livelihoods based on ination or other unidentified factors—repre- hunting, fishing, and gathering wild fruits and sents one quarter to more than one half of the nuts. Because of urbanization and deforesta- total differential. This means that although tion, however, many Pygmies have become about half the earnings differential can be sedentary, leaving them disadvantaged and influenced by improvements in human capital vulnerable. In the Democratic Republic of (education, skills, and abilities that an indig- Congo, for example, Pygmies have tradition- enous person brings to the labor market), ally been closely attached to the rain forest— another half may result from discriminatory the source of their spirituality, livelihood, labor market practices or other factors over and protection. They once lived in camps of which the indigenous person has little con- 30–40 families, maintaining regular links trol. So efforts to empower excluded groups with each other. But today their lifestyle is in and increase their human capital need to look danger, as they become more sedentary, lose at how the labor market can be made more access to the forest, and face a deteriorating fair for them. relationship with Bantu farmers. Census data for the Central African Republic and Gabon show that Pygmies lag behind the general Small and largely untouched population in wealth, education, and access groups to basic infrastructure. Indigenous peoples across the globe have very Their poverty rates are extremely high. In different exposures to nonindigenous peoples the Democratic Republic of Congo, 85 per- and varying degrees of cooperation with gov- cent of the Pygmy population is estimated to ernments and government programs. It is be poor, compared with 72 percent of the gen- hard to design programs that would benefit eral population.36 Unemployment is typically 118 A S S I S T I N G I N D I G E N O U S A N D S O C I A L LY E X C L U D E D P O P U L AT I O N S GLOBAL MONITORING REPORT 2011 lower among Pygmies, perhaps because they of their environment and the possibilities for are so poor that they cannot afford not to humans to adapt to it sustainably. They are work. But the share of workers who are not increasing their integration into broader soci- paid cash for their work is much higher, sug- ety as they become more sedentary, and this gesting low-productivity jobs. process—unmanaged and with little input School enrollment rates and average years from the Pygmies—is linked to their impov- of schooling among Pygmies in the Central erishment, exploitation, and poor health and African Republic and Gabon are dramatically education outcomes. Although these findings lower than among other groups. The Central do not point to specific policies that the gov- African Republic’s gross primary enrollment ernment could implement to improve Pyg- is only 21 percent for Pygmies, compared mies’ living conditions, they do suggest the with 73 percent for nonindigenous peoples; need for more qualitative and institutional Pygmy men average 0.3 years of schooling analysis of measures to target this especially and women average 0.1 years, compared vulnerable group. with 2.8 years for non-Pygmy men and 1.4 years for non-Pygmy women. Gabon’s gross Lack of interaction with wider society: secondary enrollment is only 4 percent for the case of the Tsimane’ of Bolivia Pygmies, but 80 percent for non-Pygmies; and the average years of schooling are 3.0 The Tsimane’, a native Amazonian society of for men and 2.8 for women among Pygmies, foragers and farmers in Bolivia, provide an compared with 6.5 years for both non-Pygmy intriguing example of the benefits and costs men and women. Lower school enrollment when an indigenous group exercises signifi- among Pygmies may result from high cost, cant autonomy in its exposure to the outside distance to schools, weaker cultural empha- world. Active withdrawal into the backlands sis, and the need for children to work. and collaboration with Protestant missionar- Pygmies in the Central African Republic ies since the late 1940s have enabled Tsimane’ and Gabon are less likely to have access to to retain a great deal of their traditional cul- safe water and electricity; more likely to live ture and territory.37 They show no change in in a dwelling with walls, roof, or floor made core cultural values38 or in ethnobotanical of temporary materials; and less likely to have knowledge,39 although a more recent unpub- adequate sanitation. Even compared with lished study by Reyes-García et al.40 suggests other rural households, Pygmies in Gabon they may be losing that knowledge at the rate fare worse in all housing dimensions. of 2 percent a decade. In the Central African Republic, the share Close to 75 percent of the population of households living in a village or area with follows the traditional preferential system a health facility is smaller for Pygmies than of cross-cousin marriage (that is, a man for the population as a whole, although the marries his mother’s brother’s daughter or differences in distances to health facilities are father’s sister’s son),41 which produces a small. Access to condoms is much lower for tight endogamic group. With ample access to Pygmies; and the share of households that farmlands, Tsimane’ daily personal income lost a member to a long-term illness is higher (adjusted for inflation and in purchasing among Pygmies than among the overall pop- power parity) reaches $9 a day per person, ulation (although the share of households that much higher than the international poverty have lost a member to HIV/AIDS is smaller). line of $1–$2 a day.42 And, over the period The Pygmies embody a valuable cultural 2002– 06, many indicators of individual heritage that should be protected and pre- well-being improved, including real income, served, one of the most original forms of real wealth, consumption, body-mass index, human adaptation to the ecology of the rain social capital, and happiness.43 Little wonder forest. They have a sophisticated knowledge that permanent outmigration is rare; but they GLOBAL MONITORING REPORT 2011 A S S I S T I N G I N D I G E N O U S A N D S O C I A L LY E X C L U D E D P O P U L AT I O N S 119 are being encroached on by highland colonist breaking the intergenerational transmission farmers, cattle ranchers, and logging and oil of social exclusion. firms.44 The international community needs a Tsimane’ experience little damage in knowledge base of what specific programs physical stature,45 widespread parasitism46 and policies work best, and for which vul- and child growth stunting, and vulnerabil- nerable populations. It is widely held that ity to adverse weather conditions (such as outright discrimination may explain some excessive rainfall). Such events during gesta- of the observed differential in poverty out- tion, infancy, and early childhood suppress comes among minority groups,48 and there adult physical stature, particularly among is evidence consistent with labor market dis- women.47 And when adverse shocks strike crimination against indigenous peoples.49 But households (illness, theft, crop loss, floods), overt tests for discrimination against indig- people weather them on their own, without enous peoples (such as tests in the United help from kith, kin, or formal institutions. States comparing call-back rates for blacks That suggests the Tsimane’ remain poorly and whites with otherwise similar profiles)50 insured against mishaps that strike either the are distinctly lacking. individual or the village. To properly address the needs of the world’s excluded populations, more and bet- ter data are needed. That is, one needs to Conclusions analyze disaggregated data that are collected Indigenous groups make up a sizable share of regularly and consistently. This data collec- the world’s people and even a greater share of tion and analysis effort can only be done the world’s poor. But their isolation reduces successfully to the extent that the minority their gains from overall growth and makes populations play a role, thus ensuring cover- them difficult to reach through targeted pov- age and relevance. erty programs. As we come closer to 2015, Extending the gains from development to reaching even more isolated people will indigenous people will not be easy. It will be become progressively more difficult. Address- necessary to continue progress in overall pov- ing their needs will require widespread and erty reduction; to overcome physical, institu- sustainable economic growth and poverty tional, and social obstacles to their participa- reduction, along with strategies to reach those tion in broader society; and to empower them who need a special lift. Helping indigenous by supporting their political and social rights. groups will require better understanding of their needs and of how they interact with the modern economy. Nevertheless, some programs have improved Notes their lot. For example, well-implemented 1. In formal terms, “indigenous” is more suit- bilingual education programs promote school able for describing socially excluded groups completion and subsequent earnings gains, in the western hemisphere and other areas of thus contributing to multiple goals; and these mass migration than in areas where majority programs can be cost effective. CCT pro- groups have been in place for many centuries. grams can promote schooling, health, and However, “indigenous peoples” has been used generally to refer to socially excluded groups, poverty reduction goals; and in countries and we will continue that usage here. with large vulnerable populations, CCTs 2. Hall and Patrinos 2011. show disproportionate gains for the minority 3. Ibid 2011. This gap is expressed as the money population. Empirical research suggests that required to raise the poor from their present investments in early childhood development income to the poverty line, as a proportion of and improving school attendance and com- the poverty line, and averaged over the total pletion are the most promising means of population. 120 A S S I S T I N G I N D I G E N O U S A N D S O C I A L LY E X C L U D E D P O P U L AT I O N S GLOBAL MONITORING REPORT 2011 4. Garcia-Moreno and Patrinos forthcoming. 49. Daley and Lui 1995; Patrinos and Sakellariou 5. Ravallion 2010. 1992; Brosnan 1984; Kimmel 1997; Hall and 6. Das et al. forthcoming. Patrinos 2006; Ohenjo et al. 2006. 7. Hoff and Pandey 2004; UNESCO 2010. 50. Bertrand and Mallainathan 2003. 8. For example, Alesina et al. 2003. 9. Lewis and Lockheed 2006. 10. WHO 2010, p. 8. References 11. Parodi, Muñoz, and Sanhueza 2010. 12. UN 2010b. Alesina, A., A. Devleeschauwer, W. Easterly, S. 13. Lewis and Lockheed 2006, p. 8. Kurlat, and R. Wacziarg. 2003. “Fraction- 14. Das et al. forthcoming. alization.” Journal of Economic Growth 8: 15. King and Van de Walle forthcoming. 155–94. 16. Figure 1.3 and table 1.1 in chapter 1 and table Bando, R., L. F. Lopez-Calva, and H. A. Patrinos. A1.1 in the appendix show the countries by 2005. “Child Labor, School Attendance, and MDG status. Indigenous Households: Evidence from Mex- 17. Borja-Vega, Lunde, and Garcia Moreno 2007. ico.” Policy Research Working Paper 3487, 18. Das et al. 2010. World Bank, Washington DC. 19. MacDonald forthcoming. Baulch, B., T.T.K. Chuyen, D. Haughton, and J. 20. Dang 2010. Haughton. 2007. “Ethnic Minority Develop- 21. Ravallion 2010. ment in Vietnam.” Journal of Development 22. World Bank 2009. Studies 43 (7): 1151–76. 23. Ravallion 2010. Becker, G. S. 1971. The Economics of Discrimi- 24. Ravallion 2010; Lanjouw and Ravallion nation. 2nd edition. Chicago: University of 1999. Chicago Press. 25. Dollar, Glewwe, and Litvack 1998. Ben-Achour, A., P. Backiny-Yetna, and Q. Wodon. 26. Baulch et al. 2007. 2011. “Socioeconomic Status of the Pygmies in 27. Hall and Patrinos 2006. the Democratic Republic of Congo.” Unpub- 28. Fiszbein and Schady 2009. lished manuscript, Human Development Net- 29. Hall and Patrinos 2011. work, World Bank, Washington, DC. 30. See, for example, Dollar and Kraay (2002). Bertrand, M., and S. Mallainathan. 2003. “Are 31. Ravallion 2010. Emily and Greg More Employable Than Lak- 32. Gertler, Patrinos, and Rodríguez-Oreggia isha and Jamal? A Field Experiment on Labor 2010. Market Discrimination.” Working Paper 9873, 33. Nopo, Chong, and Moro 2010; Patrinos, National Bureau of Economic Research, Cam- Skoufias, and Lunde 2010; Hall and Patrinos bridge, MA. 2006; Psacharopoulos and Patrinos 1994. Borja-Vega, C., T. Lunde, and V. Garcia-Moreno. 34. Godoy et al. 2005; Rubio et al. 2009. 2007. “Economic Opportunities for Indigenous 35. But see Scott (2009), who provides a compre- People in Mexico.” Background paper for the hensive overview of various Southeast Asia study Economic Opportunities for Indigenous groups who have avoided contact with mod- Peoples in Latin America, Conference Edition. ern society. World Bank, Washington, DC. 36. Ben-Achour, Backiny-Yetna, and Wodon 2011. Brosnan, P. 1984. “Age, Education and Maori- 37. Huanca 2008; Ringhofer 2009. Pakeha Income Differences.” New Zealand 38. Rubio et al. 2009. Economic Papers 18: 49–61. 39. Godoy, Reyes-García, Broesch et al. 2009. Daley, A. E., and J. Lui. 1995. “Estimating the 40. Reyes-García et al. 2009. Private Rate of Return to Education for Indig- 41. Godoy, Eisenberg et al. 2008. enous Australians.” Discussion Paper 97, Cen- 42. Godoy et al. 2005. tre for Aboriginal Economic Policy Research, 43. Godoy, Reyes-García, Gravlee et al. 2009. Australian National University, Canberra. 44. Reyes-García and others 2010. Dang, H.-A. H. 2010. “Growth in Vietnam Is 45. Godoy et al. 2006. Strong But Not Shared Equitably across Ethnic 46. Tanner et al. 2009. Groups.” Indigenous Peoples Country Brief 47. Godoy, Tanner et al. 2008. (Vietnam) No. 1, World Bank, Washington, 48. Becker 1971. DC. GLOBAL MONITORING REPORT 2011 A S S I S T I N G I N D I G E N O U S A N D S O C I A L LY E X C L U D E D P O P U L AT I O N S 121 Das, M. B., G. Hall, S. Kapoor, and D. Nikitin. Godoy, R., V. Reyes-García, C. C. Gravlee, T. 2010. “India’s Adivasis.” Indigenous Peoples Huanca, W. R. Leonard, T. W. McDade, and Country Brief (India) No. 4, World Bank, S. Tanner. 2009. “Moving Beyond a Snapshot Washington, DC. to Understand Changes in the Well-Being of ——— . Forthcoming. “India: The Scheduled Native Amazonians.” Current Anthropology Tribes” In Indigenous Peoples, Poverty and 50 (4): 563–73. Development, ed. G. Hall and H. A. Patrinos. Godoy, R., V. Reyes-García, T. Huanca, W. R. Cambridge, UK: Cambridge University Press. Leonard, V. Vadez, C. Valdés-Galicia, and de Laat, J. 2010. “Economic Costs of Roma D. Zhao. 2005. “Why Do Subsistence-Level Exclusion.” Europe and Central Asia Topic People Join the Market Economy? Testing Brief, Human Development Department, Hypotheses of Push and Pull Determinants in World Bank, Washington DC. Bolivian Amazonia.” Journal of Anthropologi- Dollar, D., P. Glewwe, and J. Litvack, eds. 1998. cal Research 61 (2): 157–78. Household Welfare and Vietnam’s Transition. Godoy, R., S. Tanner, V. Reyes-García, W. R. Washington, DC: World Bank. Leonard, T. W. McDade, M. Vento, J. Broesch, Dollar, D., and A. Kraay. 2002. “Growth Is Good I. C. Fitzpatrick, P. Giovannini, T. Huanca, for the Poor.” Journal of Economic Growth 7 and N. Jha. 2008. “The Effect of Rainfall dur- (3): 195–225. ing Gestation and Early Childhood on Adult Fiszbein, A., and N. Schady. 2009. Conditional Height in a Foraging and Horticultural Society Cash Transfers: Reducing Present and Future of the Bolivian Amazon.” American Journal of Poverty. Washington DC: World Bank. Human Biology 20 (1): 23–34. Garcia-Moreno, V., and H. A. Patrinos. Forth- Hall, G., and H. A. Patrinos, eds. 2006. Indigenous coming. “Education and Poverty in Mexico: Peoples, Poverty and Human Development in Indigenous Peoples.” Teachers College, Colum- Latin America. London: Palgrave Macmillan. bia University, New York; and Human Devel- ——— . 2011. Indigenous Peoples, Poverty and opment Department, World Bank, Washing- Development. Cambridge, UK: Cambridge ton, DC. University Press. Gertler, P., H. A. Patrinos, and E. Rodríguez-Oreg- Hartley, B. 2008. “MDG Reports and Indigenous gia. 2010. “Parental Empowerment in Mex- Peoples: A Desk Review.” Report prepared for ico: Randomized Experiment of the Apoyos the Secretariat of the United Nations Perma- a la Gestión Escolar (AGE) in Rural Primary nent Forum on Indigenous Issues, New York. Schools in Mexico: Preliminary Findings.” Heckman, J. J. 2006. “Skill Formation and the Unpublished manuscript, Human Development Economics of Investing in Disadvantaged Chil- Network, World Bank, Washington, DC. dren.” Science 312: 1900–02. Godoy, R., D.T.A. Eisenberg, V. Reyes-García, Hoff, K., and P. Pandey. 2004. “Belief Systems T. Huanca, W. R. Leonard, T. W. McDade, and Durable Inequalities: An Experimental and S. Tanner. 2008. “Assortative Mating and Investigation of Indian Caste.” Policy Research Offspring Well-Being: Theory and Empirical Working Papers 3351, World Bank, Washing- Findings from a Native Amazonian Society in ton DC. Bolivia.” Evolution and Human Behavior 29: Huanca, T. 2008. Tsimane’ Oral Tradition, Land- 201–10. scape, and Identity in Tropical Forest. La Paz, Godoy, R., W. R. Leonard, V. Reyes-Garcia, E. Bolivia: Imprenta Wagui. Goodman, T. McDade, T. Huanca, S. Tanner, Kimmel, J. 1997. “Rural Wages and Returns and V. Vadez. 2006. “Physical Stature of Adult to Education: Differences between Whites, Tsimane’ Amerindians, Bolivian Amazon, in blacks, and American Indians.” Economics of the 20th Century.” Economics and Human Education Review 16 (1): 81–96. Biology 4: 184–205. King, E. M., and D Van de Walle. Forthcoming. Godoy, R., V. Reyes-García, J. Broesch, I. C. Fitz- “Laos: Ethno-Linguistic Diversity and Disad- patrick, P. Giovannini, M. R. M. Rodríguez, T. vantage.” in Indigenous Peoples, Poverty and Huanca, W. R. Leonard, T. W. McDade, and Development, ed. G. Hall and H. A. Patrinos. S. Tanner. 2009. “Long-Term (Secular) Change Cambridge, UK: Cambridge University Press. of Ethnobotanical Knowledge of Useful Plants: Kremer, M., E. Miguel, and R. Thornton. 2009. Separating Cohort and Age Effects.” Journal “Incentives to Learn.” Review of Economics of Anthropological Research 65 (1): 51–67. and Statistics 91 (3): 437–56. 122 A S S I S T I N G I N D I G E N O U S A N D S O C I A L LY E X C L U D E D P O P U L AT I O N S GLOBAL MONITORING REPORT 2011 Lanjouw, P., and M. Ravallion. 1999. “Benefit India.” World Bank Research Observer 26 (1): Incidence, Public Spending Reforms, and the 71–104. Timing of Program Capture.” World Bank Reyes-García, V., E. Kightley, I. Ruiz-Mallén, Economic Review 13 (2): 257–73. N. Fuentes-Peláez, K. Demps, T. Huanca, Lewis, M. A., and M. E. Lockheed. 2006. Inexcus- and M. Martínez-Rodríguez. 2009. “School- able Absence: Why 60 Million Girls Still Aren’t ing and Local Environmental Knowledge: Do in School and What to Do about It. Washing- They Complement or Substitute Each Other?” ton DC: Center for Global Development. Unpublished manuscript, Tsimane’ Amazonian MacDonald, K. Forthcoming. “Indigenous Peo- Panel Study, Brandeis University, Waltham, ples and Development Goals: A Global Snap- MA. shot.” in Indigenous Peoples, Poverty and Reyes-García, V., M. Orta-Martínez, M. Gueze, Development, ed. G. Hall and H. A. Patrinos. A. C. Luz, J. Paneque-Gálvez, M. J. Macía, J. Cambridge, UK: Cambridge University Press. Pino, J. Cristobal, and TAPS Bolivian Study Nopo, H., A. Chong, and A. Moro, eds. 2010. Team. 2010. “Does Participatory Mapping Discrimination in Latin America: An Eco- Increase Conflict? A Randomized Experimen- nomic Perspective. Washington DC: World tal Evaluation in the Bolivian Amazon.” TAPS Bank. working paper 59. http://tsimane.org/research/ Ohenjo, N., R. Willis, D. Jackson, C. Nettleton, pgs/workingpapers.html. K. Good, and B. Mugarura. 2006. “Health Ringhofer, L. 2009. Fishing, Foraging and Farm- of Indigenous People in Africa.” Lancet 367: ing in the Bolivian Amazon: On a Local Soci- 1937–46. ety in Transition. New York: Springer. OSI (Open Society Foundation). 2010. No Data— Ringold, D., M. A. Orenstein, and E. Wilkens. No Progress: Country Findings. Data Collec- 2005. Roma in an Expanding Europe: Break- tion in Countries Participating in the Decade ing the Poverty Cycle. Washington DC: World of Roma Inclusion 2005–2015. New York. Bank. Parodi C. T., S. Muñoz, and A. Sanhueza. 2010. Rubio, K., E. Undurraga, O. Magvanjav, C. Grav- “Acceso y gasto de salud para grupos étnicos/ lee, T. Huanca, W. R. Leonard, T. W. McDade, raciales en la región de las Américas.” World V. Reyes-García, S. Tanner, R. Godoy, and Health Report 2010 background paper, no. TAPS Team. 2009. “Modernization and Cul- 46. http://www.who.int/healthsystems/topics/ ture Loss: A Natural Experiment among Native financing/healthreport/whr_background/en. Amazonians in Bolivia.” Tsimane’ Amazonian Patrinos, H. A. 2007. “Living Conditions of Chil- Panel Study Working Paper. http://tsimane.org/ dren.” In Solutions for the World’s Biggest research/pgs/workingpapers.html. Problems: Costs and Benefits, ed. B. Lomborg, Schlotter, M. 2010. “The Relationship between 358–75. Cambridge, UK: Cambridge Univer- Human Capital Quality, Governance and sity Press. Educational Institutions: First Steps towards Patrinos, H. A., and C. N. Sakellariou. 1992. a Panel Approach.” Unpublished manuscript, “North American Indians in the Canadian Human Development Network, World Bank, Labour Market: A Decomposition of Wage Dif- Washington, DC. ferentials.” Economics of Education Review 2 Scott, J. 2009. The Art of Not Being Governed: (3): 257–66. An Anarchist History of Upland Southeast Patrinos, H. A., E. Skoufias, and T. Lunde. 2010. Asia. Yale Agrarian Studies Series. New Haven, “Social Networks among Indigenous Peoples CT: Yale University Press. in Mexico.” Latin American Research Review Tanner, S., W. R. Leonard, T. W. McDade, V. 45 (2): 49–67. Reyes-García, R. Godoy, and T. Huanca. Patrinos, H. A., and E. Velez. 2009. “Costs and 2009. “Influence of Helminth Infections on Benefits of Bilingual Education in Guatemala: Childhood Nutritional Status in Lowland A Partial Analysis.” International Journal of Bolivia.” American Journal of Human Biol- Educational Development 29 (6): 594–98. ogy 21: 651–56. Psacharopoulos, G., and H. A. Patrinos, eds. Telles, E. E. 2007. “Race and Ethnicity and Latin 1994. Indigenous People and Poverty in Latin America’s United Nations Millennium Devel- America: An Empirical Analysis. Washington, opment Goals.” Latin American and Carib- DC: World Bank. bean Ethnic Studies 2 (2): 185–200. Ravallion, M. 2010. “A Comparative Perspective UN (United Nations). 1981. “Study of the Prob- on Poverty Reduction in Brazil, China and lem of Discrimination against Indigenous GLOBAL MONITORING REPORT 2011 A S S I S T I N G I N D I G E N O U S A N D S O C I A L LY E X C L U D E D P O P U L AT I O N S 123 Populations.” Final Report submitted by the for All Global Monitoring Report 2010: Special Rapporteur, Jose R. Martinez Cobo, Reaching the Marginalized. Oxford, UK: United Nations Economic and Social Council, Oxford University Press. Commission on Human Rights, Sub-Commis- WHO (World Health Organization). 2010. World sion on Prevention of Discrimination and Pro- Health Report 2010: Health Systems Financ- tection of Minorities, 34th session, E/CN.4/ ing: The Path to Universal Coverage. Geneva, Sub.2/476, New York. Switzerland. ——— . 2009. State of the World’s Indigenous World Bank. 2009. “From Poor Areas to Poor Peoples. New York. People: China’s Evolving Poverty Reduc- ——— . 2010a. Human Rights and the Mil- tion Agenda. An Assessment of Poverty and lennium Development Goals in Practice: A Inequality in China. Washington, DC. Review of Country Strategies and Reporting. ——— . 2010. “Roma Inclusion: An Economic New York. Opportunity for Bulgaria, Czech Republic, ——— . 2010b. The Millennium Development Romania and Serbia.” Policy Note, Human Goals Report. New York. Development Sector Unit, Europe and Central UNESCO (United Nations Educational, Scientific, Asia Region, Washington, DC. and Cultural Organization). 2010. Education 5 Progress in the International Development and Trading Framework T he global economy has weathered engage with them at the Fourth High Level the fi nancial crisis, but the recovery Forum on Aid Effectiveness, to be held in remains fragile. Governments and Busan, Republic of Korea, at the end of international institutions cooperated in the 2011. The objective is to forge a new consen- face of recession, thus avoiding the perils of sus on development assistance coming out a protectionist spiral, maintaining aid levels, of Busan—one that reflects the growing role and boosting emergency financial resources. of non-OECD-DAC countries. Developing Because a beneficial global environment also countries are also receiving more assistance supports progress toward the Millennium from private sources. Philanthropic flows to Development Goals (MDGs), this chapter developing countries from the private sector reviews recent changes in the international in advanced countries may have exceeded development framework—in the context of $60 billion in 2008, more than 10 times the the gradual transformation of development amount received early in the decade. policies over the past decade. Nonetheless, new donor flows will not Aid in 2010 rose in real terms over 2009, compensate for any significant fall in aid despite fears that the sharp rise in fiscal defi- from traditional donors, particularly if they cits during the crisis would severely constrain pursue different development priorities and donor budgets. Even so, recent research practices. The changing aid landscape could suggests that declines in aid have tended to also have implications for the transparency deepen for a period of years following past of official flows and for the policies and pro- banking crises. Some major donor govern- grams that aid supports. ments remained dedicated to maintaining World trade, recovering at about double aid in 2010, while others announced cuts. the 2002– 08 rate of growth, remains well New donors that are not members of the below the precrisis peak and even lower than Organisation for Economic Co-operation it would have been if trade had continued to and Development (OECD) Development follow the rising 1995–2008 trend. The new Assistance Committee (DAC) may provide protectionist measures during the crisis, which some $12 billion to $15 billion in aid dis- hit the exports of least-developed countries bursements a year. There is keen interest to (LDCs), appear to be receding. Solidifying an A human flood chokes the mud-slick remnants of a highway, the main vein for commerce through the Ituri, Democratic Republic of the Congo. Rough and rutted 125 during dry weather, the roads are nearly impassable in the wet season. But rain can’t stop the toleka traders, who push goods hundreds of miles by bicycle. 126 PROGRESS IN THE DEVELOPMENT AND TRADING FRAMEWORK GLOBAL MONITORING REPORT 2011 open, rules-based international trade regime trade flows were followed by the worst eco- can best be accomplished by (finally) conclud- nomic crisis since the Great Depression. ing the Doha Round. Efforts to increase trade Wars have led to shifts in priorities among integration of the poorest developing coun- international donors. And China and India, tries should focus on extending duty-free, with some other developing countries, have quota-free access to their exports, providing emerged as economic powerhouses. financial resources and technical assistance Over the last decade, donors reaffirmed to improve trade facilitation, and simplify- their commitments to increasing aid flows ing rules of origin in preference agreements. and improving aid quality, as in the Monter- Regional trade agreements should support rey Declaration at the 2002 Financing for open trade through low external tariffs, and Development conference in Mexico and the technical assistance to developing-country Gleneagles G-8 and Millennium +5 sum- trade negotiators would support deeper mits in 2005. At the September 2010 United regional integration. International efforts to Nations Summit on MDGs, held in New avoid a collapse in trade finance during the York, representatives of key stakeholders— crisis were timely and successful. But further including governments, international organi- steps are required to ensure that low-income zations, and private and public sectors—reaf- countries can obtain trade finance at reason- firmed their commitment to work together to able cost, along with improvements in data achieve the MDGs in the five years left. and a review of whether banking regulations impose excessive capital requirements on Will the crisis undermine recent trade finance transactions. commitments? The past tumultuous decade has ush- ered in significant changes in the assistance This reaffirmation was all the more policies of the international institutions. remarkable, given the potential impact of the The multilateral development banks have crisis on donor budgets. These may be threat- developed a results-based, country-driven ened as high income countries struggle to assistance framework grounded in regular control fiscal deficits. Even so, aid flows from reviews of country strategy and indepen- DAC donors rose in 2010 (to $128.7 billion in dent evaluations. A more diverse and flexible 2010 from $119.8 billion in 2009, a 6.5 per- range of financing facilities has helped tailor cent increase in real terms). Despite various assistance to the needs of particular countries shortfalls against commitments, the impact (such as those affected by conflict) and in par- of the financial crisis on aid, at least in 2009 ticular situations (such as disaster relief and and 2010, was relatively mild. Official devel- crisis assistance). The international financial opment assistance (ODA) as a percentage of institutions (IFIs) also are doing more in gross national income (GNI) rose from 0.26 providing knowledge in the context of loans percent in 2004 to 0.31 percent in 2009 and and technical assistance to their clients, to 0.32 percent in 2010. both through global projects and programs The United States remained the largest and through the free provision of informa- donor in 2009 ($28.8 billion), with France tion to the global community. The financing ($12.6 billion), Germany ($12.1 billion, the provided to crisis-affected countries jumped United Kingdom ($11.5 billion), and Japan sharply with the onset of the crisis, leading to ($9.5 billion) following. With a net inflow a general increase in IFI resources. of $6.1 billion, Afghanistan was the larg- est recipient in 2009, followed by Ethiopia ($3.8 billion), Vietnam ($3.7 billion), West An unprecedented decade for aid Bank and Gaza ($3.0 billion), and Tanzania Economic and political events since the Mil- ($2.9 billion) (box 5.1). The amount of aid for lennium Declaration in 2000 have reordered debt relief and technical assistance as a share the aid framework. Booming economic of total net official development assistance growth, higher capital flows, and surging declined. GLOBAL MONITORING REPORT 2011 PROGRESS IN THE DEVELOPMENT AND TRADING FRAMEWORK 127 BOX 5.1 Aid recipients Aggregate aid flows often reflect geopolitical priorities ($253), Georgia ($206), Afghanistan ($168), Serbia and/or responses to major global events. It is therefore ($142), and the Democratic Republic of Congo ($140). not a surprise that Iraq and Afghanistan were the larg- Low-income countries captured the largest share est aid recipients over the last decade (top table below). net official development assistance in 2008 and In per capita terms, the pattern is also mixed at the received the highest amount in per capita terms. By country level. The top 10 recipients in per capita net geographic region, Sub-Saharan Africa received the official development assistance in 2008 were West highest aggregate aid flows, but it is second to the Bank and Gaza ($659), Botswana ($373), Liberia Middle East and North Africa in per capita terms ($330), Iraq ($321), Lebanon ($257), Timor-Leste (lower table below) Largest aid recipients by decade billions, constant 2008 US$ 1960–69 1980–89 2000–09 India 64.7 India 42.4 Iraq 68.1 Pakistan 25.1 Egypt, Arab Rep. 31.0 Afghanistan 35.5 Algeria 20.7 Bangladesh 29.6 Nigeria 29.2 Vietnam 17.5 Israel 23.2 Ethiopia 27.4 Korea, Rep. 14.5 Indonesia 20.9 Vietnam 27.4 Brazil 13.1 Pakistan 20.5 Tanzania 25.4 Turkey 12.6 China 19.5 Congo, Dem. Rep. 24.8 Indonesia 12.4 Syrian Arab Republic 18.7 Pakistan 23.4 Congo, Dem. Rep. 8.4 Sudan 18.6 Mozambique 21.1 Israel 8.1 Tanzania 16.9 India 20.6 1970–79 1990–99 Source: DAC database. India 44.8 Egypt, Arab Rep. 41.1 Egypt, Arab Rep. 44.8 China 33.6 Indonesia 25.1 India 25.0 Pakistan 21.2 Bangladesh 21.3 Bangladesh 19.5 Indonesia 18.9 Syrian Arab Republic 18.2 Mozambique 16.3 Vietnam 15.6 Tanzania 15.2 Israel 13.6 Israel 14.4 Jordan 12.4 Pakistan 14.0 Korea, Rep. 10.5 Ethiopia 13.3 Aid flows by income groups and region, 2008 Income groups and regions Total ($ billions) Per capita ($) Income Low 41.4 42 Lower middle 37.1 10 Higher middle 11.8 12 Regions East Asia and Pacific 9.1 5 Europe and Central Asia 8.2 19 Latin America and the Caribbean 9.3 16 Middle East and North Africa 23.6 73 South Asia 12.3 8 Sub-Saharan Africa 40.1 49 Source: Word Development Indicators database. 128 PROGRESS IN THE DEVELOPMENT AND TRADING FRAMEWORK GLOBAL MONITORING REPORT 2011 The largest increases in real terms in ODA Portugal—appear unlikely to reach it. Simi- between 2009 and 2010 were recorded by larly, the United States set the goal of dou- Australia, Belgium, Canada, Japan, Korea, bling its aid to Sub-Saharan Africa between Portugal, and the United Kingdom. Many 2004 and 2010. It achieved this in 2009. countries (for example, the United Kingdom) Japan, by contrast, fell short of its Gleneagles have maintained their original commitments promise that its aid during 2005–09 would to aid flows made at the Gleneagles G-8 and equal $10 billion more than if it had remained Millennium +5 summits in 2005. Other donor at the 2004 level, owing to serious budgetary countries, however, have announced shifts in constraints. their aid policy because aid budgets are under Overall, donor countries have remained strain and economic growth in many high- pretty much on track in meeting their aid income countries is slow in recovering to pre- commitments, despite the economic down- crisis levels. The Dutch government plans to turn. But we will only really know the true cut foreign aid from 0.8 percent to 0.7 per- impact of the crisis on aid several years down cent of gross national income. Ireland and the road. Reinhart and Rogoff argue that the Spain have already cut their aid budgets, and full impact on aid flows of a recession fol- France has announced a postponement of its lowing a banking crisis can be felt for sev- commitments. eral years after the crisis, owing to enduring The year 2010 marked an important effects of the crisis on the country’s fiscal target for DAC countries that are part of position and growth prospects. 2 Similarly, the European Union. In 2005, they collec- Dang, Knack, and Rogers find that aid flows tively agreed to increase aid to 0.56 percent during 1977–2007 fell by 20–25 percent, on of gross national income within five years, average, from donor countries with bank- with a minimum country requirement of ing crises, beyond any income-related effects 0.51 percent. Some countries—including (figure 5.1).3 And the effects can last as long Belgium, Finland, Denmark, Ireland, Lux- as 10 years after the onset of a crisis.4 Is this embourg, the Netherlands, Spain, Sweden, pattern going to repeat itself? and the United Kingdom—appear to be on track to meet this goal.1 Others—including New donors appear on the scene Austria, France, Germany, Greece, Italy, and During most of the last half of a century, DAC donors, with international organiza- FIGURE 5.1 Impact of banking crises on net disbursed aid provided tions, accounted for most aid flows. Since by crisis-affected donors, 1977–2007 1960, the United States, Japan, Germany, France, and the United Kingdom have been 0.3 the largest donors in absolute terms; and the estimated banking-crisis coefficient 0.2 Nordic countries have given the most, rela- tive to gross national income.5 0.1 But the global balance of aid has started 0 to shift rapidly in the past five years because –0.1 some developing countries’ recent strong –0.2 growth gives them more resources to share in development. There is a clear lack of reli- –0.3 able data. Until recently, the DAC Creditor –0.4 Reporting System, which captures only a 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 partial picture of total aid flows, was the only years after onset of crisis available database. DAC donors are required to report their financial flows to the OECD, whereas non-DAC donors can voluntarily Source: Dang, Knack, and Rogers 2009. provide the information to the organization. GLOBAL MONITORING REPORT 2011 PROGRESS IN THE DEVELOPMENT AND TRADING FRAMEWORK 129 Twenty non-DAC donors now report to the a new consensus on aid effectiveness. With so OECD. Some of them—such as Saudi Ara- many actors, donor activities are less likely to bia, Thailand, and Turkey—provide sig- be coordinated effectively in recipient coun- nificant amounts of aid. Overall, non-DAC tries. Aid-funded projects are also likely to donors reporting to the DAC contributed follow different rules, increasing the risk of 5 percent of net aid ($6.7 billion in current more fragmentation and higher transaction dollars) in 2009. But other non-DAC donors costs for donors and recipients. Moreover, making significant contributions to develop- differences in aid policies may undermine ment assistance are not recorded by DAC, aid effectiveness through a lack of transpar- including Brazil, China, India, Malaysia, the ency, possible violations of international aid Russian Federation, and República Bolivari- and governance standards, and unfair com- ana de Venezuela. petition. Incorporating the viewpoints of new In total, non-DAC donors provided about actors is likely to enrich the discussions and $12 billion to $15 billion in aid in 2008.6 bring key voices of the international commu- Arab countries together would constitute the nity into this important global dialogue. largest non-DAC donor, contributing $5.9 billion.7 Development assistance from the Aid effectiveness BRIC countries (Brazil, Russia, India, and China) is estimated in the range of $2.3–$5.1 Aid effectiveness is a critical agenda for billion in fiscal 2006/07, with Chinese assis- partner-countries, donors, and other aid tance at $1.4 billion to $3.0 billion. These providers—including the World Bank. The figures may be underestimated. China’s aid importance of results of development efforts, to Africa alone was estimated at nearly $2.5 including external assistance, is the fourth billion in 2009.8 pillar of new research directions at the World Flows from multilateral institutions grew Bank.12 The 2005 Paris Declaration on Aid to $28 billion in 2009, and the number of Effectiveness and the 2008 Accra Agenda for institutions has increased. There are now Action have focused on the following prin- more than 260 multilateral aid agencies, ciples, intended to improve the quality and including development banks, the United effectiveness of aid: Nations, and specialized funds that have become important or have been set up in 1. strengthening country ownership of recent years.9 These funds tend to deliver aid development, with a special focus, such as HIV prevention 2. building more effective and inclusive part- and treatment, vaccines, or disaster preven- nerships for development, and tion and relief. 3. delivering and accounting for development Other key aid flows in recent years have results. come from civil society organizations and private philanthropists. No reliable figures Monitoring tools have been developed to exist for total aid from this group, but one improve the accountability of donors and recent estimate has their total aid reaching aid recipients. The Survey on Monitoring the $53 billion in 2008, more than 10 times the Paris Declaration was developed to monitor amount early in the decade.10 The largest is the implementation of these principles by the Gates Foundation, which has a total asset measuring progress against the commitments trust endowment of more than $35 billion made by both partner-countries and donors. and annual disbursements of $3 billion to $4 To measure progress between 2005 and billion.11 Additional work is clearly needed to 2010, surveys—consisting of 12 quantitative systematically gather more reliable data. indicators and qualitative information—were Still, the emergence of new aid flows and conducted in 2006 and 2008, with the third actors has led to changes and increased the survey scheduled for 2011. An evaluation tool complexity of the aid architecture, requiring called the Evaluation of the Implementation 130 PROGRESS IN THE DEVELOPMENT AND TRADING FRAMEWORK GLOBAL MONITORING REPORT 2011 of the Paris Declaration is intended to pro- (ADB) scored the highest among multilat- vide more comprehensive understanding of eral donors, followed by the World Bank, how increased aid effectiveness contributes to the International Monetary Fund (IMF), meeting development objectives (results forth- the International Fund for Agricultural coming in the spring of 2011). Development, and the Global Alliance for Several assessment tools have focused on Vaccines and Immunisation. the quality or effectiveness of donors, includ- • The QuODA Index, compiled by Nancy ing by observers such as Easterly and Pfutze,13 Birdsall and Homi Kharas, takes into Roodman,14 and the Brookings Institution’s account four broad categories covering Quality of Official Development Assistance 30 aid-related indicators: maximizing effi- Assessment (QuODA, developed in 2010). ciency, fostering institutions, reducing the Other attempts sponsored by bilateral donors burden on recipients, and transparency include the Multilateral Organizational Per- and learning.16 Different from Knack, formance Assessment Network, the United Rogers, and Eubank, QuODA does not Kingdom’s Multilateral Aid Review, and aggregate across sectors because of the low self-evaluations such as the Common Per- correlations among the four categories. It formance Assessment System. More recently, also looks at both bilateral and multilateral Knack, Rogers, and Eubank developed an agencies. The International Development index to compare donor performance and Association (IDA) and Ireland scored in assess how well donors are delivering on the top 10 for all four categories. their mandates.15 Two of these assessments are described here to provide a sense of the Transparency is emerging as a major com- results: ponent of aid effectiveness because it is essen- tial to enhancing accountability. For donors, • In Knack, Rogers, and Eubank 2010, the one of the most important initiatives is the selectivity subindex reviews policies and International Aid Transparency Initiative, measures of poverty reduction, based pri- which aims at establishing common global marily on GDP per capita and the World reporting standards on aid and publishing aid Bank’s Country Policy and Institutional information in a way that is accessible to var- Assessment index. The index covers the ious stakeholders for the purposes of global four most common dimensions of aid qual- reporting and improving budget management ity in the relevant literature—selectivity, by partner-countries. Progress on how donors alignment, harmonization, and special- rank on their transparency has been moni- ization. Alignment focuses on ways that tored in the 2010 report on “Aid Transpar- donor countries match their aid to support ency Assessment.”17 For partner-countries, country priorities and policies outlined in transparency in public finances is critical for country development strategies. Harmoni- accountability. Several frameworks attempt zation involves coordination among donors to measure fiscal transparency—frameworks that is designed to avoid duplication and such as the IMF’s Good Practices on Fiscal high transaction costs in recipient coun- Transparency; the OECD’s Best Practices for tries. The specialization subindex attempts Budget Transparency; the Public Expenditure to capture the fragmentation of donor aid, and Financial Accountability assessments including geographic concentration of aid, supported by the World Bank Group (WBG); concentration of aid by sector, and average and public sector accounting standards from size of projects. In the overall ranking of the International Federation of Accountants, the quality index (the average of the four the International Organization of Supreme subindexes), Denmark came first among Audit Institutions, and the International Bud- bilateral donors, followed by Ireland, the get Project’s Open Budget Index. Netherlands, Sweden, and the United The aim of these indexes is to give donor Kingdom; the Asian Development Bank countries and organizations a metric for GLOBAL MONITORING REPORT 2011 PROGRESS IN THE DEVELOPMENT AND TRADING FRAMEWORK 131 comparison as well as an objective frame- Much aid is, of course, targeted at improv- work for ways to improve the quality of aid ing health, nutrition, and education, as well over time. But aid is measured; a more funda- as at such nondevelopmental objectives as mental question remains: does aid contribute disaster relief. Such interventions are primar- to development at all? ily directed at welfare, particularly of the poor, although they also can have an impor- tant impact on growth. For example, the The impact of aid German and Irish governments’ funding for For decades, researchers have grappled with the Kalomo cash transfer program in Zam- the question of whether foreign aid inflows bia was likely intended to reduce poverty and lead to positive changes in a recipient coun- malnourishment, rather than to raise GDP try—economic and otherwise—and, if so, directly. in what form. Does aid lead to long-term Some researchers have looked at the improvements in the recipient country in the impact of social sector aid on country prog- form of increased growth? Or does it harm ress in achieving the MDGs. Michaelowa the economy in some way? And if it helps, and Weber fi nd aid for education (both per what type of aid helps best? capita and as a share of GDP) increases pri- Resolving these questions is critical for mary enrollment and completion rates. 26 aid policy, particularly given the huge sums. Dreher, Nunnenkamp, and Thiele fi nd that Some studies find that aid has a positive increased per capita aid for education signifi- impact on a country’s growth, others that a cantly increases primary school enrollment.27 positive impact depends on certain economic Mishra and Newhouse, using a data set of and political factors, and still others that aid 118 countries for 1973–2004, find that dou- has no impact on growth. Indicators include bling per capita health aid is associated with macroeconomic factors, openness or trade a 2 percent reduction in the infant mortality variables, governance and political condi- rate. Therefore, increasing per capita health tions, and geographic location.18 Burnside and aid from $1.60 to $3.20 will lead to roughly Dollar’s seminal paper concluded that aid has 1.5 fewer deaths per 1,000 births—a lift that a positive effect on growth in countries with is encouraging but small, relative to the MDG good policy environments.19 But according to targets.28 Doucouliagos and Paldam, no clear conclu- These results show that not all aid is effec- sion emerges from the aid literature.20 tive. Bourguignon and Sundberg point out Analyzing the effectiveness of particular the need to “open the black box” and bet- forms of aid, Clemens, Radelet, and Bhavnani ter understand the relationship between aid find a positive relationship between economic and development outcomes. 29 This can be growth and “short-impact aid.”21 Mavro- done through analyses that focus on three tas, reviewing the impact of different types basic relationships: the impact of policies on of aid on Uganda’s fiscal sector, finds that outcomes (knowledge, either ex post from project and food aid reduces public invest- impact evaluation or ex ante from model- ment, whereas program aid and technical ing or economic reasoning), the influence of assistance increase it. 22 Reddy and Minoiu, policy makers on policy, and the influence of distinguishing between what they call devel- external donors and international fi nancial opmental aid and geopolitical aid, 23 find that institutions on policy makers. the former has a strong, positive impact on Myriad factors play on whether aid growth, and the latter has a weak but nega- achieves its desired impact—or any positive tive relationship with growth.24 Heyman, in impact. Some examples include the level of her sectoral analysis of the link between aid recipient-country ownership and the condi- and economic improvement, fi nds that only tions imposed by donors, the predictability of humanitarian aid has a positive impact on future flows of aid, the investment climate of per capita GDP.25 the recipient country, whether aid is tied, and 132 PROGRESS IN THE DEVELOPMENT AND TRADING FRAMEWORK GLOBAL MONITORING REPORT 2011 the apparatus to monitor and evaluate prog- benefit from greater trade integration. To ress. The changing aid architecture, which has maintain the recent momentum toward new faces, new priorities, and a new structure, recovery, concerted efforts will be required will play a major role in the next years.30 Aid to keep protectionist tendencies in check and quality, together with strong international to recommit to building a stronger and more coordination, will continue to be a focus to effective multilateral trading system that improve the delivery and effectiveness of aid serves developing countries through the con- and its impact on developing countries clusion of the Doha Round. It will be impor- tant to ensure that regional trade agreements, which can enhance market access, promote Trade and the global economic open regionalism. Also needed are measures recovery to support access to trade fi nance and trade In the aftermath of the recent financial crisis, facilitation to connect vulnerable low-income trade has recently begun its path to recovery. countries, landlocked countries, and lag- Sustaining it will be critical. But many devel- ging regions to regional and international oping countries continue to face considerable markets—and an expansion of aid for trade challenges in reaching global markets as a to ensure that the developing countries can result of limited or weak trade-related infra- implement such reforms and reap the gains of structure and institutions, unfavorable busi- trade integration. ness climates, limited access to trade finance, and protectionist restrictions. High and vola- Trade remains crucial after the “great” tile food prices also create vulnerabilities, trade collapse particularly for low-income countries with high shares of food imports and limited fis- World merchandise trade has recovered since cal space. early 2009, but it remains below precrisis lev- After a brief review of the ongoing trade els (see figure 5.2). The world import value recovery, this section discusses the key chal- increased nearly 30 percent since its low point lenges and opportunities to strengthen the in February 2009, recovering at twice its international system and enhance the abil- growth rate during 2002–08. The recovery in ity of developing countries—particularly the trade is much faster among developing coun- more vulnerable low-income countries—to tries than among high-income countries. In the first 10 months of 2010, high-income-country FIGURE 5.2 Exports have yet to recover their precrisis trend export volumes grew at 10.4 percent a year, compared with 15.5 percent among developing countries. But the world import value remains log of real merchandise exports, US$ 2005 14 18 percent lower than the precrisis peak, even lower than if the world economy had contin- 13 ued to grow at the 1995–2008 trend. The pace of trade recovery is now decelerating, and the 12 prospects are unclear. Although the global GDP growth projection for 2011 has been revised upward (by a quarter of a point to 4.4 11 percent) in anticipation of stronger growth in the United States, persistent high food prices 10 threaten to undermine it. Jan Jan Jan Jan Jan Jan Jan Jan Jan Jan Jan 1991 1993 1995 1997 1999 2001 2003 2005 2007 2009 2011 Trade bounced back in all develop- ing regions (figure 5.3), driven by a vibrant high-income countries developing countries rebound in emerging economies. By 2010, all developing regions recovered to their precri- Sources: Datastream; World Bank Development Prospects Group. sis export volumes, with East Asia and the GLOBAL MONITORING REPORT 2011 PROGRESS IN THE DEVELOPMENT AND TRADING FRAMEWORK 133 FIGURE 5.3 Merchandise export volume 160 140 index (Jan 1, 2007 = 100) 120 100 80 60 M1 M3 M5 M7 M9 M11 M1 M3 M5 M7 M9 M11 M1 M3 M5 M7 M9 M11 M1 M3 M5 M7 M9 M11 M1 2007 2008 2009 2010 2011 East Asia and Pacific Latin America and the Caribbean South Asia Europe and Central Asia Middle East and North Africa Sub-Saharan Africa Source: World Bank Development Economics Prospects Group data. Pacific and South Asia (especially China and Benefits from global trade require India) leading this recovery. Export volumes continuing vigilance against in South Asia and East Asia and the Pacific protectionist tendencies peaked higher than precrisis levels. Trade vol- umes in the Middle East and North Africa Given the importance of trade and invest- also rebounded, but continue to be con- ment for the global economic recovery, the strained by the sluggish growth in their main G-20 commitment to resist all forms of pro- markets in the European Union. The down- tectionist measures, to keep markets open, turn in exports in Europe and Central Asia and to liberalize trade and investment to pro- turned a corner in August 2010, led by strong mote economic progress for all is welcome. intraregional trade that saw the region more This commitment, reiterated during the than recover its precrisis levels by the end November 2010 meeting in Seoul, Republic of 2010. Exports in Latin America and the of Korea, has been a centerpiece of the G-20 Caribbean remained lackluster, with sharp response to the global economic crisis. It may drops in October 2010 in Argentina and have spared the world economy from falling Chile as a result of strikes and disruptions in into a 1930s Depression-era scenario. metal supply and weather-related agricultural The World Trade Organization (WTO) shocks. Export growth in Africa has recov- reports some progress in lifting protectionist ered from its trough during the crisis, but measures imposed in the wake of the fi nan- remains volatile. Globally, low-income coun- cial crisis (figure 5.4). The latest WTO Trade tries have been insulated not only from the Policy Review (November 2010) indicates economic downturn, but also from the recov- that 415 measures were reported by mem- ery, reflecting their lack of integration in the bers between November 2009 and October world economy. Their exports and imports, 2010, down from 430 in the aftermath of in both volume and value terms, remained the crisis between October 2008 and Octo- comparatively flat during both the crisis and ber 2009. Whereas more than three-quarters the postcrisis recovery. of the measures were trade restrictive in the 134 PROGRESS IN THE DEVELOPMENT AND TRADING FRAMEWORK GLOBAL MONITORING REPORT 2011 FIGURE 5.4 Trade-restrictive measures have introduced since the end of 2008 now account fallen since 2008 for about 1.9 percent of world imports. G-20 countries remained the most active 350 instigators of restrictive—and liberalizing— no. of trade restrictive measures 15 300 20 measures. About 60 percent of the trade- 250 105 restrictive measures implemented between 200 13 25 November 2009 and October 2010 were 150 62 imposed by G-20 nations, led by BRICs (85 measures) with India at the head, followed by 100 184 high-income OECD countries (35 measures). 50 122 The United States had imposed the second- 0 Oct 2008–Oct 2009 Nov 2009–Oct 2010 highest number of trade-restrictive measures in the year after the crisis but reported only trade remedy export four new restrictive measures in the last border other year. Whereas restrictions imposed in high- Source: WTO Trade Policy Review OV 13. income OECD countries were mainly the result of trade remedy investigations—almost all antidumping initiations by the European first year, newly reported restrictive measures Union—BRICs brought out both nontariff came down to two-thirds in the second year. measures and trade remedy investigations. Retrenchment in protectionism since the More detailed information on the number of start of the crisis has nonetheless been slow, WTO member–initiated temporary trade bar- as only 15 percent of the restrictive measures rier investigations (antidumping, safeguards, introduced in 2008 have been terminated so and countervailing duty policies) from the far. Moreover, newly imposed export mea- World Bank’s Temporary Trade Barriers sures increased by 25 percent over the same database indicates that their number has gone period—primarily export bans and quotas down since the third quarter of 2009 and is on agricultural products partly resulting back to precrisis levels (figure 5.5). But over- from higher food prices. According to WTO all estimates of the impact of these border estimates, total import-restrictive measures and behind-the-border measures (including FIGURE 5.5 Newly initiated trade remedy investigations have peaked 50 newly initiated trade remedy investigations 45 40 35 30 25 20 15 10 5 0 Q1 Q2 Q3 Q4 Q1 Q2 Q3 Q4 Q1 Q2 Q3 Q4 Q1 Q2 Q3 Q4 2007 2008 2009 2010 initiated by developing economies initiated by developed economies Source: World Bank Temporary Trade Barriers database. GLOBAL MONITORING REPORT 2011 PROGRESS IN THE DEVELOPMENT AND TRADING FRAMEWORK 135 bailouts and subsidies) implemented since FIGURE 5.6 Trade-restrictive and trade-liberalizing measures, the crisis indicate that they contributed to an November 2009–October 2010 annual aggregate distortion to global trade of at least $35 billion.31 Algeria Armenia The number of trade-liberalizing mea- Argentina sures introduced in 2010 increased by 71 Australia percent over those introduced in 2009. More Azerbaijan Gulf Cooperation Council than a third of the new measures were tariff Bangladesh reductions, introduced primarily by BRICs Belarus Belarus, Kazakhstan, Russian Federation and lower-middle-income countries, such as Bolivia Pakistan and Bolivia. Reductions in nontariff Brazil measures made up a quarter of all liberalizing Canada Chile measures, mainly from the BRICs. Despite China global economic uncertainty, new measures to Colombia Costa Rica liberalize trade increased in 2010 (figure 5.6). Dominican Republic The LDCs are particularly harmed by Ecuador Egypt, Arab Rep. G-20 protectionism. New analysis by the European Union Global Trade Alert 32 indicates that, since Honduras November 2008, 141 trade measures imposed India Indonesia by countries worldwide harmed the commer- Iraq cial interests of LDCs. Of these, about 100 Israel Jamaica measures (70 percent of those imposed) were Japan introduced by G-20 members. Among the Jordan G-20, developing countries initiated 70 per- Kazakhstan Kenya cent of the measures, led, in order, by India, Korea, Rep. Argentina, Indonesia, and Russia. Among Macedonia, FYR Malaysia LDCs, Bangladesh has been affected by the Mercosur largest number of measures, followed by Mexico Tanzania, the Republic of Yemen, Senegal, Morocco Nigeria and Sudan. These measures may significantly Pakistan restrict LDC exports, particularly for some Paraguay Peru products they specialize in, thus contravening Philippines the undertaking at the G-20 summit. Russian Federation Rwanda High and volatile food prices are also Saudi Arabia creating vulnerabilities, particularly for low- Serbia income countries with high shares of food South Africa Sri Lanka imports and limited fiscal space. Whereas indi- Sudan vidual governments can insulate themselves to Switzerland Taiwan, China some degree from increases in world prices Thailand by imposing export bans or making tactical Turkey reductions in import tariffs, these measures Ukraine United States increase the volatility of world prices. Trade Uruguay policy is generally not the appropriate instru- Venezuela, R.B. de Vietnam ment to achieve food security and rural devel- Yemen, Rep. opment. High protection hurts households Zimbabwe that are net consumers. Long-term solutions 0 10 20 30 40 restrictions liberalizing include, among other things, higher agricul- tural productivity to raise farmer incomes and lower consumer prices, liberalization around Source: World Bank International Trade Department calculations based on WTO Trade Policy Review Overview Annex 1, November 2010. the world and commitment to multilateral Note: Of the 486 trade-related measures listed in the annex, 19 were unclassifiable because of insuf- trade rules, public access to information on ficient information regarding their trade effect. 136 PROGRESS IN THE DEVELOPMENT AND TRADING FRAMEWORK GLOBAL MONITORING REPORT 2011 the quality and quantity of grain stocks, and As unemployment rates rise despite the global effective safety nets.33 economic recovery, governments should con- Keeping markets open for LDC exports tinue to ensure that benefits of an open mul- would help lift those countries out of poverty. tilateral trading system are not compromised Extending 100 percent duty-free, quota-free by short-run pressures to protect domestic (DFQF) access to all exports of LDCs would markets. Keeping markets open will also be promote new export opportunities for LDCs, paramount to countering the effects of the particularly market access for products in withdrawal of expansionary fiscal and mon- which LDCs have a comparative advantage. etary policies. Although trade preference programs provide An effective and strong multilateral trad- high levels of product coverage, preferences ing system remains critical for improving are often underused and exclude products market access and accommodating new important to LDC exporters—for example, dynamics of trade. The structure of global agricultural goods (for African exporters) production has undergone a major trans- and textiles, apparel, and footwear (for Asian formation, with global value chains much exporters). more prevalent and elaborate, involving Extending DFQF from 97 percent to 100 more developing countries (including low- percent could have a significant impact. income countries). Although weak domestic Overall export gains for WTO-member policies, regulations, and institutions (par- LDCs could be on the order of $2 billion if ticularly those increasing the cost of logistics) G-20 OECD countries extended to 100 per- are the main binding constraint to progress, cent product coverage; this amount could be more gains may be achieved through a con- even larger if BRICs also offered full market certed, internationally integrated approach to access. For example, full DFQF access to all reform, given the multilateral nature of verti- G-20 markets could increase Malawi’s GDP cal production chains. by some 10 percent; for Tanzania, the ben- Now is the right time to reach an agree- efits of full DFQF granted by non-OECD ment on Doha. Commissioned by a group of G-20 countries are double those generated developed and developing economies, Peter through DFQF enacted by only the OECD Sutherland and Jagdish Bhagwati made the members. Given the size of the economies of case for a deadline to complete the talks low-income countries and the apparent com- before the end of 2011, highlighting three pre- plementarities with the G-20 countries, the vailing conditions as opportune. First, higher additional imports are unlikely to pose undue commodity prices can dissipate resistance by costs or disruptions to G-20 markets.34 More farmers benefiting from subsidies. Second, comprehensive efforts to harmonize and coor- the weak recovery enhances the attraction of dinate trade preferences would also increase a potential stimulus of approximately $360 the overall effectiveness of trade preferences. billion on global trade. And, third, the new global trade dynamics need to be addressed.35 Consider the growth and rising complexity of Concluding the Doha Round global supply chains: they mean that today’s The global financial crisis underlined the crit- protectionists are more likely to resort to ical role of a multilateral, rules-based trade targeted rules than to tariffs, making trade system. The WTO’s monitoring and report- negotiators’ traditional goal of dismantling ing of trade measures during the crisis helped tariffs less relevant. distinguish WTO-compatible policies from The November 2010 G-20 meeting in Seoul discriminatory policies, crucial to ensuring and the Japan summit of the Asia-Pacific Eco- a transparent and fair trading system. It also nomic Cooperation renewed hopes for prog- certainly contributed to restraining govern- ress in the Doha negotiations in the first half ments from adopting more pervasive and of 2011. G-20 and Asia-Pacific Economic numerous trade-restrictive policy measures Cooperation leaders saw a new window of to curtail the domestic impact of the crisis. opportunity to intensify negotiations in 2011, GLOBAL MONITORING REPORT 2011 PROGRESS IN THE DEVELOPMENT AND TRADING FRAMEWORK 137 MAP 5.1 Across the world, 884 million people lack access to safe water—around 80 percent of them in rural areas IBRD 38468 APRIL 2011 Access to safe water Greenland (Den) Faeroe Norway Iceland Islands Finland (Den) Sweden The Netherlands Russian Federation Estonia Canada Isle of Man (UK) Denmark Russian Latvia Fed. Lithuania United Ireland Kingdom Germany Poland Belarus Channel Islands (UK) Belgium Ukraine Luxembourg Moldova Kazakhstan Liechtenstein Romania Mongolia Switzerland France Italy Andorra Bulgaria Georgia Azer- Uzbekistan Kyrgyz D.P.R. Armenia baijan Turkmenistan Rep. United States Portugal Spain Monaco Turkey Tajikistan of Korea Japan Greece Syrian Gibraltar (UK) Cyprus Rep. of Tunisia Malta Lebanon Arab Islamic Rep. Afghanistan China Korea Israel Rep. Iraq of Iran Bermuda Morocco Kuwait (UK) West Bank and Gaza Jordan Bahrain Pakistan Nepal Bhutan Former Algeria Libya Arab Rep. Qatar The Bahamas Spanish of Egypt Saudi Cayman Is. (UK) Sahara Arabia United Arab Bangladesh Cuba Turks and Caicos Islands (UK) India Mexico Emirates Myanmar Lao Haiti Mauritania Oman P.D.R. Belize Jamaica Cape Verde Mali N. Mariana Islands (US) Niger Rep. of Thailand Vietnam Guatemala Honduras Senegal Chad Eritrea Yemen Guam (US) El Salvador Nicaragua The Gambia Burkina Sudan Marshall Faso Djibouti Cambodia Philippines Islands Guinea-Bissau Guinea Benin Federated States of Costa Rica Panama Sri Micronesia R.B. de Guyana Sierra Leone Côte Ghana Nigeria Central Ethiopia Lanka Venezuela Suriname Liberia D’Ivoire African Rep. Brunei Cameroon Somalia Palau Colombia French Guiana (Fr) Togo Malaysia Equatorial Guinea Uganda Maldives Ecuador São Tomé and Príncipe Gabon Congo Rwanda Kenya Singapore Nauru Dem. Rep. of Burundi Seychelles Papua Solomon Congo Indonesia New Guinea Islands Tanzania Comoros Peru Brazil Timor-Leste Angola Malawi Mayotte Zambia (Fr) Vanuatu Bolivia Zimbabwe Mauritius Namibia Madagascar Paraguay Botswana Mozambique Réunion (Fr) New Caledonia Australia (Fr) Swaziland South Lesotho Africa Chile Uruguay Argentina Access to safe water New British Virgin Dominican Puerto Poland Percent of population without access Zealand Germany Republic Rico (US) Islands (UK) Ukraine Anguilla (UK) Czech Rep. Slovak Rep. to improved water source U.S. Virgin Islands Antigua and Barbuda (US) St. Maarten (Neth) St. Kitts Guadeloupe (Fr) Austria Hungary ≥ 50 and Nevis Dominica Slovenia Romania 30–49 Curaçao Montserrat (UK) Martinique (Fr) Croatia (Neth) St. Lucia Bosnia and 10–29 Aruba Bonaire St. Vincent and San Herz. Serbia Bulgaria (Neth) (Neth) The Grenadines Barbados Marino Kosovo 6–9 Montenegro Grenada FYR ≤5 Italy Macedonia Trinidad Vatican Albania R.B. de Venezuela and Tobago City Greece no data MDG performance Greenland (Den) Faeroe Norway Iceland Islands Finland (Den) Sweden The Netherlands Russian Federation Estonia Canada Isle of Man (UK) Denmark Russian Latvia Fed. Lithuania United Ireland Kingdom Germany Poland Belarus Channel Islands (UK) Belgium Ukraine Luxembourg Moldova Kazakhstan Liechtenstein Romania Mongolia Switzerland France Italy Andorra Bulgaria Georgia Azer- Uzbekistan Kyrgyz D.P.R. Turkey Armenia baijan Turkmenistan Rep. United States Portugal Spain Monaco Tajikistan of Korea Japan Greece Syrian Gibraltar (UK) Cyprus Rep. of Tunisia Malta Lebanon Arab Islamic Rep. Afghanistan China Korea Israel Rep. Iraq of Iran Bermuda Morocco Kuwait (UK) West Bank and Gaza Jordan Bahrain Pakistan Nepal Bhutan Former Algeria Libya Arab Rep. Qatar The Bahamas Spanish of Egypt Saudi Cayman Is. (UK) Sahara Arabia United Arab Bangladesh Cuba Turks and Caicos Islands (UK) India Mexico Emirates Myanmar Lao Haiti Mauritania Oman P.D.R. Belize Jamaica Cape Verde Mali N. Mariana Islands (US) Niger Rep. of Thailand Vietnam Guatemala Honduras Senegal Chad Eritrea Yemen Guam (US) El Salvador Nicaragua The Gambia Burkina Sudan Marshall Faso Djibouti Cambodia Philippines Islands Guinea-Bissau Guinea Benin Federated States of Costa Rica Panama Sri Micronesia R.B. de Guyana Sierra Leone Côte Ghana Nigeria Central Ethiopia Lanka Venezuela Suriname Liberia D’Ivoire African Rep. Brunei Cameroon Somalia Palau Colombia French Guiana (Fr) Togo Malaysia Equatorial Guinea Uganda Maldives Ecuador São Tomé and Príncipe Gabon Congo Rwanda Kenya Singapore Nauru Dem. Rep. of Burundi Seychelles Papua Solomon Congo Indonesia New Guinea Islands Tanzania Comoros Peru Brazil Timor-Leste Angola Zambia Malawi Mayotte (Fr) Vanuatu Bolivia Zimbabwe Mauritius Namibia Madagascar Paraguay Botswana Mozambique Réunion (Fr) New Caledonia Australia (Fr) Swaziland South Lesotho Africa Chile Uruguay Argentina MDG performance New Dominican Puerto British Virgin Poland MDG 7.C Halve by 2015 the proportion of Zealand Germany Republic Rico (US) Islands (UK) people without sustainable access to safe Ukraine Anguilla (UK) Czech Rep. U.S. Virgin Islands Antigua and Barbuda Slovak Rep. drinking water (US) Austria St. Maarten (Neth) St. Kitts Guadeloupe (Fr) Hungary and Nevis Slovenia on target Dominica Romania Curaçao Montserrat (UK) Martinique (Fr) Croatia close to target (Neth) Bosnia and Bonaire St. Vincent and St. Lucia San Herz. Serbia Aruba far from target Bulgaria (Neth) (Neth) The Grenadines Barbados Marino This map was produced by the Map Design Unit of The World Bank. Montenegro Kosovo no data The boundaries, colors, denominations and any other information Grenada FYR Italy Macedonia shown on this map do not imply, on the part of The World Bank Trinidad Vatican Albania Group, any judgment on the legal status of any territory, or any and Tobago City Greece R.B. de Venezuela endorsement or acceptance of such boundaries. Source: World Bank staff calculations based on data from the World Development Indicators database. 138 PROGRESS IN THE DEVELOPMENT AND TRADING FRAMEWORK GLOBAL MONITORING REPORT 2011 calling for multilateral cooperation to pursue through parallel efforts to support unilateral a comprehensive and balanced trade agree- liberalization or open access. 38 Supporting ment. WTO Director-General Pascal Lamy the regional integration efforts of low-income asked WTO ambassadors to engage in infor- countries through the promotion of trade mal sessions with the goal of breaking long- facilitation and regional infrastructure could standing stalemates (including sessions on accelerate the pace of achievement for many the sidelines of the World Economic Forum MDGs. But regional trade agreements could Davos meeting in January 2011). He called be trade diverting and further complicate the chairs of the negotiating committees for developing countries’ global integration pros- revised texts by late-April 2011 and urged pects—for example, through complex rules of WTO ambassadors to also pursue bilateral origin. Developing countries should be strate- and small-group discussions to resolve key gic about what is included in preferential trade differences and catch up with work among agreements, be selective about which trading the full membership on refining texts. Nego- partners to include in such agreements and tiations and consultations are scheduled to treaties, and avoid signing on to one-size-fits- commence on all Doha topics—agriculture, all trade agreements that might not be suit- nonagricultural market access, services, dis- able for their developmental stages. pute settlement, facilitation, trade and envi- In light of the growing number of prefer- ronment, intellectual property rights, and the ential trade agreements, efforts should con- development agenda. centrate on building the capacity of trade Locking in the Doha Round is also cru- negotiators from developing countries. With cial so that the international community can some 474 regional trade agreements notified address emerging issues beyond those existing to the WTO as of July 2010, regional inte- in the Doha agenda.36 A fundamental shift is gration has become an unavoidable feature taking place in the world economy, but the of the international trading regime. More multilateral trading system has not adapted. of these agreements included “deep regional Mattoo and Subramanian point out that integration” measures, involving regulatory the world economy is moving broadly from and policy issues such as investment, intellec- conditions of relative abundance to relative tual property rights, environmental protec- scarcity, so economic security has become tion, technical and sanitary standards, migra- a paramount concern for consumers, work- tion and labor policies, and an array of other ers, and ordinary citizens.37 Addressing these behind-the-border regulations. The breadth new concerns—relating to food, energy, and and coverage of deep regional integration is economic security—requires a wider agenda new for many developing countries. Negoti- of multilateral cooperation involving not just ating and implementing such policies could the WTO but other multilateral institutions be complicated and, without proper guidance as well. Although these issues are growing and sufficient information, they could be tac- in importance and require a multilateral tically and commercially destructive. Many approach, abandoning the present Doha developing-country trade negotiators and negotiations in favor of an entirely new round government agencies may not have the legal of talks with a more up-to-date agenda, as or policy knowledge to strategically negoti- some have advocated, has even less chance to ate the appropriate regulatory components of get anywhere than has the Doha effort. regional agreements. To provide more information about pref- erential trade agreements to developing coun- Building capacity to support deeper but tries and the international community, the open regional integration World Bank recently launched a Web-based Regional integration can complement multi- portal, the Global Preferential Trade Agree- lateral liberalization. Regional trade agree- ment Database, 39 which provides real-time ments can mitigate potential adverse effects information on preferential trade agreements GLOBAL MONITORING REPORT 2011 PROGRESS IN THE DEVELOPMENT AND TRADING FRAMEWORK 139 around the world (including agreements as they aim to become more integrated with that have not been notified to the WTO). the global economy. Designed to help trade policy makers, schol- As a follow-up to the Multi-Agency Sup- ars, and business operators better understand port Team’s work on NTMs that revised the and navigate the world of preferential trade NTM classification and collected data in agreements, the database was merged with a pilot phase, the World Bank, the Interna- the revamped Web-based World Integrated tional Trade Centre, and the United Nations Trade Solutions, multiagency software pro- Conference on Trade and Development have viding access to trade-related data and inte- initiated data collection and dissemination grated trade simulation tools. efforts worldwide. As of December 2010, A services “platform” could help fill in the data had been collected in about 30 countries information gap for services negotiations. and should soon populate the World Inte- Bringing together government agencies, reg- grated Trade Solutions database.41 Data col- ulators, and private sector stakeholders to lection and dissemination are accompanied discuss regulatory reforms related to services by capacity-building programs at country would improve policy makers’ expertise.40 and regional levels to improve the governance Considering that services trade is paramount of NTMs, streamline them, and assess their to the export strategies for many LDCs and impact on trade. More transparent informa- that multilateral negotiations on further tion on NTMs should lead to more informed services liberalization and issues of labor policy making for developing countries and mobility (beyond the General Agreement on facilitate the harmonization and streamlining Trade and Services) are inevitably tied to the of these measures. eventual conclusion of the Doha Round, this To further increase the benefits of their would provide important information that preferential trade programs, developed coun- developing countries need as they pursue tries should also consider simplifying the services trade liberalization through unilat- rules of origin pertaining to their preference eral reforms or regional and bilateral trade programs. The estimated administrative cost agreements. for LDCs to comply with such rules is rela- As developing countries pursue regional tively high—in the range of 3–4 percent of integration, a better understanding of non- the value of the goods traded. But because tariff measures (NTMs)—their prevalence, the majority of most-favored-nation tariffs in impact, and ways to streamline them—should high-income countries are below this range, be at the forefront of governments’ export many LDCs choose not to use the preference competitiveness agenda. By definition, NTMs margin. Harmonizing multilaterally would encompass a broad range of barriers, includ- bring the administrative costs down. Those ing quotas, import licensing systems and costs can also be reduced by adopting more procedures, sanitary regulations, technical trader-friendly approaches, such as allow- requirements, product standards, and label- ing self-certification methods. The rules can ing requirements. As a result of this broad be further eased by allowing less restrictive defi nition, there is a lot of discretion in the cumulation rules (such as diagonal or full application of NTMs, whether for legitimate cumulation), allowing duty drawback, and policy objectives (health and consumer pro- setting higher de minimis levels.42 tection) or as protectionist measures. Poorly designed NTMs can also harm the competi- Trade facilitation for LDCs, landlocked tiveness of developing countries by constrain- countries, and lagging regions ing the ability of companies to outsource key inputs, putting them at a competitive disad- Exporters in many developing countries con- vantage on global markets. Enhancing trans- tinue to face weak transport infrastructure parency on the adoption and application of and burdensome trade procedures. Land- NTMs is thus crucial for developing countries locked LDCs, in particular, face unique 140 PROGRESS IN THE DEVELOPMENT AND TRADING FRAMEWORK GLOBAL MONITORING REPORT 2011 BOX 5.2 The donor community is enhancing Africa’s trade facilitation The international community has reinforced its West African States and the Economic Community support to LDCs, especially in Africa. In addition of Central African States). The facility also has pro- to the Multi-Donor Trade Facility, the World Bank vided funding for regional transportation integra- multidonor Trade Facilitation Facility, launched in tion in East Asia and the Pacifi c, Europe and Cen- July 2009, helps developing countries implement tral Asia, and Latin America. The aim is to provide trade facilitation reforms and scale up trade-related initial funding to support technical advisory services infrastructure and institutions. As of June 2010, the and capacity building for project preparation that facility was supporting 29 projects, with $20 mil- would later lead to concrete and measurable trade lion approved. Most projects are in Africa (Camer- facilitation improvements. In addition, the facil- oon, Democratic Republic of Congo, Lesotho, Nige- ity supports the Doha negotiations by funding case ria, Rwanda, South Africa, Togo, Zimbabwe, and studies to assess the implementation challenges of regional projects with the Economic Community of trade facilitation agreements for LDCs. infrastructure and institutional hurdles to and better-connected leading areas. More- overseas markets. For many landlocked over, logistics costs are exceptionally high LDCs, connecting to global supply chains is a because of poor logistics services and infra- formidable challenge caused by long distances structure, low economies of scale (transport- from seaports and the need to cross borders. ing small consignments across long distances), Landlocked countries tend to trade, on aver- and highly fragmented supply chains with age, 30 percent less than coastal countries. numerous intermediaries exacting time and Connecting landlocked LDCs to global sup- cost at each link. So producers and exporters ply chains could expand their employment, in lagging areas do not reap the full benefits income, and consumption opportunities. of trade. Policy actions should ensure better With 20 of 54 low-income countries land- coordination and cooperation among pro- locked (mostly in Sub-Saharan Africa), trade ducers within the supply chain and leverage facilitation that connects landlocked regions public-private partnerships to improve trans- with international markets will open consid- port infrastructure, to apply information erable economic opportunities (box 5.2). It technology, and to reduce the information could include streamlining transit regimes, asymmetry within supply chains (figure 5.7). implementing border and customs reform, improving transport service quality, and Trade facilitation through support to coordinating multimodal nodes within the trade finance in low-income countries supply chain. Limited competition on domes- tic routes is an explanatory factor behind the The international community should continue high prices charged for domestic shipments.43 to increase the availability of trade fi nance So constraining monopoly power and remov- in developing countries—particularly low- ing barriers to entry and exit would lower income countries—to facilitate trade. Trade costs for importers and exporters. fi nance, which includes a range of fi nancial Logistical improvements can connect instruments for trade transactions, underpins rural and remote areas within developing the financial infrastructure that enables coun- countries.44 Even with the sharper focus on tries and firms to trade with one another. The connecting countries to trade corridors and lack of trade fi nance can have severe impli- supply chains, there is still a gaping divide cations for a pro-development global trading between lagging areas in developing countries system. The issue of trade finance availability GLOBAL MONITORING REPORT 2011 PROGRESS IN THE DEVELOPMENT AND TRADING FRAMEWORK 141 MAP 5.2 With half the people in developing regions without sanitation, the 2015 target appears to be out of reach IBRD 38469 APRIL 2011 Access to sanitation Greenland (Den) Faeroe Norway Iceland Islands Finland (Den) Sweden The Netherlands Russian Federation Estonia Canada Isle of Man (UK) Denmark Russian Latvia Fed. Lithuania United Ireland Kingdom Germany Poland Belarus Channel Islands (UK) Belgium Ukraine Luxembourg Moldova Kazakhstan Liechtenstein Romania Mongolia Switzerland France Italy Andorra Bulgaria Georgia Azer- Uzbekistan Kyrgyz D.P.R. Turkey Armenia baijan Turkmenistan Rep. United States Portugal Spain Monaco Tajikistan of Korea Japan Greece Syrian Gibraltar (UK) Cyprus Rep. of Tunisia Malta Lebanon Arab Islamic Rep. Afghanistan China Korea Israel Rep. Iraq of Iran Bermuda Morocco Kuwait (UK) West Bank and Gaza Jordan Bahrain Pakistan Nepal Bhutan Former Algeria Libya Arab Rep. Qatar The Bahamas Spanish of Egypt Saudi Cayman Is. (UK) Sahara Arabia United Arab Bangladesh Cuba Turks and Caicos Islands (UK) India Mexico Emirates Myanmar Lao Haiti Mauritania Oman P.D.R. Belize Jamaica Cape Verde Mali N. Mariana Islands (US) Niger Rep. of Thailand Vietnam Guatemala Honduras Senegal Chad Eritrea Yemen Guam (US) El Salvador Nicaragua The Gambia Burkina Sudan Marshall Faso Djibouti Cambodia Philippines Islands Guinea-Bissau Guinea Benin Federated States of Costa Rica Panama Sri Micronesia R.B. de Guyana Sierra Leone Côte Ghana Nigeria Central Ethiopia Lanka Venezuela Suriname Liberia D’Ivoire African Rep. Brunei Cameroon Somalia Palau Colombia French Guiana (Fr) Togo Malaysia Equatorial Guinea Uganda Maldives Ecuador São Tomé and Príncipe Gabon Congo Rwanda Kenya Singapore Nauru Dem. Rep. of Burundi Seychelles Papua Solomon Congo Indonesia New Guinea Islands Tanzania Comoros Peru Brazil Timor-Leste Angola Malawi Mayotte Zambia (Fr) Vanuatu Bolivia Zimbabwe Mauritius Namibia Madagascar Paraguay Botswana Mozambique Réunion (Fr) New Caledonia Australia (Fr) Swaziland South Lesotho Africa Chile Uruguay Argentina Access to sanitation New British Virgin Dominican Puerto Poland Percent of population without access Zealand Germany Republic Rico (US) Islands (UK) Ukraine Anguilla (UK) Czech Rep. Slovak Rep. to sanitation facilities U.S. Virgin Islands Antigua and Barbuda (US) St. Maarten (Neth) St. Kitts Guadeloupe (Fr) Austria Hungary ≥ 50 and Nevis Dominica Slovenia Romania 30–49 Curaçao Montserrat (UK) Martinique (Fr) Croatia (Neth) St. Lucia Bosnia and 10–29 Aruba Bonaire St. Vincent and San Herz. Serbia Bulgaria (Neth) (Neth) The Grenadines Barbados Marino 6–9 Montenegro Kosovo Grenada FYR ≤5 Italy Macedonia Trinidad Vatican Albania R.B. de Venezuela and Tobago City Greece no data MDG performance Greenland (Den) Faeroe Norway Iceland Islands Finland (Den) Sweden The Netherlands Russian Federation Estonia Canada Isle of Man (UK) Denmark Russian Latvia Fed. Lithuania United Ireland Kingdom Germany Poland Belarus Channel Islands (UK) Belgium Ukraine Luxembourg Moldova Kazakhstan Liechtenstein Romania Mongolia Switzerland France Italy Andorra Bulgaria Georgia Azer- Uzbekistan Kyrgyz D.P.R. Turkey Armenia baijan Turkmenistan Rep. United States Portugal Spain Monaco Tajikistan of Korea Japan Greece Syrian Gibraltar (UK) Cyprus Rep. of Tunisia Malta Lebanon Arab Islamic Rep. Afghanistan China Korea Israel Rep. Iraq of Iran Bermuda Morocco Kuwait (UK) West Bank and Gaza Jordan Pakistan Bhutan Bahrain Nepal Former Algeria Libya Arab Rep. Qatar The Bahamas Spanish of Egypt Saudi Cayman Is. (UK) Sahara Arabia United Arab Bangladesh Cuba Turks and Caicos Islands (UK) India Mexico Emirates Myanmar Lao Haiti Mauritania Oman P.D.R. Belize Jamaica Cape Verde Mali N. Mariana Islands (US) Niger Rep. of Thailand Vietnam Guatemala Honduras Senegal Chad Eritrea Yemen Guam (US) El Salvador Nicaragua The Gambia Burkina Sudan Marshall Faso Djibouti Cambodia Philippines Islands Guinea-Bissau Guinea Benin Federated States of Costa Rica Panama Sri Micronesia R.B. de Guyana Sierra Leone Côte Ghana Nigeria Central Ethiopia Lanka Venezuela Suriname Liberia D’Ivoire African Rep. Brunei Cameroon Somalia Palau Colombia French Guiana (Fr) Togo Malaysia Equatorial Guinea Uganda Maldives Ecuador São Tomé and Príncipe Gabon Congo Rwanda Kenya Singapore Nauru Dem. Rep. of Burundi Seychelles Papua Solomon Congo Indonesia New Guinea Islands Tanzania Comoros Peru Brazil Timor-Leste Angola Malawi Mayotte Zambia (Fr) Vanuatu Bolivia Zimbabwe Mauritius Namibia Madagascar Paraguay Botswana Mozambique Réunion (Fr) New Caledonia Australia (Fr) Swaziland South Lesotho Africa Chile Uruguay Argentina MDG performance New Dominican Puerto British Virgin Poland MDG 7.C Halve by 2015 the proportion of Zealand Germany Republic Rico (US) Islands (UK) people without sustainable access to Ukraine Anguilla (UK) Czech Rep. U.S. Virgin Islands Antigua and Barbuda Slovak Rep. basic sanitation (US) Austria St. Maarten (Neth) St. Kitts Guadeloupe (Fr) Hungary and Nevis Slovenia on target Dominica Romania Curaçao Montserrat (UK) Martinique (Fr) Croatia close to target (Neth) Bosnia and Bonaire St. Vincent and St. Lucia San Herz. Serbia Aruba far from target Bulgaria (Neth) (Neth) The Grenadines Barbados Marino This map was produced by the Map Design Unit of The World Bank. Montenegro Kosovo no data The boundaries, colors, denominations and any other information Grenada FYR Italy Macedonia shown on this map do not imply, on the part of The World Bank Trinidad Vatican Albania Group, any judgment on the legal status of any territory, or any and Tobago City Greece R.B. de Venezuela endorsement or acceptance of such boundaries. Source: World Bank staff calculations based on data from the World Development Indicators database. 142 PROGRESS IN THE DEVELOPMENT AND TRADING FRAMEWORK GLOBAL MONITORING REPORT 2011 FIGURE 5.7 Local-to-global connectivity Subnational logistics Logistics activities value addition ➤ consolidation local collection low-capacity vehicles value addition district district ➤ consolidation hub hub shared storage provincial ➤ shared storage hub packaging logistics exchanges International corridor logistics national trucking checkpoints border trucking transit checkpoints gateway hub rail crossing rail national transit international transit Source: Kunaka 2010. became especially relevant during the global prices actually rose postcrisis, the evidence is financial crisis in 2008–09, when higher lend- consistent with a negative supply shock that ing costs, higher risk premiums, and liquidity might be associated with credit constraints.46 pressures caused by scarcity of capital led to The trade collapse was mostly a result of the a sudden shortage in trade finance. spillover of the financial crisis to the real Lack of affordable trade finance has prob- economy and lower activity and inventory ably constrained the trade activities of small destocking. Nevertheless, the lack of afford- and medium enterprises (SMEs), particularly able trade fi nance likely constrained trading in low-income countries. Results from finan- activities of SMEs, especially those based in cial market and firm surveys (undertaken low-income countries that have underdevel- during the crisis by the IMF, the International oped financial systems and banks that could Chamber of Commerce, and the World Bank not fulfi ll the counterparty criteria of over- to overcome the lack of data on trade finance) seas banks. Postcrisis surveys and data on and postcrisis empirical analyses all indicate trade finance indicate signs of improvement. the prevalence of tighter trade finance condi- The institutional response in providing tions during the crisis and significant adverse trade finance during the crisis was timely effects on trade flows. The lack of data can and substantial. In the midst of the crisis, the hamper clear conclusions. Although survey international community responded swiftly— results indicate that the shortfall in trade spearheaded by the G-20—in committing finance has contributed to the sharp drop in $250 billion over the course of two years global trade flows during the crisis, it seems in funding for cofinancing arrangements that it played a moderate role in one study.45 that support trade transactions. This was In manufacturing, where volumes fell but implemented through a partnership between GLOBAL MONITORING REPORT 2011 PROGRESS IN THE DEVELOPMENT AND TRADING FRAMEWORK 143 development banks, export credit agencies, the request of the World Bank and the WTO, foreign commercial banks, private insurance the G-20 will take stock of the situation at its underwriters, and investment funds. Whereas 2011 meeting. the G-20 support was mainly directed at large banks and international banking institutions, Renewed support for effective aid for the World Bank’s International Finance Cor- trade poration (IFC) and the regional development banks stepped in as well to target their efforts The G-20 has renewed its commitment to sup- at smaller banks and banks in developing port LDC trade capacity building through aid countries. With the continued uncertainty in for trade. At the November 2010 Seoul meet- trade fi nance markets, particularly for low- ing, the G-20 pledged to maintain aid for trade income countries and small firms, govern- commitments in 2011 and beyond, despite fis- ments and international organizations need cal and budgetary constraints. Launched at the to be cautious of the timing and pace of with- G-8 meeting in Gleneagles and the follow-on drawal of such trade finance programs. WTO Ministerial Conference in Hong Kong Despite recent positive developments, SAR, China, in 2005, aid for trade aims to concerns about trade fi nance have not com- facilitate the integration of developing coun- pletely dissipated. At the Seoul G-20 meeting tries (particularly the LDCs) into the global in November 2010, the international com- economy by lowering cumbersome trade barri- munity expressed particular concern about ers, reducing transaction costs, and enhancing low-income countries that may still be facing trade competitiveness. Aid for trade remains severe difficulties in accessing trade fi nance an effective way to support LDCs to improve at affordable cost. These concerns are exac- supply-side capacities and reap the benefits of erbated by the lack of available data for trade increased market access.47 finance, which has constrained the ability to After a sharp increase in donor aid for properly monitor and evaluate existing trade trade commitments in 2007 and 2008, the finance flows and to measure their impact latest OECD data indicate that new com- on trade flows during and after the crisis. mitments in 2009 plateaued at $39.5 billion The building up of the Trade Registry by the (2008 constant prices). In 2009, the World International Chamber of Commerce is a sig- Bank’s concessional arm, IDA, was the larg- nificant step forward because it will create a est provider of aid for trade, accounting for living database of the trade fi nance market about 21 percent of the total. Japan and and may help demonstrate the resilience of the United States were the largest bilateral the trade finance business. donors, providing 14 percent and 12 percent Other concerns relate to the new Basel of total aid for trade, respectively. Africa regulations, which are viewed as potentially was, for the fi rst time, the largest recipient constraining the supply of trade finance. of aid for trade (41 percent); and the level Banks argue that the increase in the new of commitments increased over those of liquidity and capital prudential requirements, 2008. Aid for trade in Africa was mainly and the nonrecognition of trade assets as a directed to Nigeria, Uganda, Kenya, and highly liquid and safe asset, would lead to a Ethiopia. Asia was the second-largest recipi- significant increase in banks’ cost in provid- ent of aid for trade (39 percent), with Viet- ing trade fi nance—a situation that will lead nam, India, and Afghanistan the major to lower supply, higher prices, or both. Regu- recipients. Aid for trade mainly supported lators have insisted that the Basel II and III projects involving trade infrastructure and increase in capital for trade finance exposures capacity building (particularly trade devel- is not any greater than for other exposures. In opment programs), while technical assis- light of these different views, the Basel Com- tance for institutional capacity building and mittee established a working group to study training for trade regulations have been on the impact of regulations on trade finance. At the decline (figure 5.8). 144 PROGRESS IN THE DEVELOPMENT AND TRADING FRAMEWORK GLOBAL MONITORING REPORT 2011 FIGURE 5.8 Regional trends in aid for trade, 2002–09 International financial institutions’ support for 20,000 50,000 development The IFIs are adapting their policies, proce- 16,000 40,000 dures, and tools to country heterogeneous circumstances and needs.48 They strive to US$ millions US$ millions 12,000 30,000 provide all member-countries with devel- opment solutions appropriate to meet their 8,000 20,000 development needs and capacity by using a 10,000 mix of lending policies, fi nancial facilities, 4,000 and knowledge services, and by supporting 0 0 the global public goods agenda. Moreover, 2002 2003 2004 2005 2006 2007 2007 2009 the financial crisis, and the food and fuels cri- (left axis) Africa Asia America sis preceding it, have shown the crucial role Middle East Europe Oceania of IFIs in supporting responses to crises and all developing countries (right axis) development emergencies, swiftly shifting toward a more countercyclical approach to mitigate social and economic impacts of the Source: OECD, Aid for Trade Statistical Queries 2010, http://www.oecd.org/document/21/0,3746,en_ 2649_34665_43230357_1_1_1_1,00.html. crises. The response to recent crises provided a snapshot of the untapped potential of the Besides concerted efforts to channel aid IFIs to address the challenges and opportuni- for trade to developing countries, the interna- ties of the postcrisis world. It also provided a tional community is exploring the effective- glimpse of how transformed IFIs will work ness of aid for trade. At the initiative of the on an ongoing rather than emergency basis. WTO and the OECD, donors are currently In recognition of the changing development collecting case studies that illustrate the landscape and country development needs, impact of trade-related assistance and lending the IFIs have embarked on comprehensive programs and the lessons learned from these modernization agendas involving actions and projects. These case studies will be discussed reforms to enhance financial capacity, trans- at the July 2011 WTO/OECD Aid for Trade form governance, reform organization, and meeting. Moreover, with the emergence of a refresh and refi ne their priorities. Efforts to multipolar world with rising economic pow- bolster their performance, ensuring relevance ers and multinational private sector firms, and results focus on country strategies and the quality of delivery of aid for trade has operations in the postcrisis world (including become more critical. This new paradigm helping countries develop their own capacity calls for more public-private partnerships to to manage for results) are key objectives of strengthen trade capacity-building assistance the ongoing reform agendas and the driving delivered to LDCs. The IFIs can focus more force in the development of the results-based on leveraging partnerships between govern- approaches that will guide IFIs’ actions in the ment agencies and private sector firms. By future. combining the policy experience of the gov- ernment and the ingenuity of the private sec- The evolution of assistance policies and tor, trade development projects may have the resources potential to be more far-reaching and effec- tive. A good illustration of such programs is Given developing countries’ diverse needs and the deployment of modern information and heterogeneous characteristics, the approach of communications technologies from the pri- the multilateral development banks (MDBs) vate sector to streamline trade facilitation has increasingly been tailored to country con- and minimize at-the-border barriers. ditions and grounded in national strategies GLOBAL MONITORING REPORT 2011 PROGRESS IN THE DEVELOPMENT AND TRADING FRAMEWORK 145 setting out countries’ development visions, expanded over the last decade. For instance, objectives, and priorities.49 All of the major the World Bank has enhanced its financial MDBs now formulate country strategy docu- support to provide increased access to highly ments, generally, every two to four years, concessional resources for small IDA coun- with more frequent progress reports and inde- tries. Two measures were agreed to strengthen pendent evaluations. At the World Bank, the IDA support for small states during the IDA16 Country Assistance Strategy has evolved over period: fi rst, to eliminate the maximum per the past decade to adopt an explicit results capita allocation ceiling, currently set at spe- framework that specifies the expected links cial drawing right (SDR) 19.8, which has con- between the World Bank’s interventions and strained the allocations of several small states; long-term development goals, along with and, second, to raise the base allocation for measurable indicators to monitor results.50 small states from SDR1.5 million to SDR3.0 Donors have stepped up financial resources million per year. This second measure would to low-income countries through the MDBs’ result in a substantial increase of IDA resources affiliates that provide grants and no- (or for small states with small populations. low-) interest credits. Disbursements by the IDA also has continued its efforts to help World Bank’s IDA, the African Development countries recover from natural disasters. Fund, the Asian Development Fund, the Inter- Following the 2009 tsunami, IDA was able American Development Bank (IADB)’s Fund to provide significant additional resources for Special Operations, and the European to both Samoa and Tonga. Following the Bank for Reconstruction and Development earthquake in Haiti, the World Bank pledged (EBRD)’s Early Transition Countries Initia- $479 million toward Haiti’s reconstruction tive (under which the EBRD accepts higher for the first 24 months after the earthquake, risks in projects in low-income members) have two-thirds of which was delivered in the steadily risen over the years. Strong pledges fi rst 12 months. The portfolio in Haiti was from both traditional and new donors in the restructured toward meeting pressing post- last IDA replenishment recognize that restor- earthquake priorities; and, by end-2010, the ing momentum to the MDGs leading to 2015 WBG had provided $340.0 million to Haiti, requires ambitious efforts to deliver on the including $139.5 in new grants, $129.0 mil- economic growth and access agenda for basic lion in disbursements, debt cancellation of services (health, education, and basic infra- $39.0 million (May 2010), and investments of structure). A total of 51 donors endorsed a $32.5 million from the IFC (box 5.3). Simi- $49.3 billion funding package for the next larly, the IMF established a Post-Catastrophe cycle of IDA (IDA16, which covers July 2011 Debt Relief Trust that allows the fund to to June 2014), an increase of 18 percent from join international debt relief efforts for very the previous round. At the African Develop- poor countries hit by the most catastrophic of ment Bank (AfDB), the level of concessional natural disasters. Such debt relief is intended aid resources was similarly maintained as to free up additional resources to meet excep- a reflection of donor commitment to the tional balance-of-payments needs created by achievement of MDGs by 2015. This com- the disaster and the recovery, complementing mitment was similarly demonstrated during fresh donor financing and the fund’s conces- the 12th General Replenishment of the Afri- sional liquidity support. The Post-Catastrophe can Development Fund, whereby negotiations Debt Relief Trust Fund provided timely assis- with 27 donor countries resulted in a 10.6 per- tance to Haiti equivalent to around $268 mil- cent increase from the previous replenishment lion in fund-financed debt stock relief, elimi- cycle and a funding package of approximately nating Haiti’s entire outstanding debt to the $9.5 billion to support the fund’s work in low- IMF. This decision adds to the $1.2 billion of income countries in Africa between January debt relief delivered to Haiti by the IMF and 2011 and December 2013. MDBs’ operational other fi nancial organizations in June 2009, and financial approach to address specific under the Heavily Indebted Poor Countries challenges or vulnerabilities has evolved and and Multilateral Debt Relief initiatives. 146 PROGRESS IN THE DEVELOPMENT AND TRADING FRAMEWORK GLOBAL MONITORING REPORT 2011 BOX 5.3 Special reconstruction allocation for Haiti The earthquake that struck Haiti on January 12, 2010, WBG disbursements and debt relief. The World was a human tragedy. More than 230,000 people per- Bank portfolio currently includes 16 active projects ished, 300,000 more were wounded, and well more for a total of $337 million. The Bank has also pro- than a million were displaced. The earthquake ravaged vided $42.5 million in budget support to the govern- cities including Port-au-Prince, the capital, destroying ment of Haiti since January 2010. IDA, the Bank’s whole neighborhoods, wiping away roads, collaps- fund for the poorest countries, has disbursed more ing public buildings, and damaging businesses. The than $11 million a month, on average, since the earth- disaster struck the country’s political, economic, and quake. Following the earthquake, the Haiti portfolio administrative nerve center, where an estimated 65 was restructured to direct resources toward urgent percent of GDP and 85 percent of government revenues post-earthquake priorities through the reallocation were generated. A year later, conditions remain diffi- of funds and the modification of activities to adapt cult for the Haitian population. About 1 million people to institutional capacity changes. In May 2010, the continue to live in tent camps and depend on aid orga- Bank was able to cancel Haiti’s remaining $39 million nizations for water, food, sanitation, health, and edu- debt to IDA, thanks to contributions by IDA. cation services. Furthermore, on October 21, 2010, an New WBG programs. In early 2011, the WBG outbreak of cholera was confirmed. The latest United will launch the following programs and projects from Nations data indicate that about 149,000 cases have newly approved resources: $95 million for a neighbor- been reported and more than 3,000 people have died. hood upgrading and housing reconstruction program International response. Innumerable public and financed by IDA and the HRF; $15 million for an private organizations were mobilized in an unprec- emergency cholera project; $11 million for a line item edented outpouring of international support after the budget support operation to cover specific expendi- earthquake. At the International Conference on Haiti tures in education and agriculture; and $3 million to in March 2010, $9 billion was pledged ($5.3 billion establish a $35 million Partial Risk Guarantee Fund of which was for 2010–11 to support the Government for private operators in Haiti, cofi nanced by the Inter- Action Plan for Reconstruction and National Devel- American Development Bank (IADB), the U.S. Trea- opment. At the end of 2010, the international com- sury, and the HRF. munity had delivered more than half of its pledge for Support for the private sector. Since the earth- the fi rst 24 months: $1 billion in debt relief had been quake, the WBG’s IFC has approved five projects provided and $2.7 billion approved for projects and for a total of $49.6 million. Three of these projects programs, with $1.2 billion of these funds spent. (worth $15.3 million) in the garments, hotels, energy, Haiti Reconstruction Fund. At the request of the and mining industries, are already under implementa- government of Haiti, in April 2010, the World Bank tion, in addition to a trade finance guarantee facility, set up the Haiti Reconstruction Fund (HRF), a multi- an equity investment in a local bank to expand access donor trust fund. The World Bank serves as a trustee to fi nance to SMEs, and a 30-megawatt energy proj- for the HRF. In March 2010, donors pledged more ect approved prior to the quake. IFC advisory services than $500 million to the fund. To date, 12 donors have also ramped up operations to foster a better invest- confirmed their pledges for a total of $267 million, all ment climate, improve access to finance, and develop of which has been received. Since June 2010, the HRF management skills for more than 600 small entrepre- has allocated $193 million for reconstruction. neurs. The IFC’s combined investment and advisory WBG support. The World Bank Group pledged projects are supporting the creation of 5,000 new $479 million toward Haiti’s reconstruction for the jobs and safeguarding 5,000 existing jobs. The IFC fi rst 24 months after the earthquake, two thirds of also completed the structuring of the international which was delivered in the first 12 months. The port- bidding process for TELECO, which is bringing the folio in Haiti was restructured toward meeting press- country’s largest foreign direct investment since the ing post-earthquake priorities; and, by end-2010 the earthquake—a $100 million investment by Vietnam’s WBG had provided $340.0 million to Haiti, includ- biggest mobile telephone operator, Viettel—to expand ing $139.5 in new grants, $129.0 million in disburse- telecommunications services in Haiti. ments, debt cancellation of $39.0 million (May 2010), and investments of $32.5 million from the IFC. GLOBAL MONITORING REPORT 2011 PROGRESS IN THE DEVELOPMENT AND TRADING FRAMEWORK 147 The approach to address the specific chal- policies on debt relief (through the joint World lenges of fragile and confl ict-affected coun- Bank-IMF Heavily Indebted Poor Countries tries has also increasingly evolved over the and Multilateral Debt Relief initiatives) and last decade. In terms of resources, eligible from grants allocated through IDA’s grant postconflict and reengaging countries receive allocation framework to help them avoid the exceptional allocations, and some have ben- reemergence of debt problems. efited from exceptional pre-arrears clearance Although all fragile states are characterized grants and allocations to help clear arrears. by weak policies and institutions, country The IDA phase-out period for exceptional context varies considerably and operational allocations, for instance, was increased from approaches must be carefully calibrated to 3 to 6 years, resulting in 10 years of support take this into account. The World Bank will for postconfl ict countries, which otherwise incorporate the lessons of the 2011 World would have exited the window during IDA15. Development Report (box 5.4)—a report For the IDA16 period, postconflict countries focused on conflict, security, and devel- will also be eligible for an extension of their opment challenges in fragile and conflict- phase-out period on a case-by-case basis, pro- affected countries—into the revision of its pol- vided that they meet a predetermined set of icies. Moreover, the World Bank committed criteria. Fragile and confl ict-affected coun- in the IDA16 replenishment to develop plans tries have also benefited significantly from to enhance the implementation of United BOX 5.4 y World Development Report 2011: Conflict, Security, and Development Repeated cycles of political and criminal violence external stresses. States or subnational governments have left more than 440 million people in poverty. do not provide protection and access to justice. Mar- Strengthening the institutions that provide justice, kets do not provide legal employment opportunities. jobs, and citizen security is crucial to break these And communities have lost the social cohesion to cycles. Restoring confidence and transforming secu- contain conflict. rity, justice, and economic institutions is possible What works in breaking repeated cycles of vio- within a generation. But that requires determined lence? The key differences between situations of national leadership and an international system violence and stable developing environments are the “refitted” to address 21st-century risks of violence: need to restore confidence before undertaking wider refocusing assistance on preventing criminal and institutional reforms; the priority given to basic insti- political violence, reforming the procedures of tutional functions of security, justice, and provision international agencies, and renewing cooperative of employment; and the role of regional and interna- efforts between lower-, middle-, and higher-income tional action to contain external stresses. Pragmatic, countries. responsive, and accountable leadership at national Why do some parts of the world face repeated and local levels also plays a crucial role. cycles of violence? Violence can be spurred by a What policy tools make a difference, and how can range of internal and external stresses—infi ltration they be adapted to country circumstances? The pro- of trafficking networks and foreign fighting forces; grams listed in the table on the next page have been youth unemployment; economic shocks; and ten- used successfully in different country circumstances. sions between ethnic, regional, or religious groups. Few countries have combined all of them, and most But other areas of the world, including some very have faced challenges at some point either on fast poor countries, have managed and contained similar confidence-building or on longer-term institutional pressures. The common “missing factor” explaining transformation. The toolbox presented here thus pro- repeated cycles of violence is that state and societal vides a mechanism to learn from both successes and institutions are too weak to manage high internal and failures. continued 148 PROGRESS IN THE DEVELOPMENT AND TRADING FRAMEWORK GLOBAL MONITORING REPORT 2011 y BOX 5.4 World Development Report 2011: Conflict, Security, and Development (continued) A policy toolbox Tools Politics and justice Security Jobs and services Restoring and building Political action to build national Military and police Multisectoral community confidence and international stakeholder doctrine and planning programs, combining security support and manage expectations: adapted to citizen operations with action on Forming inclusive-enough strategic communications, security goals employment, local justice, coalitions appointments, decision-making infrastructure, education, and processes Designed to deliver community empowerment Delivering early results benefits in ease of Policy decisions and realistic production, trade, Regulatory simplification and timelines to political reform, workplace attendance, recovery of basic infrastructure decentralization, and transitional access to services, and transit facilities to remove justice where relevant human rights principles, bottlenecks to private sector job and simple mechanisms creation Extended reach and to redress citizen responsiveness of local justice complaints Humanitarian delivery to support services improvements in intercommunity and state-community relations All areas: transparency and willingness to consider independent monitoring by independent or multisectoral government agencies to lock in the credibility of changes Transforming Justice sector reform: focus on Security sector reform: Employment policy: mix of long- institutions basic functions and use of state- capacity increases linked term public programs at the community and state-traditional to repeated realistic community level, employability Prioritizing and systems for justice performance outcomes; programs, value chain approaches, sequencing institutional vetting and civilian infrastructure, informal sector reform to underpin Phased anticorruption measures to oversight support, and labor migration security, justice, and jobs first show that national resources agreements can be used for the public good; Traditional low-capital Taking pragmatic then to combat grand corruption; systems for rural and Temporary options for second- approaches—first-best and then to roll back minor community policing best service delivery: high service in the political context, corruption coverage, high unit cost, under second-best in technical phase-out agreements; planned terms transitions from international humanitarian support Global and regional Platforms to coordinate supply- Supplementary capacity Pooled supplementary approaches for external and demand-side responses and over-the-horizon administrative capacity stresses that national between producing and guarantees actors can initiate consuming countries Cross-border development programming Joint investigations and prosecutions across jurisdictions Source: World Bank 2011. GLOBAL MONITORING REPORT 2011 PROGRESS IN THE DEVELOPMENT AND TRADING FRAMEWORK 149 Nations–World Bank partnership agreements, important refinements to the development evaluate IDA’s work in fragile and confl ict- policy lending (DPL) instrument were made affected countries during 2012–13, strengthen at the World Bank immediately prior to collaboration with partners on multidonor and during the crisis. First, the World Bank trust funds administered by the World Bank, adopted enhancements for DPL with a and revise the postconflict performance indi- deferred drawdown option (DDO) that allows cators and publicly disclose them before the International Bank for Reconstruction and start of IDA16. Development (IBRD) borrowers to postpone Like IDA’s special assistance, the African disbursement of a loan for a defi ned period, Development Fund’s postconflict enhance- instead of drawing down funds immediately ment factor enables countries to benefit from after approval. Second, the World Bank additional resources beyond their perfor- introduced a DPL instrument that provides mance-based allocation for a limited period liquidity immediately after a natural disas- after they are designated postconfl ict coun- ter (the CAT DDO) to support a disaster risk tries. The AfDB has established the Post- management program, assuming a state of Conflict Country Facility to help countries emergency is declared. And, third, the Bank emerging from conflict clear these arrears on adopted reforms for the Special DPL, avail- their debt. Similarly, the ADB has deployed able to countries that are approaching or are innovative means to strengthen the effective- in a macroeconomic crisis, contingent on a ness of country-led models of engagement disbursing IMF program being in place. Since in the complex environments of fragile and April 2008, the board has approved 12 DPL conflict-affected situations. In particular, DDOs for a total of $5.6 billion, and $4.8 ADB has sought to sustain its commitments billion of that amount had been disbursed by for longer periods; partner closely with other the middle of fiscal 2011. Several CAT DDOs funding agencies; and pursue deeper, more for countries vulnerable to natural catastro- flexible, and longer engagements in capac- phes have since been approved, totaling $400 ity and institutional development. The IADB million. Moreover, in recognition of the vul- used special measures for its engagement in nerabilities that low-income countries face Haiti—including simplified startup require- in terms of both financial crises and natural ments, broader eligible expenditure catego- disasters, the World Bank and the IDA depu- ries, and elimination of counterpart financing ties have recently agreed to establish a dedi- requirements. In 2004, the EBRD launched cated Crisis Response Window (CRW) within the Early Transition Countries Initiative to IDA to enhance its capacity to respond to increase its impact in the region’s lowest- crises. The CRW will provide IDA countries income and least-advanced transition coun- with timely access to additional resources to tries51 by developing specific financial facili- respond to the impact of severe economic cri- ties targeted at local enterprises, including the ses or natural disasters; and it will strengthen separate facilities to fund equity investments, IDA’s capacity to respond rapidly to such directly finance long-term loans, cofinance crises in collaboration with other agencies, loans with local commercial banks, and lend development banks, and donors. Resources to microfinance institutions. for the CRW have been capped at 5 percent of total IDA16 resources, and an amount of SDR1,335 million has been set aside to Development finance finance expenditures under the CRW. Over the years, the MDBs have expanded the Finally, a major effort is also under way range of their instruments and modernized to reform the World Bank’s investment lend- them to make them more responsive, results ing model to focus on results and risks, with oriented, and amenable to an enhanced busi- the goal of improving its responsiveness to ness model and country needs. For example, borrowers’ needs and to a changing global 150 PROGRESS IN THE DEVELOPMENT AND TRADING FRAMEWORK GLOBAL MONITORING REPORT 2011 environment.52 Guarantees are currently one Fund; and enhanced Policy Advisory Support of three development finance instruments to support countries where advisory capacity (along with DPLs and investment lending) remains challenging. that the World Bank offers, and a number The EBRD has also expanded the exist- of actions are being put in place to improve ing pool of financial instruments avail- the attractiveness of the guarantee prod- able to the private sector in its countries of uct and encourage country teams and client operation. This includes the introduction governments to actively consider guarantee of various facilities and frameworks aimed operations. at different aspects of private sector devel- Similarly, the ADB has introduced new and opment—regional energy efficiency, direct innovative instruments to fi nance large indi- mid-size lending, and regional micro, small, vidual projects and longer-term investment and medium enterprises lending. The crisis programs. The following instruments expand response packages introduced in 2008 are the options available in ADB’s development good examples of the EBRD’s reaction to the fi nance portfolio while preserving its credit changing economic environment in its coun- standing: Multitranche Financing Facility, tries of operation. 54 Similarly, the IFC also which allows the ADB to finance a long-term undertook new initiatives to support private investment program based on a sector assess- sector activities in the wake of the crisis by ment (road map) and organized in a series providing liquidity support, rebuilding finan- of funding blocks, or tranches; Nonsover- cial infrastructure and mobilizing investment eign Public Sector Financing Facility, which in distressed asset pools, and providing advi- provides debt finance (loans and guarantees) sory services to enhance the efficiency of dis- directly to selected nonsovereign public sec- tressed asset markets.55 tor entities without a central government (sovereign) guarantee; Local Currency Loan Knowledge services Product, which allows the ADB to offer loans denominated in a local currency to reduce The IFIs are also increasingly providing more the mismatch between income received in a flexible and ad hoc analysis, research, global domestic currency and debt repayments in a public goods in the form of databases and foreign currency; and new forms of cofinanc- free data, advice, and technical assistance. ing through active financial syndications and In low-income countries and some lower- risk sharing with commercial financing part- middle-income countries, knowledge services ners that have since been mainstreamed into are often bundled with lending, which is seen the ADB’s operations.53 as a tool of public policy for designing and The global financial crisis undoubt- implementing reform/knowledge agendas. edly also had a lasting effect on the AfDB’s More sophisticated middle-income countries lending. In March 2009, the AfDB’s board are mainly interested in obtaining stand-alone endorsed four initiatives aimed at better knowledge support in cutting-edge areas. responding to the needs of its client countries: Between these two extremes, there is a vari- an Emergency Liquidity Facility to assist eli- ety of hybrid requests that are accommodated gible AfDB regional member-countries and by an increased flexibility in the delivery nonsovereign operations suffering from lack modalities. For instance, the EBRD’s unique of liquidity ($1.5 billion); a Trade Finance mandate of fostering transition toward open Initiative ($1 billion) that comprises a line market–oriented economies has shaped its of credit ($500 million) and a Global Trade specific role in knowledge services. Techni- Liquidity Program ($500 million) to allow cal cooperation provided to clients (including eligible African commercial banks to sup- through donor programs) is aimed at differ- port trade finance operations; an accelerated ent “transition-oriented” aspects of an opera- replenishment process for the 12th General tion, such as development of management Replenishment of the African Development skills, legal advice to promote regulatory GLOBAL MONITORING REPORT 2011 PROGRESS IN THE DEVELOPMENT AND TRADING FRAMEWORK 151 development and improve legislation and around selected strategically significant corporate governance, and skills transfer. issues. For instance, the World Bank knowl- Another example of knowledge services is the edge platform initiative focuses resources— EBRD’s Turn Around Management and Busi- both within the World Bank and across the ness Advisory Services programs, which have development community—on strategic, and been valuable instruments for the promotion cross-sectoral transformational issues to fill of good management in the SME sector, pro- critical knowledge gaps and seek cogeneration viding consultancy advice at the enterprise of knowledge from diverse sources and insti- level. Turn Around Management supports tutions. The first three knowledge platforms the introduction of international best practice are becoming operational, and they focus SME enterprises, while the Business Advi- on urbanization, the e-Transform Initiative sory Services Program acts as a facilitator (information and communications technol- for the use of local, private sector consultants ogy for accountability and development); and by micro, small, and medium enterprises to green development. The World Bank also is obtain a diverse array of services. taking steps to improve access to its data and Increasingly, cutting-edge knowledge is research (box 5.5). generated in the South through the countries’ own development experiences. And as policy Global programs and partnerships analysis and development research are increas- ingly conducted by countries themselves, the The MDBs are stepping up their efforts to IFIs can play a vital role both as generators address common social sector and environ- of research and as providers of a forum for mental challenges through global programs bringing together knowledge from both the and partnerships. World Bank support for North and the South—acting as a global con- global programs began three decades ago, nector of practitioner knowledge, a broker with the establishment of the Consultative of development, and a facilitator of capacity Group on International Agricultural Research development and client learning. The World that was recently restructured. Global pro- Bank, for instance, has established partner- grams are now reflected in World Bank cor- ships with researchers and analysts in devel- porate strategy papers and operational activi- oping countries: the China Center for Eco- ties, with more than 100 programs (managed nomic Research, the New Economic School by either the World Bank or external recipi- in Moscow, the Economic Research Forum in ents). Recent initiatives include the World Cairo, the Global Development Network in Bank’s role in the Climate Investment Funds, collaboration with 11 regional partners and including the Clean Technology Fund and headquartered in New Delhi, and others. A Special Climate Change Fund, which pro- South-South Experience Exchange Facility vide financing for projects to address climate has been set up to help developing countries change (box 5.6). The Pilot Program on Cli- share expertise. 56 So far, that facility has mate Resilience helps highly vulnerable coun- received 88 proposals to support South-South tries integrate risk and resilience into core activities in World Bank products and services development planning; the Forest Investment in more than 40 countries. Program supports efforts to reduce defores- As MDBs strengthen their role as provid- tation and forest degradation, cut emissions, ers of global public goods (such as data and and maintain carbon reservoirs; and the Scal- research) and their role in linking country ing Up Renewable Energy Program aims to practitioners and policy makers to sources and demonstrate low-carbon energy development centers of knowledge and innovation dispersed in low-income countries. around the world, knowledge platforms that The regional banks are also involved involve external stakeholders in coproduc- in global and regional programs covering ing knowledge have been set up to provide a fi nancial stability, trade, environment, post- framework for sustained global collaboration confl ict assistance, and knowledge; and all 152 PROGRESS IN THE DEVELOPMENT AND TRADING FRAMEWORK GLOBAL MONITORING REPORT 2011 BOX 5.5 Open data, open knowledge, and open solutions User expectations for public information are chang- strategic themes: economic transformations around ing rapidly; and clients and other stakeholders expect emerging issues, such as macroeconomic growth, easy access to World Bank–generated data, knowl- agriculture, and rural development, urbanization, edge, and solutions. Open access is meant to increase and green development; the broadening of oppor- the development impact of the World Bank’s informa- tunities by trade and integration and access to tion. Three areas are receiving top priority: finance; the understanding of risk and vulnerability across countries and sectors; and the measuring of • Open data—In April 2010, the World Bank results and aid effectiveness. released its development data free of charge. The • Open solutions—To engage in the “wholesaling of Open Data work builds on the success of this initia- research,” the World Bank plans a greater focus on tive. Its priorities are to expand the Web site and the creation of software tools, training products, create a data resource center to serve as a clearing- and researcher communities. Significant efforts are house for researchers, to develop new applications under way to improve and to broaden the focus of to enable easy access to data across platforms and such existing tools as ADePT and iSimulate. The devices, and to strengthen capacity-building activi- programs are expected to be collaborative efforts ties in developing countries to help ensure the qual- within the World Bank and with technology part- ity of data. ners (such as Google and Microsoft) and multilat- • Open knowledge—With a shift in the focus of eral organizations (such as the United Nations and research teams, there is now an emphasis on four the IMF). but the EBRD are involved in the control areas in the provision of regional and global of infectious diseases. In many cases, the public goods—fi nancial sector assessments, banks are focused on regional public goods regional integration, curbing of infectious (RPGs) or on regional aspects of global pub- diseases, promotion of environmental ser- lic goods, looking to the World Bank on vices, and support for research in agriculture the global aspects. They also are involved and regional policy dialogue. It has prepared in helping their regional clients build coun- a new policy framework for its support for try capacity to meet requirements under RPGs, including a fi nancing facility geared global agreements. For the AfDB, critical to providing grant fi nancing for what it calls issues are postconflict assistance and health, “early-stage RPGs” (where dialogue among especially in the face of the HIV/AIDS epi- countries is needed), “later-stage RPGs” demic. For the ADB, key issues are the (where larger institutional resources to man- environment, health, and knowledge, with age the emerging program are needed), and a particular focus on those issues where the initial stages of “club RPGs” (which will there are spillover effects within the region likely be fi nancially self-sustaining once they or within the ADB’s subregional coverage. are up and running). For the EBRD, nuclear safety is an area of The regional development banks are also special focus, where the World Bank has stepping up their efforts to promote regional the international lead in supporting transi- cooperation and integration. The AfDB sup- tion countries in decommissioning capacity ports the New Partnership for Africa’s Devel- and resolving other environmental liabilities opment Secretariat, the African Union, the from the Soviet era. Another focus area is Global Environment Facility (especially on fi nancial stability, especially the adoption of the development of the Environmental Action the standards and codes underpinning mar- Plan for Africa), and the Africa Regional ket economies. The IADB has five priority Coordination Unit for the United Nations GLOBAL MONITORING REPORT 2011 PROGRESS IN THE DEVELOPMENT AND TRADING FRAMEWORK 153 BOX 5.6 Climate change, natural disasters, and the World Bank Group To address the threat of climate change, develop- Policy Credit supports efforts to integrate climate risk ment needs to become “climate smart.” It must deal management into agriculture-led growth. In Bangla- with adaptation efficiently and mitigate the growth desh, 320,000 homes have been provided with solar of greenhouse gases. This will require additional electricity, and thus with lighting for education and resources, beyond existing development finance. Syn- for women’s economic empowerment. The IFC is sup- ergies between climate-related investment and devel- porting climate-related investments that help meet the opment need to be fully exploited. Adaptation pro- needs of the poor for modern services in an efficient grams can be pro-poor, such as converting degraded manner, such as the Lighting Africa Program. cropland into resilient agroforestry systems. Natural disasters are another area where the WBG The WBG is helping countries address climate is increasingly focusing its attention. The Global Facil- change through projects and partnerships. A key ity for Disaster Recovery and Reconstruction has partnership is the Climate Investment Funds (estab- established a Disaster Risk Financing and Insurance lished in 2008) to which contributing countries have Program to enhance capacity building and knowledge pledged more than $6 billion for climate-resilient and sharing on disaster risk financing and to mainstream low-emissions development. It is a partnership among disaster risk financing and insurance into World the MDBs through which developed and develop- Bank operations. New regional initiatives can build ing countries come together with stakeholders. The the Caribbean Catastrophe Risk Insurance Facility, World Bank plays a critical tripartite role—as trustee which is in its fourth year of providing hurricane and/ of the funds, as implementing agency, and as host of or earthquake insurance to 16 Caribbean countries the administrative unit. More than 40 countries are and territories. With reinsurance, it has the capacity to currently undertaking Climate Investment Funds pilot handle claims arising from a series of events having a programs. The Pilot Program on Climate Resilience statistical probability of occurring only once in 1,000 helps highly vulnerable countries integrate risk and years without drawing more than $20 million of its resilience into core development planning. The Forest own assets (which now exceed $100 million). The Investment Program supports efforts to reduce defor- World Bank’s convening power is proving valuable estation and forest degradation, cut emissions, and to other pooling initiatives designed to reduce costs maintain carbon reservoirs. Mitigation is addressed and achieve efficiencies in risk fi nancing. In partner- through the Clean Technology Fund, which supports ship with the ADB and regional partners, the Pacific scaled-up fi nancing for low-carbon technologies in Catastrophe Risk Assessment and Financing Initiative middle-income countries. The Scaling Up Renewable has been developing several disaster risk management Energy Program aims to demonstrate low-carbon applications: disaster risk financing solutions, includ- energy development in low-income countries. ing a regional insurance facility; disaster risk prepared- The WBG also addresses climate change through ness through identification of critical infrastructure; Development Policy Operations, technical assistance, and disaster risk reduction through the identifica- investment lending, and analytical work. Develop- tion of high-risk areas. Other similar new programs ment Policy Operations provide support to govern- include the Catastrophe Risk Insurance Facility for ments implementing climate-related policies. In Southeastern Europe and the Africa Risk Capacity, an Ghana, for example, the Agriculture Development effort to pool drought risk across the continent. Convention to Combat Desertification. In programs cover the Pacific Islands, Central Asia, the Greater Mekong Subregion Pro- Asian regional economic cooperation, South gram has long been a prominent area of ADB Asian subregional economic cooperation, and support, promoting cross-country coopera- the Association of Southeast Asian Nations. tion in a number of sectors through invest- Similarly, the IADB works very closely with ments in infrastructure, policy initiatives, regional associations in Central and South and institutional mechanisms. Other regional America and the Caribbean. 154 PROGRESS IN THE DEVELOPMENT AND TRADING FRAMEWORK GLOBAL MONITORING REPORT 2011 IFIs’ response to crises since July 2008 when the full force of the financial crisis began to hit, the WBG has The IMF has taken the lead in responding to committed more than $152 billion and dis- crises by providing financial support to coun- bursed more than $96 billion in loans, grants, tries to permit orderly adjustment to payments equity investments, and guarantees in support crises. With the onset of the global financial of its clients. In fiscal 2010, the IBRD’s com- crisis, the IMF moved quickly to establish new mitments of nonconcessional resources were instruments to better assist its members—for $44.2 billion, up from a record high of $33 example, a new Flexible Credit Line (FCL) to billion in 2009. Low-income countries have provide large and up-front financing to mem- not had access to additional resources to the bers with very strong fundamentals and poli- same extent as middle-income countries. IDA cies. The facility can be used on a precaution- commitments in fiscal 2010 reached $14.5 bil- ary basis or for actual balance-of-payments lion, a 3.5 percent increase from 2009. Con- needs. Because access to the FCL is restricted cessional funds from MDBs increased only to those countries that meet strict qualifica- slightly, given the fi xed envelope of conces- tion criteria, drawings under it are not tied sional windows. To accelerate their response to ex post conditionality. Countries that do to the crisis, however, the IFIs have boosted not qualify for the FCL may receive financial flows to the poorest countries by frontload- assistance under High Access Precaution- ing available resources. Finally, the support ary Stand-by Arrangements, which can be to private sector and nonsovereign has also frontloaded and must take account of the been substantial. IFC investments increased strength of a country’s policies and the exter- by 21 percent in fiscal 2010, reaching $12.7 nal environment. The Precautionary Credit billion, with an additional $5.0 billion mobi- Line (PCL) was introduced to meet the needs lized from other sources. Almost half went to of countries that, despite having sound poli- IDA recipients—255 projects totaling $4.9 cies and fundamentals, have some remaining billion. Guarantees from the Multilateral vulnerabilities that preclude them from using Investment Guarantee Agency reached $1.5 the FCL. Decisions have been made on a dou- billion, up from $1.4 billion in 2009. bling of access levels, and conditionality has The role of the IFIs, of course, extends been reformed to make it more focused and beyond financing. Crucial roles for the IFIs in tailored to country circumstances. the context of the global economic crisis have IMF actions were complemented by the been to inform policy making by analyzing MDBs. Much of the increase in MDB financ- the international spillovers of national policy ing over the past two years took the form of actions and bringing out the interconnected budget support to quickly disburse funds to nature of the challenges and to highlight the protect the most vulnerable people against need to ensure that national responses are the fallout of the crisis, to maintain planned consistent with the global good. Amid ris- infrastructure investment, and to sustain ing pressures for policies to turn inward, the private sector–led economic growth and IFIs’ role in warning against the risks of trade employment creation. The countercyclicality protectionism and fi nancial mercantilism is of the fi nancial support provided during the indispensable. Drawing policy lessons from crisis by the IFIs was particularly important, the current crisis, especially but not only in given that the overall aid levels stagnated in financial regulation, will be another key area. real terms. The IMF has a particularly important role in The IFIs’ response was tailored to the enhanced surveillance of risk in the global- gravity, speed, and center of attention of the ized financial markets. event. The IMF has made commitments of As the recovery progresses, it is clear more than $250 billion since mid-2008, and that the crisis has dramatically altered the the MDBs had record commitments and dis- development challenges facing low-income bursements in the same period. For instance, and middle-income countries, and hence GLOBAL MONITORING REPORT 2011 PROGRESS IN THE DEVELOPMENT AND TRADING FRAMEWORK 155 those facing the international community. 23. Developmental aid comprises multilateral aid Managing the availability and allocation of and aid from some bilateral donors, with rela- resources will remain a challenge for the IFIs tively high aid quality; geopolitical aid is aid as the recovery proceeds. To this end, the from other bilateral donors. 24. Reddy and Minoiu 2006. IFIs’ shareholders increased the ability of the 25. Heyman 2010. IFIs to cope with crises and address devel- 26. Michaelowa and Weber 2006. opment needs through the increase in IMF 27. Dreher, Nunnenkamp, and Thiele 2006. resources and recent agreements on MDBs’ 28. Mishra and Newhouse 2007. capital increases. 29. Bourguignon and Sundberg 2007. 30. See, for example, Shafik (2007). Notes 31. Henn and McDonald 2010. 32. The Global Trade Alert database is available 1. OECD 2010. at http://www.globaltradealert.com. 2. Reinhart and Rogoff 2008. 33. Chauffour 2008, Jeonghoi 2010, and Martin 3. Dang, Knack, and Rogers 2009. See also and Anderson 2010. Mold et al. (2010). 34. Bouët et al. 2010. 4. They look at banking crises in 24 donor coun- 35. Sutherland and Bhagwati 2011. tries over 1977–2007. The data and the defini- 36. For examples, see Martin and Mattoo (2010) tion of “banking crisis” are based on Laeven and Hoekman, Martin, and Mattoo (2009). and Valencia (2008). 37. Mattoo and Subramanian 2008. 5. Norway has spent 0.85 percent of its gross 38. This implies simultaneously cutting trade national income on official development assis- barriers to external members or allowing tance over the past half-century, followed by their participation within the rules envisaged the Netherlands (0.79 percent), Sweden (0.77 under the agreement. percent), Denmark (0.77 percent), and Lux- 39. The database is available at http://data.world embourg (0.57 percent). bank.org/news/global-preferential-trade- 6. Davies 2010. agreement-database. 7. Ibid. 40. See Hoekman and Mattoo (2010). 8. IPRCC 2010. See also Qi (2007). 41. The World Integrated Trade Solution data- 9. Kharas 2009. base is available at http://wits.worldbank.org/ 10. Hudson Institute 2010. wits/. 11. Bishop and Green 2008. 42. Notes on some technical terms are as fol- 12. World Bank 2010. lows: “Rules of origin” determine whether 13. Easterly and Pfutze 2008. an import shipment has “originated” in a 14. Roodman 2009. preference-eligible exporting country and thus 15. Knack, Rogers, and Eubank 2010. whether it qualifies for a reduced tariff rate. 16. Birdsall and Kharas 2010. The rules may be specified according to the 17. Publish What You Fund: The Global Cam- share of “value added” that originated in the paign for Aid Transparency, http://www. preference-eligible country, or in other ways. publish whatyoufund.org/. “Cumulation” provisions allow intermediate 18. Statistical and technical problems include inputs that originate in certain third countries endogeneity, difficulty in determining the (usually other preference-receiving countries) direction of causality or controlling for to count toward fulfi lling the rule of origin, country-specific characteristics, multicol- thus providing flexibility in sourcing inputs linearity, autocorrelation, and selection of from low-cost providers. “De minimis” refers instruments in the analysis (Roodman 2008). to rules permitting exemption from notifica- 19. Burnside and Dollar 2000. tion for state aid under certain thresholds to 20. Doucouliagos and Paldam 2005, 2006. farmers, fishermen, and processing and mar- 21. Clemens, Radelet, and Bhavnani 2004. Short- keting companies. See the WTO Web site impact aid is defi ned as an aid disbursement (http://www.wto.org) for more discussion of funding an intervention that plausibly could various terms. increase growth within four years (balance- 43. Raballand, Kunaka, and Giersing 2008. of-payments support, infrastructure finance). 44. See Arvis et al. 2010. 22. Mavrotas 2005. 45. Chauffour and Malouche forthcoming. 156 PROGRESS IN THE DEVELOPMENT AND TRADING FRAMEWORK GLOBAL MONITORING REPORT 2011 46. Haddad, Harrison, and Hausman 2010. ity crunch and enhance DMCs’ capacity to 47. See Hoekman and Wilson (2010) and OECD/ provide fiscal stimulus. WTO (2010). 54. The EBRD has stepped up its support for 48. Our focus is on the five largest multilateral banks adversely affected by the crisis in the development banks—the African Develop- transition region by providing equity and ment Bank (AfDB), the ADB, the European debt finance that was otherwise unavail- Bank for Reconstruction and Development able on the fi nancial market. The enterprise (EBRD), the Inter-American Development response package aims at meeting short- Bank (IADB), and the World Bank—and the term refi nancing needs and sustaining exist- IMF. ing investment programs. The launch of 49. For most low-income countries, the Poverty the Corporate Support Facility ensured the Reduction Strategy Paper (PRSP) is the vehicle provision of quick-disbursing financing to used by the authorities to set out the national help companies weather the impact of the strategy. It serves as the foundation for the crisis and of opportunities to strengthen and strategy. For middle-income countries, there extend transition impact through loan con- is not an agreed format like the PRSP for set- ditionality. The EBRD is virtually doubling ting out the national strategy, and the MDBs its Trade Facilitation Programme (to a maxi- rely on a wide variety of country-specific vehi- mum of €1.5 billion), which plays a crucial cles as a basis for their country lending and role in keeping trade flowing to and from the nonlending programs. region in times of severely restricted access 50. Every Country Assistance Strategy since Jan- to fi nance. uary 2005 is results based (following a suc- 55. The IFC’s new initiatives focused on three main areas: (1) providing liquidity support cessful pilot project during FY03/04). through the expansion of the Global Trade 51. The EBRD countries of operation within Finance Program from $1 billion to $3 billion the Early Transition Countries Initiative are and creating the Global Trade Liquidity Pool, Armenia, Azerbaijan, Belarus, Georgia, Kyr- which was endorsed at the most recent G-20 gyz Republic, Moldova, Mongolia, Tajikistan, summit in London to provide credits to sup- Turkmenistan, and Uzbekistan. port trade fi nance over the next three years; 52. World Bank 2009. (2) launching the Infrastructure Crisis Facil- 53. The ADB’s Trade Finance Program (TFP) ity to provide liquidity for infrastructure proj- provides guarantees and loans to partner ects, creating the Microfinance Enhancement banks in support of international trade. A Facility in partnership with the German devel- substantial portion of the TFP’s portfolio opment bank, and rebuilding fi nancial infra- supports small and medium-size enterprises, structure by creating the Bank Capitaliza- and many transactions occur either intrare- tion Fund in which the IFC and the Japanese gionally or between developing member- government have invested $1 and $2 billion, countries (DMCs). In March 2009, following respectively, to provide additional capital for the financial crisis and market needs, the TFP banks in developing countries; and (3) design- limit was increased to $1 billion and transac- ing advisory programs in risk management tion tenor was extended to three years. The and nonperforming-loans management and ADB established a $3 billion Countercycli- implementing a Distressed Asset Recovery cal Support Facility in June 2009 to provide Program that would mobilize investment into support to fiscal stimulus by middle-income distressed asset pools and provide advisory countries. The facility helped these middle- services to enhance the efficiency of distressed income DMCs sustain critical development asset markets. expenditures for fiscal stimulus to counter 56. It is envisaged that the initiative could help the adverse impacts of the global economic developing countries share expertise in areas crisis during 2009–10. The facility is a short- such as managing commodity windfalls, term lending instrument and complements developing efficient tax systems, adapting to conventional program loans aimed at sup- new technologies, selecting public investment porting structural reforms over an extended projects with high economic and social rates period. Its aim is to provide fast-disbursing of return, reforming pensions, and creating crisis assistance to address short-term liquid- social safety nets that benefit the poor. GLOBAL MONITORING REPORT 2011 PROGRESS IN THE DEVELOPMENT AND TRADING FRAMEWORK 157 References ——— . 2006. “Aid Effectiveness on Accumula- tion: A Meta Study.” Kyklos 59 (2): 227–54. Arvis, J.-F., M. A. Mustra, L. Ojala, B. Shepherd, Dreher, A., P. Nunnenkamp, and R. Thiele. 2006. and D. Saslavsky. 2010. “Connecting to Com- “Does Aid for Education Educate Children? pete 2010: Trade Logistics in the Global Econ- Evidence from Panel Data.” Working Paper, omy. The Logistics Performance Index and Its Kiel Institute for the World Economy, Kiel, Indicators.” World Bank, Washington, DC. Germany. Arvis, J.-F., G. Raballand, and J.-F. Marteau. Easterly, W., and T. Pfutze. 2008. “Where Does 2010. The Cost of Being Landlocked: Logis- the Money Go? Best and Worst Practices in tics Costs and Supply Chain Reliability. Direc- Foreign Aid.” Working Paper 21, Brookings tions in Development Series. Washington, DC: Global Economy and Development, Washing- World Bank. ton, DC. Bishop, M., and M. Green. 2008. Philanthrocapi- Haddad, M., A. Harrison, and C. Hausman. talism: How Giving Can Save the World. Lon- 2010. “Decomposing the Great Trade Col- don: Bloomsbury. lapse: Products, Prices, and Quantities in the Birdsall, N., and H. Kharas. 2010. “Quality of 2008–2009 Crisis.” Working Paper 16253, Official Development Assistance Assessment.” National Bureau of Economic Research, Cam- Center for Global Development, Washington, bridge, MA. DC. Henn, C., and B. McDonald. 2010. “Crisis Protec- Bouët, A., D. Laborde Debucquet, E. Dienesch, tionism: The Observed Trade Impact.” http:// and K. Elliott. 2010. “The Costs and Benefits www.voxeu.org/index.php?q=node/5960. of Duty-Free, Quota-Free Market Access for Heyman, A. 2010. “Drilling Down Aid: Aid by Poor Countries: Who and What Matters?” Sector.” PhD diss., Bocconi University, Milan, Working Paper 990, International Food Policy Italy. Research Institute, Washington, DC. Hoekman, B., and A. Mattoo. 2010. “Services Bourguignon, F., and M. Sundberg. 2007. “Aid Trade Liberalization and Regulatory Reform: Effectiveness— Opening the Black Box.” Re-invigorating International Cooperation.” American Economic Review 97 (2): 316–21. Policy Research Working Paper 5517, World Burnside, C., and Dollar, D. 2000. “Aid, Policies, Bank, Washington, DC. and Growth.” American Economic Review 90 Hoekman, B., W. Martin, and A. Mattoo. 2009. (4): 847–68. “Conclude Doha: It Matters!” Policy Research Chauffour, J.-P. 2008. “Global Food Price Cri- Working Paper 5135, World Bank, Washing- sis—Trade Policy Options.” PREM Notes 120 ton, DC. (June). World Bank, Washington, DC. Hoekman, B., and J. Wilson. 2010. “Aid for Chauffour, J.-P., and M. Malouche. Forthcoming. Trade: Building on Progress Today for Tomor- Trade Finance during the Great Trade Col- row’s Future.” Policy Research Working Paper lapse. Washington, DC: World Bank. 5361, World Bank, Washington, DC. Clemens, M., S. Radelet, and R. Bhavnani. 2004. Hudson Institute. 2010. The Index of Global Phi- “Counting Chickens When They Hatch: The lanthropy and Remittances. Washington, DC. Short-Term Effect of Aid on Growth.” Work- IPRCC (International Poverty Reduction Center ing Paper 44, Center for Global Development, in China). 2010. “Development Partnerships Washington, DC. for Growth and Poverty Reduction.” Working Dang, H.-A., S. Knack, and H. Rogers. 2009. Paper 07, Beijing. “International Aid and Financial Crises in Jeonghoi, K. 2010. “Recent Trends in Export Donor Countries.” Policy Research Working Restrictions.” Trade Policy Working Paper Paper 5162, World Bank, Washington, DC. 101, Organisation for Economic Co-operation Davies, P. 2010. “A Review of the Roles and and Development, Paris, France. Activities of New Development Partners.” Kharas, H. 2009. “Development Assistance in Concessional Finance and Global Partnerships the 21st Century.” Paper prepared for the VIII Working Paper Series No. 4, World Bank, Salamanca Forum, “The Fight Against Hunger Washington, DC. and Poverty,” Salamanca, Spain, July 2–4. Doucouliagos, H., and M. Paldam. 2005. “The Knack, S., F. H. Rogers, and N. Eubank. 2010. Aid Effectiveness Literature: The Sad Result of “Aid Quality and Donor Rankings.” Policy 40 Years of Research.” Working Paper 2005- Research Working Paper 5290, World Bank, 15, University of Aarhus, Denmark. Washington, DC. 158 PROGRESS IN THE DEVELOPMENT AND TRADING FRAMEWORK GLOBAL MONITORING REPORT 2011 Kunaka, C. 2010. Logistics in Lagging Regions: Qi, G. 2007. “China’s Foreign Aid: Policies, Overcoming Local Barriers to Global Con- Structure, Practice and Trend.” Presentation nectivity. Washington, DC: World Bank. at the Oxford/Cornell University Conference, Laeven, L., and F. Valencia. 2008. “Systemic “New Directions in Development Assistance,” Banking Crises: A New Database.” Working Oxford, UK, June 11–12. Paper 08/224, International Monetary Fund, Raballand, G., C. Kunaka, and B. Giersing. 2008. Washington, DC. “The Impact of Regional Liberalization and Martin, W., and K. Anderson. 2010. “Trade Harmonization in Road Transport Services: A Distortions and Food Price Surges.” Paper Focus on Zambia and Lessons for Landlocked prepared for the World Bank–University of Countries.” Policy Research Working Paper California-Berkeley Conference on Agricul- 4482, World Bank, Washington, DC. ture for Development-Revisited, Berkeley, CA, Reddy, S., and C. Minoiu. 2006. “Development October 1–2. Aid and Economic Growth: A Positive Long- Martin, W., and A. Mattoo. 2010. “The Doha Run Relation.” Department of Economic and Development Agenda: What’s on the Table?” Social Affairs, United Nations, New York. Journal of International Trade and Economic Reinhart, C. M., and K. Rogoff. 2008. “Bank- Development 19 (1): 81–107. ing Crises: An Equal Opportunity Menace.” Mattoo, A., and Arvind Subramanian. 2008. Working Paper 14587, National Bureau of Eco- “Multilateralism Beyond Doha.” Policy nomic Research, Cambridge, MA. Research Working Paper 4735, World Bank, Roodman, D. 2008. “Through the Looking Glass, Washington, DC. and What OLS Found There: On Growth, For- Mavrotas, G. 2005. “Aid Heterogeneity: Looking eign Aid, and Reverse Causality.” Working at Aid Effectiveness from a Different Angle.” Paper 137, Center for Global Development, Journal of International Development 17 (8): Washington, DC. 1019–36. ——— . 2009. “An Index of Donor Performance.” Michaelowa, K, and A. Weber. 2006. “Aid Effec- Working Paper 67, Center for Global Develop- tiveness Reconsidered: Panel Data Evidence for ment, Washington, DC. the Education Sector.” Discussion Paper 264, Shafik, N. 2010. “The Future of Aid.” Eighth Hamburg Institute of International Econom- Luca d’Agliano Lecture in Development Eco- ics. Hamburg, Germany. nomics, Fondazione Luigi Einaudi, Turin, Italy, Mishra, P., and D. Newhouse. 2007. “Health Aid December 17. and Infant Mortality.” Working Paper 07/100, Sutherland, P., and J. Bhagwati. 2011. “The Doha International Monetary Fund, Washington, Round: Setting a Deadline, Defi ning a Final DC. Deal.” Interim report of the High Level Trade Mold, A., A. Prizzon, E. Frot, and J. Santiso. Experts Group. http://www.acp-eu-trade. 2010. “Aid Flows in Times of Crisis.” Presenta- org/library/fi les/High%20level%20trade%20 tion to the Conference on Development Coop- experts%20group_ EN _ 010111_The%20 eration In Times of Crisis and on Achieving the doha%20round%20setting%20a%20dead- MDGs, Madrid, Spain, June 9–10. line.pdf. OECD (Organisation for Economic Co-operation World Bank. 2009. “Moving Ahead on Invest- and Development). 2010. “Development Aid ment Lending Reform: Risk Framework and Rose in 2009 and Most Donors Will Meet Implementation Support.” SecM2009-0442, 2010 aid Targets.” Paris, France. http://www. September 9, Washington, DC. oecd.org/document/11/0,3343,en_2649_3444 ——— . 2010. “Research for Development—A 7_44981579_1_.1_1_37413,00.html World Bank Perspective on Future Directions OECD/WTO (Organisation for Economic Co- for Research.” Policy Research Working Paper operation and Development/ World Trade 5437, Washington, DC. Organization). 2010. “Aid for Trade: Is It ——— . 2011. World Development Report 2011: Working?” Policy Note. OECD, Paris, France. Confl ict, Security, and Development. Wash- ington, DC. Appendix 1 Technical Issues in Estimating Progress in the Millennium Development Goals Measuring Millennium progress continues until 2015 at the same Development Goals performance rate as progress from 1990 to the latest available year; T he Millennium Development Goals • off target: the country will not achieve the (MDGs) are typically defined in terms 2015 development goal if progress contin- of the number or percentage of people ues until 2015 at the same rate as progress (for example, halving the number of poor from 1990 to the latest available year; people or achieving 100 percent access to pri- • close to the target or off target and above mary education). Whereas data are generally average: the country is off target, but per- collected on a country basis, the influence of forming better than the average off-target each country in the global average depends country; or on the size of its population. When large • far from the target or off target and below countries like China and India are doing average: the country is off target and per- well—as on the poverty MDG—their prog- forming worse than the average off-target ress will be reflected very visibly in the global country. average, but will also hide progress (or a lack of it) in smaller countries. To examine how To determine whether a country is on tar- poor countries are doing, the data in chap- get, we calculate the linear annualized rate ter 1 are also presented in terms of progress of improvement from the 1990 value of each in individual countries—not to replace the MDG indicator needed to reach the 2015 standard approach, but to provide additional goal. We restrict our attention to the six information. MDGs and nine development targets with MDG performance is measured by devia- quantifiable 2015 goals. A country is classi- tions from target values required to reach fied as on target if the observed MDG per- development goals. The reference year for formance is equal to or above this required measuring progress is officially set as 1990. achievement path; a country is considered Countries are classified as off target if MDG progress is below this path. • on target: the country has already achieved Within the off-target category, countries or will meet the 2015 development goal if are classified in relation to the group’s average 159 160 APPENDIX 1 GLOBAL MONITORING REPORT 2011 TABLE A1.1 Data availability by MDG, income, and geographic region, fiscal 2011 MDG 3.a MDG 3.a MDG 7.c MDG 2.a gender gender MDG 4.a access Total MDG 1.a primary parity in parity in child MDG 5.a to safe MDG 7.c Income level and number of extreme MDG 1.c education primary secondary mortality maternal drinking access to geographic region countries poverty hunger completion education education under five mortality water sanitation Income Low income 40 22 24 29 36 31 40 40 40 40 Lower-middle income 56 34 20 44 50 47 56 46 51 51 Upper-middle income 48 29 11 39 41 42 48 38 41 39 Region East Asia and Pacific 24 8 3 15 19 17 24 15 21 21 Europe and Central Asia 22 20 4 18 19 18 22 21 18 16 Latin America and the Caribbean 30 20 12 26 26 26 30 24 29 29 Middle East and North Africa 13 7 7 11 11 12 13 12 13 13 South Asia 8 5 4 6 8 8 8 8 7 7 Sub-Saharan Africa 47 25 25 36 44 39 47 44 44 44 Source: World Development Indicators. distance to be on target. (The average here on reaching the 2015 MDGs). Performance is the mean of the off-target group, not the data for these two subgroups are presented in entire sample of countries.) Two subgroups tables A1.2, A1.3, and A1.4. are identified: countries that are off target and above average or close to the target (that is, countries for which development goals are Data constraints possibly within reach); and countries that Detailed historical data on MDG performance are off target and below average or far from are required to calculate the achievement path the target (that is, countries lagging the most for each country to meet each of the MDGs.1 TABLE A1.2 MDG performance: share of countries close to the target in the total number of countries off target, by geographic region percent MDG 7.c MDG 2.a MDG 3.a MDG 3.a MDG 4.a access MDG 1.a primary gender parity gender parity child MDG 5.a to safe MDG 7.c extreme MDG 1.c education in primary in secondary mortality maternal drinking access to Geographic region poverty hunger completion education education under five mortality water sanitation East Asia and Pacific 100 100 100 75 100 33 64 56 62 Europe and Central Asia 0 — 100 100 100 92 40 50 27 Latin America and 25 100 100 91 100 85 43 60 39 the Caribbean Middle East and 0 50 80 67 33 50 80 44 50 North Africa South Asia 100 67 67 50 50 60 80 33 80 Sub-Saharan Africa 81 53 48 47 50 25 38 69 29 Source: World Bank staff calculations based on data from the World Development Indicators database. Note: — = not applicable. Shading indicates the share of countries close to target among all off-target countries that is below 50 percent. GLOBAL MONITORING REPORT 2011 APPENDIX 1 161 TABLE A1.3 MDG performance: share of countries close to the target in the total number of countries off target, by income level percent MDG 3.a MDG 3.a MDG 7.c MDG 2.a gender gender MDG 4.a access MDG 1.a primary parity in parity in child MDG 5.a to safe MDG 7.c extreme MDG 1.c education primary secondary mortality maternal drinking access to Income level poverty hunger completion education education under five mortality water sanitation Low-income 88 50 41 36 35 34 51 67 31 countries Lower-middle- 38 80 84 72 85 53 53 43 48 income countries Upper-middle- 17 50 100 89 100 63 37 67 39 income countries Source: World Bank staff calculations based on data from the World Development Indicators database. Note: Shading indicates the share of countries close to target among all off-target countries that is below 50 percent. TABLE A1.4 MDG performance: share of countries close to the target in the total number of countries off target, by alternative typologies percent MDG 3.a MDG 3.a MDG 7.c MDG 2.a gender gender MDG 4.a access MDG 1.a primary parity in parity in child MDG 5.a to safe MDG 7.c extreme MDG 1.c education primary secondary mortality maternal drinking access to Alternative typology poverty hunger completion education education under five mortality water sanitation A. IDA classification Non-IDA countries 27 64 100 82 92 57 45 57 49 IDA countries 74 58 50 52 46 42 52 61 29 B. State fragility Little to low 25 50 100 100 100 53 47 63 47 Moderate 33 100 100 83 100 65 32 50 45 Serious 73 50 56 83 50 65 87 73 50 High to extreme 70 67 27 23 38 26 33 64 23 C. Export sophistication High 67 — 100 100 — 67 17 50 57 Medium high 43 60 67 100 33 59 68 50 65 Medium low 58 64 53 62 62 33 40 75 38 Low 60 50 69 67 70 52 63 58 25 Source: World Bank staff calculations based on data from the World Development Indicators database, Marshall and Cole 2010, and Lall 2000. Note: — = not applicable; IDA = International Development Association. Shading indicates the share of countries close to target among all off-target countries that is below 50 percent. Unfortunately, such data are not available in starting point are significantly different from many countries for 1990, although estimates the level of MDG performance in 1990 or if for recent years tend to be more complete. If the sample period does not capture the latest no country data are available for 1990, we progress. The latter is a particularly impor- used the closest available information in the tant issue now because data generally are not late 1980s or early 1990s as substitutes for available for 2009, the crisis year. In addi- the starting point (table A1.5), and then cal- tion, for countries without at least two data culated the rate of progress required from points, progress cannot be measured, even that point to meet the MDG. This approach if data are available for a recent year. Even may be inaccurate if the data for the available so, the approach enables us to include more 162 APPENDIX 1 GLOBAL MONITORING REPORT 2011 TABLE A1.5 Development indicators, average levels circa 1990 MDG 3.a MDG 3.a MDG 7.c MDG 2.a gender gender MDG 4.a access MDG 1.a primary parity in parity in child MDG 5.a to safe MDG 7.c extreme MDG 1.c education primary secondary mortality maternal drinking access to MDG performance  poverty hunger completion education education under five mortality water sanitation On target 24.14 19.47 88.61 90.76 97.26 73.37 150.73 23.75 30.43 Close to the target 55.97 25.66 68.22 91.97 79.75 87.94 525.39 38.15 51.19 Far from the target 9.70 23.04 26.09 73.65 50.92 112.47 485.98 24.52 53.03 Total 29.43 22.27 68.99 89.14 87.00 93.44 419.49 28.26 46.42 Source: World Bank staff calculations based on data from the World Development Indicators database. countries than if we relied only on data from target (off target and above average), and 3 1990 and 2008. Table A1.1 summarizes the for countries on target. Avoiding the use of number of observations by MDG, income the actual value of MDG indexes is impor- group of countries, and region (World Bank’s tant for two reasons. First, the index num- classification, table A1.13). bers that indicate progress in many MDG indicators display substantial variability for countries performing well below or above The multinomial logit estimates average. Taking account of this variability Empirically analyzing the MDGs from a would require some form of data trimming, cross-country perspective imposes serious outlier identification procedure, or inclusion challenges, with frequent data gaps and mea- of control variables that would reduce the surement errors. Plausible functional forms available degrees of freedom and therefore are therefore difficult to derive. And several decrease the reliability of estimates, in a con- MDGs—such as access to safe drinking text of small data samples. Second and more water and health targets—are likely to be sig- important, our goal is to assess the likelihood nificantly cross-correlated. Here we discuss of each country achieving or being on track some of the most significant issues affecting to achieve the MDGs, conditional on cur- our methodological approach.2 rent development performance, an empirical We employ the multinomial logit estima- approach consistent with the use of models tion method because it is well suited to exam- of categorical dependent variables. We are ine the likelihood that countries fall into one not trying to determine how much per cap- of the three country groups explained above ita GDP must grow or institutions and poli- (on target, close to the target, and far from cies must improve to attain the development the target), given changes in economic growth goals by 2015—for which observed values and the policy framework. This method is of development indicators and linear regres- typically employed to model individual dis- sion models are better suited (although these crete choices, such as the occupational choice models, as well as nonlinear approaches, may of households in microsimulations or demand suffer from endogeneity and multicollinearity for modes of transportation. problems).3 Dependent variable and estimation At first glance, an ordinal regression model method. The multinomial logit model does seems appropriate to analyze the extent to not use the actual values of MDG perfor- which GDP growth and the policy frame- mance indexes. Instead, MDG performance work determine the likelihood of being on is defined in terms of three values: 1 for track to achieve the MDGs. Our initial work countries far from the target (off target and therefore involved estimating this relationship below average), 2 for countries close to the using an ordered logistic regression model. GLOBAL MONITORING REPORT 2011 APPENDIX 1 163 However, a fundamental assumption of such that the current level of the CPIA indexes models is that the explanatory variables have partially reflects past performances, given the the same impact across different values for slowly evolving nature of institutions. the dependent variable (the proportional The independence of irrelevant alter- odds hypothesis), and this assumption is con- natives. One assumption of our estima- sistently rejected in most of the nine specifi- tion procedure is that the results satisfy the cations under consideration. These rejections independence of irrelevant alternatives (IIA) imply that the coefficients associated with per assumption. This assumption, which is often capita growth and institutions are not equal used in the context of public choice theory, is across levels of MDG performance. For this that the odds of an outcome (in this exercise, reason, alternative and less restrictive mod- being on target, close to target, or far from els that can integrate a differentiated impact target) do not depend on other alternatives of growth and policy on the dependent vari- that are available (alternatives are dissimilar). able (MDG performance) are required. Con- This means that the coefficients on indepen- sequently, we turn to the multinomial logit dent variables (in this exercise, growth and model, a nominal outcome estimation tech- the quality of institutions) would not change nique that reduces the risk of bias resulting significantly if we were to suppress one of the from the rejection of the proportional odds categories of the dependent variable and rees- hypothesis in the ordinal regression approach, timate the model. We performed Small-Hsiao but at the cost of a potential loss of efficiency, tests6 to determine whether results satisfy the given the many parameters in the model.4 IIA assumption, with generally satisfactory Functional form, model specification, results (calculations of the Small-Hsiao test and the use of the Country Policy and Insti- are shown in table A1.7).7 In any event, it is tutional Assessment index. The empiri- generally acknowledged that IIA tests have lit- cal model follows a simple structure that tle power in small samples and may even pro- addresses the fundamentals of the Global vide conflicting results.8 According to McFad- Monitoring Report framework, 5 builds on den, the multinomial logit model “should be the current MDG literature, and takes into limited to situations where the alternatives account data limitations. Progress toward can plausibly be assumed to be distinct and the 2015 goals is expressed as a function of weighted independently in the eyes of each initial conditions and development progress decision-maker.”9 Therefore, the validity of over time (table A1.6). In this context, mea- our conclusions (in terms of the IIA assump- suring the quality of policies and institutions tion) relies more on the fact that our catego- represents a major challenge. We use the ries are conceptually independent than on this Country Policy and Institutional Assessment econometric test. (CPIA) index, which provides a comprehen- Endogeneity, reverse causality. Indica- sive assessment of social policies and public tors of progress in human development (our sector management that is fundamental for dependent variables) can have an impact on MDG attainment. For example, the CPIA growth and the quality of institutions (our index takes into consideration the extent to independent variables). Thus, our estimations which the pattern of public expenditures and could be subject to reverse causality, which revenue collection affects the poor and is con- would mean that the coefficients on the inde- sistent with national poverty reduction pri- pendent variables are not correctly estimated. orities (criterion 8, equity of public resource However, such concerns are likely to be less use). However, an important limitation of of a problem in our estimation than in regres- the CPIA index is that values are not strictly sions using the levels of MDGs (for example, comparable across years because of numerous where the level of the poverty head count methodological changes over time. Therefore, is the dependent variable). This is because we focus on current index values, assuming small changes in the dependent variable (for 164 APPENDIX 1 TABLE A1.6 Growth has a clear, pervasive effect, but better policy also helps improve the odds of achieving the MDGs multinomial logit estimates (1) (2) (3) (4) (5) (6) (7) (8) (9)   MDG 1.a MDG 1.c MDG 2.a MDG3.a (primary) MDG 3.a (secondary) MDG 4.a MDG 5.a MDG 7.a (water) MDG 7.a (sanitation) Independent Above On Above On Above On Above On Above On Above On Above On Above On Above On variables average target average target average target average target average target average target average target average target average target   coef/se coef/se coef/se coef/se coef/se coef/se coef/se coef/se coef/se coef/se coef/se coef/se coef/se coef/se coef/se coef/se coef/se coef/se Annual growth in GDP per capita, 0.262 0.359*** 0.335*** 0.427** 0.766*** 1.304*** 0.627 0.470 0.922* 1.411** 0.240* 0.518** 0.187* 0.546 0.105 0.253*** 0.260 0.383*** average for 1990–2009 (2005 Id PPP) 0.217 0.131 0.051 0.177 0.187 0.197 0.440 0.340 0.513 0.558 0.132 0.211 0.101 0.397 0.120 0.038 0.187 0.129 CPIA 2009 0.609 0.211 1.071*** 2.784*** – 0.158 0.128 0.725 0.925*** –1.007 –0.745 1.119** 1.670*** 0.864*** 1.382*** –2.033** –1.203 0.690 0.409 0.892 1.525 0.398 0.330 0.258 0.732 0.524 0.351 0.664 1.437 0.553 0.579 0.173 0.367 1.005 0.990 0.591 0.368 GDP per capita 1990 –0.010 0.041* 0.001 0.002 0.079** 0.087** 0.008 0.005 0.146** 0.141** –0.005 –0.001 –0.025** 0.002 0.051*** 0.046*** –0.005 –0.012 (2005 Id PPP) 0.059 0.023 0.014 0.012 0.034 0.039 0.013 0.016 0.071 0.069 0.005 0.010 0.012 0.008 0.014 0.010 0.006 0.016 CPIA 1996 0.906 0.740 0.585 0.294 0.726 –0.511 –0.601* 0.363** 1.295* 2.209*** –0.801** 0.047 0.183 –0.360 1.652*** 1.927** 0.103 –0.355 1.020 1.404 0.741 0.341 0.549 0.958 0.321 0.153 0.673 0.629 0.400 0.412 0.126 0.231 0.582 0.760 0.278 0.457 GLOBAL MONITORING REPORT 2011 Extreme poverty 0.097*** 0.079*** circa 1990 0.015 0.017 Hunger –0.004 –0.040 circa 1990 0.036 0.049 Primary education 0.167*** 0.221*** completion circa 1990 0.027 0.029 Gender parity in primary 0.062** 0.054* education circa 1990 0.026 0.031 GLOBAL MONITORING REPORT 2011 Gender parity in secondary 0.089*** 0.127*** education circa 1990 0.007 0.014 Child mortality –0.006 –0.005 under five circa 1990 0.005 0.006 Maternal mortality –0.000 –0.004 circa 1990 0.001 0.003 Access to safe drinking 0.077*** 0.041*** water circa 1990 0.011 0.015 Access to sanitation 0.003 –0.033 circa 1990 0.017 0.022 _cons –7.751** –5.864*** –5.259** –9.741*** –9.462*** –11.842*** –5.422*** –6.887*** –7.736*** –14.483*** –0.988 –7.021** –3.069*** –4.741** –1.408 –3.540* –3.612 0.266 3.479 0.897 2.281 2.609 2.271 3.435 1.896 2.277 1.148 3.278 2.204 2.794 0.507 2.101 2.251 2.085 2.263 2.259 Number of 77 49 90 105 95 114 104 106 107 observations, adjusted R2 0.293 0.185 0.515 0.161 0.415 0.146 0.211 0.192 0.123 Source: World Bank staff calculations. Note: Robust estimates with regional clusters. CPIA = Country Policy and Institutional Assessment; coef/se = coefficient/standard error; Id = international dollars; PPP = purchasing power parity. *p < 0.1; **p < 0.05; ***p < 0.01. APPENDIX 1 165 166 APPENDIX 1 GLOBAL MONITORING REPORT 2011 TABLE A1.7 Small-Hsiao tests of IIA assumption—null hypothesis: odds (outcome-j versus outcome-k) are independent of other alternatives Equation Omitted outcome Lnl(full) Lnl(omit) chi2 df P>chi2 Close to the target –9.65 –8.21 2.88 6.0 0.82 (1) On target –2.20 0.00 4.40 6.0 0.62 Close to the target –0.01 0.00 0.03 6.0 1.00 (2) On target –0.14 0.00 0.27 6.0 1.00 Close to the target –0.03 0.00 0.05 6.0 1.00 (3) On target –8.05 –3.36 9.37 6.0 0.15 Close to the target –21.13 –11.46 19.33 6.0 0.00 (4) On target –10.43 –5.90 9.07 6.0 0.17 Close to the target –0.01 0.00 0.02 6.0 1.00 (5) On target –0.02 0.00 0.04 6.0 1.00 Close to the target –13.66 –12.33 2.65 6.0 0.85 (6) On target –24.72 –22.87 3.69 6.0 0.72 Close to the target –14.32 –11.65 5.34 6.0 0.50 (7) On target –22.85 –20.16 5.38 6.0 0.50 Close to the target –12.54 –9.20 6.67 6.0 0.35 (8) On target –20.60 –11.08 19.06 6.0 0.00 Close to the target –24.05 –19.67 8.76 6.0 0.19 (9) On target –19.08 –16.15 5.86 6.0 0.44 Source: World Bank staff calculations. Note: IIA = independence of irrelevant alternatives. example, poverty head count) in the latter in secondary education, and access to safe case may have a direct impact on the inde- water and sanitation. In addition, CPIA scores pendent variable (for example, growth). In appear to have significant marginal effects, at our estimations, the dependent variable is average values, on the probabilities of being inclusion in a group defined by deviations on target (positive signs) and/or far from tar- from an exogenously determined path (for get (negative signs) for several health-related example, the rate of change in poverty nec- MDGs (hunger, child mortality, and mater- essary to achieve the goal). The connection nal mortality). Note that for several devel- between inclusion in one of the three groups opment goals, the predicted probability of a and growth is much more tenuous. country being close to target is significantly Predicted probabilities, marginal effects, and inversely related to changes in per capita and odds ratios. Table A1.8 summarizes the growth and the CPIA index (that is, higher effect of marginal changes in independent growth may reduce the probability of being variables on the probability of a country being close to the target). This does not imply that in one of our three groups at average sample high growth is correlated with poor perfor- values. Results show that a one-unit marginal mance. Rather, countries with relatively high increase in per capita GDP growth is signifi- growth may be on track to meet the goals, cantly and inversely related to the probabil- instead of off-track but close to the target. ity of a country being far from the target in The odds ratios or factor change coef- all MDGs, excluding completion of primary ficients (table A1.9) illustrate the dynamics schooling and gender parity in secondary among MDG performance outcomes. These education. Conversely, a one-unit increase in coefficients depict the expected change in the GDP per capita growth significantly raises the probability of a country being on target ver- probability of a country being on target by at sus far from the target and on target versus least 0.05, holding other variables at their close to the target, following a one-standard- mean, for primary completion, gender parity deviation increase in development drivers and GLOBAL MONITORING REPORT 2011 TABLE A1.8 Predicted probabilities and marginal effects MDG 3.a MDG 3.a MDG 7.a MDG 7.a   MDG 1.a     MDG 1.c     MDG 2.a     (primary)     (secondary)     MDG 4.a     MDG 5.a     (water)     (sanitation)   Close Far Close Far Close Far Close Far Close Far Close Far Close Far Close Far Close Far Predicted On to from On to from On to from On to from On to from On to from On to from On to from On to from probabilities target target target   target target target   target target target   target target target   target target target   target target target   target target target   target target target   target target target Predicted 0.69 0.19 0.13 0.43 0.40 0.17 0.42 0.58 0.00 0.75 0.22 0.03 0.73 0.27 0.00 0.20 0.44 0.36 0.10 0.44 0.47 0.57 0.27 0.15 0.23 0.31 0.46 probabilities Change in predicted probabilities following 0.04 –0.01 –0.04 0.05 0.01 –0.06   0.13 –0.13 0.00 –0.02 0.03 –0.01   0.10 –0.10 0.00 0.06 0.01 –0.08   0.04 0.02 –0.06   0.05 –0.02 –0.03   0.05 0.03 –0.08 a one-unit increase in GDP per capita growth Change in predicted probabilities –0.03 0.07 –0.03 0.50 –0.22 –0.28 0.07 –0.07 0.00 0.05 –0.03 –0.02 0.05 –0.05 0.00 0.17 0.13 –0.30 0.09 0.15 –0.24 0.02 –0.22 0.16 0.02 0.12 –0.14 following a one-unit increase in CPIA score Source: World Bank staff calculations. Note: CPIA = Country Policy and Institutional Assessment. Predicted probabilities and changes in predicted probabilities for each category are computed at average sample values. Results are not comparable across indica- tors. Numbers in bold type denote significant changes at 0.10 level or better. APPENDIX 1 167 168 APPENDIX 1 GLOBAL MONITORING REPORT 2011 TABLE A1.9 Effects of a one-standard-deviation increase in selected development drivers percent   Due to increase in GDP per capita growth   Due to increase in CPIA index on target vs. on target vs. on target vs. on target vs. MDG far from target close to target   far from target close to target MDG 1.a extreme poverty 93 19 12 –19 MDG 1.c hunger 88 15 281 128 MDG 2.a primary completion rate 1,111 180 7 16 MDG 3.a gender parity (primary) 141 –25 67 12 MDG 3.a gender parity (secondary) 1,191 143 –34 16 MDG 4.a child mortality under five 163 68 152 36 MDG 5.a maternal mortality 189 101 120 34 MDG 7.c access to safe water 61 32 –48 58 MDG 7.c access to sanitation 102 25 26 –15 Source: World Bank staff calculations. Note: Percentage variations are not comparable across indicators. Average standard deviation increase in GDP per capita growth = 1.8. Average standard deviation increase in CPIA index = 0.5. Numbers in bold type denote significance at 0.10 level or better. holding all other variables constant. (Results measurements of policy and institutional are discussed in chapter 1.) performance, we test the robustness of our Disaggregation of the CPIA index. Our results by including in the analysis other main results use the aggregate CPIA index as indicators of government performance: state the independent variable. However, we also fragility, as measured by Marshall and Cole’s did estimations using the four components of index10 and governance indicators produced the CPIA index in 2009 (economic manage- by Kaufmann, Kraay, and Mastruzzi.11 The ment, structural policies, policies for social impact of these institutional variables, as inclusion and equity, and public sector man- well as the CPIA index, on the odds of being agement and institutions) separately. These on target is summarized in tables A1.10 and results show that policies for social inclusion A1.11. and equity (gender equality, equity of pub- Table A1.10 shows the links between fra- lic resource use, building human resources, gility and governance variables and the odds social protection, and labor) are significantly of a country being on target or far from the and positively correlated with the odds of a target. Results are broadly consistent with our country being on target to achieve the MDGs previous estimates using the CPIA index: per- for extreme poverty, primary completion, ceptions of political stability and regulatory gender parity, and child mortality. Results quality are positively related to the likelihood are mixed for the three remaining indica- of achieving the hunger target; perceptions tors (secondary education, access to water of government effectiveness, regulatory qual- and sanitation, and maternal mortality) and ity, rule of law, and control of corruption are vary according to whether a country is or is positively related to achieving the targets for not close to the target. However, these results gender equality and primary education; per- have to be interpreted with caution because ceptions of state fragility are inversely related of the high collinearity between CPIA sub- to child mortality; and perceptions of regu- categories (pairwise correlations above 0.65) latory quality are positively correlated with and the limited degrees of freedom, given the achieving the target for maternal mortality. substantial number of parameters and the rel- Table A1.11 displays the relationship atively small size of available data samples. between institutional indicators and the prob- Alternative measures of policy and insti- ability of being on target versus close to the tutions. Given the uncertainties surrounding target. Previous results are partly confirmed. GLOBAL MONITORING REPORT 2011 APPENDIX 1 169 TABLE A1.10 Probability of being on target or far from the target average MDG 3.a MDG 3.a MDG 7.c MDG 7.c Indicator MDG 1.a MDG 1.c MDG 2.a (primary) (secondary) MDG 4.a MDG 5.a (water) (sanitation) CPIA index (2009) + + + + State fragility (1995–2009) – – – Voice and accountability Political stability + + – Government effectiveness + + + + Regulatory quality + + + + Rule of law + – Control of corruption – + + Source: World Bank staff calculations. Note: CPIA = Country Policy and Institutional Assessment. Significant at 0.1 level or better. Empty cells = no significant correlation. State fragility (1995–2009) is the difference between index values in 2009 and 1996. Governance indicators correspond to 2009 rankings. TABLE A1.11 Probability of being on target or close to the target average MDG 3.a MDG 3.a MDG 7.c MDG 7.c Indicator MDG 1.a MDG 1.c MDG 2.a (primary) (secondary) MDG 4.a MDG 5.a (water) (sanitation) CPIA index (2009) + + + State fragility (1995–2009) – Voice and accountability – Political stability + – – + Government effectiveness + – – Regulatory quality + + – Rule of law Control of corruption + Source: World Bank staff calculations. Note: CPIA = Country Policy and Institutional Assessment. Significant at 0.1 level or better. Empty cells = no significant correlation. State fragility (1995–2009) is the difference between index values in 2009 and 1996. Governance indicators correspond to 2009 rankings. Political stability and government effective- off target and below average (table A1.10); ness are positively linked to the hunger target. however, per capita GDP growth loses its And political stability, regulatory quality, and significance in the case of political stability control of corruption exhibit positive correla- and regulatory quality (results not shown). tions with access to safe drinking water. Conversely, voice and accountability—that These estimations provide additional is, civil and political rights—are not signifi- and interesting links between indicators of cantly related to most MDGs (tables A1.10 institutional quality and progress toward and A1.11); and when a significant correla- the MDGs. For instance, political stability, tion is found (table A1.11, extreme poverty), regulatory quality, and government effective- the observed sign is negative (the result is ness are positively correlated with poverty counterintuitive). In addition, a negative reduction for countries on target versus those relationship is found between some other 170 APPENDIX 1 GLOBAL MONITORING REPORT 2011 governance variables (particularly, politi- education coverage). In part, the selection cal stability and government effectiveness) of MDG indicators reflected the availabil- and the attainment of several MDGs (child ity of data. In what follows, we try to assess and maternal mortality and access to water how important institutions are to achieving and sanitation [tables A1.10 and A1.11]). A higher levels of literacy—a broadly avail- complete analysis of the role of governance able outcome-based measure of educational in achieving the MDGs is beyond the scope attainment. of this appendix. However, these apparently An explicit target for literacy was not counterintuitive outcomes are consistent with included in the MDGs, so defining outcomes the fact that many of the poorest countries are in terms of our three categories (on track, off making important progress toward achieving track but above average, and off track and the MDGs, thanks to sustained growth and below average) would be difficult. Instead, despite significant institutional weaknesses— we calculate the growth rate of the literacy a finding that highlights the necessity of a rate between 1990 and 2009 (or the closest better understanding of the mechanisms available year) and divide the countries into through which policies and institutions pro- three categories: mote development. Education outcomes versus outputs. Our • countries below percentile (33), thus exhib- results show that policies and institutions are iting the slowest progress; significantly and positively correlated with • countries above percentile (33) and below several health-related MDGs, whereas cor- percentile (66), thus close to the median; relations with education attainment are not and as strong. One possibility is that these dif- • countries above percentile (66), the best ferences arise from the fact that the health performers. targets are defined in terms of outcomes (for example, child and maternal mortal- To provide a rough comparison with ity), whereas the education goals are defined access-based indicators, we perform a similar in terms of access (for example, primary procedure on the primary completion rate. Several multinomial logit models, reproduc- ing our core specification, are then estimated linking education outcomes and access to TABLE A1.12 Probability of being in percentile (66) or above or in percentile (33) or below various proxies of institutional quality, using as the reference category the group of coun- Primary completion tries below percentile (33). rate, total Literacy rate, adult total (% of relevant (% of people aged 15 The results are mixed and, to some Indicator age group) and above) extent, counterintuitive (see table A1.12) so CPIA index (2009) + it is impossible to draw strong conclusions. State fragility However, it is interesting that the CPIA is not (1995–2009) significantly related to the outcome indica- Voice and accountability tor (literacy) but in this specification (unlike Political stability + Government effectiveness – our earlier results) is significantly related to Regulatory quality – + the access indicator (primary completion Rule of law + rate). Although these results are difficult to Control of corruption + + interpret, it does not appear that the dif- Source: World Bank staff calculations. fering results for the impact of growth and Note: CPIA = Country Policy and Institutional Assessment. Significant at 0.1 level or better. Empty cells = no significant correlation. State fragility (1995–2009) is the difference between institutions on progress toward achieving the index values in 2009 and 1996. Governance indicators correspond to 2009 rankings. health and education MDGs can be ascribed simply to the use of outcome versus access indicators. GLOBAL MONITORING REPORT 2011 APPENDIX 1 171 FIGURE A1.1 Odds of achieving the MDGs improve with growth and better policy a. Annual per capita GDP growth and predicted probabilities MDG 1.a extreme poverty –0.28 0.29 MDG 1.c hunger –0.73 0.54 MDG 2.a primary completion rate –0.33 0.48 MDG 3.a gender parity (primary education) –0.58 0.38 MDG 3.a gender parity (secondary education) –0.46 0.65 MDG 4.a child mortality under five –0.77 0.76 MDG 5.a maternal mortality –0.82 0.53 MDG 7.c access to safe water –0.27 0.61 MDG 7.c access to sanitation –0.86 0.51 –1.00 –0.80 –0.60 –0.40 –0.20 0.00 0.20 0.40 0.60 0.80 1.00 pairwise correlations between predicted probabilities and annual per capita GDP growth, average 1990–2009 far from target on target b. CPIA index (2009) and predicted probabilities MDG 1.a extreme poverty –0.20 0.55 MDG 1.c hunger –0.93 0.88 MDG 2.a primary completion rate –0.38 0.41 MDG 3.a gender parity (primary education) –0.60 0.73 MDG 3.a gender parity (secondary education) –0.44 0.62 MDG 4.a child mortality under five –0.78 0.80 MDG 5.a maternal mortality –0.78 0.64 MDG 7.c access to safe water –0.19 0.71 MDG 7.c access to sanitation –0.66 0.32 –1.00 –0.80 –0.60 –0.40 –0.20 0.00 0.20 0.40 0.60 0.80 1.00 pairwise correlations between predicted probabilities and CPIA index, 2009 far from target on target Source: World Bank staff calculations. Note: CPIA = Country Policy and Institutional Assessment. Pairwise correlations are significant at the 0.10 level or better. 172 APPENDIX 1 GLOBAL MONITORING REPORT 2011 TABLE A1.13 Classification of economies by region and income, fiscal 2011 East Asia and Pacific Latin America and the Caribbean Sub-Saharan Africa High-income OECD economies American Samoa UMC Antigua and Barbuda UMC Angola LMC Australia Cambodia LIC Argentina UMC Benin LIC Austria China LMC Belize LMC Botswana UMC Belgium Fiji UMC Bolivia LMC Burkina Faso LIC Canada Indonesia LMC Brazil UMC Burundi LIC Czech Republic Kiribati LMC Chile UMC Cameroon LMC Denmark Korea, Dem. Rep. LIC Colombia UMC Cape Verde LMC Estonia Lao PDR LIC Costa Rica UMC Central African Republic LIC Finland Malaysia UMC Cuba UMC Chad LIC France Marshall Islands LMC Dominica UMC Comoros LIC Germany Micronesia, Fed. Sts. LMC Dominican Republic UMC Congo, Dem. Rep. LIC Greece Mongolia LMC Ecuador LMC Congo, Rep. LMC Hungary Myanmar LIC El Salvador LMC Côte d'Ivoire LMC Iceland Palau UMC Grenada UMC Eritrea LIC Ireland Papua New Guinea LMC Guatemala LMC Ethiopia LIC Israel Philippines LMC Guyana LMC Gabon UMC Italy Samoa LMC Haiti LIC Gambia, The LIC Japan Solomon Islands LIC Honduras LMC Ghana LIC Korea, Rep. Thailand LMC Jamaica UMC Guinea LIC Luxembourg Timor-Leste LMC Mexico UMC Guinea-Bissau LIC Netherlands Tonga LMC Nicaragua LMC Kenya LIC New Zealand Tuvalu LMC Panama UMC Lesotho LMC Norway Vanuatu LMC Paraguay LMC Liberia LIC Poland Vietnam LMC Peru UMC Madagascar LIC Portugal St. Kitts and Nevis UMC Malawi LIC Slovak Republic Europe and Central Asia St. Lucia UMC Mali LIC Slovenia St. Vincent and Mauritania LIC Spain Albania UMC the Grenadines UMC Mauritius UMC Sweden Armenia LMC Suriname UMC Mayotte UMC Switzerland Azerbaijan UMC Uruguay UMC Mozambique LIC United Kingdom Belarus UMC Venezuela, R.B. de UMC Namibia UMC United States Bosnia and Herzegovina UMC Niger LIC Bulgaria UMC Middle East and North Africa Nigeria LMC Other high-income economies Georgia LMC Rwanda LIC Kazakhstan UMC Algeria UMC São Tomé and Principe LMC Andorra Kosovo LMC Djibouti LMC Senegal LMC Aruba Kyrgyz Republic LIC Egypt, Arab Rep. LMC Seychelles UMC Bahamas, The Lithuania UMC Iran, Islamic Rep. UMC Sierra Leone LIC Bahrain Macedonia, FYR UMC Iraq LMC Somalia LIC Barbados Moldova LMC Jordan LMC South Africa UMC Bermuda Montenegro UMC Lebanon UMC Sudan LMC Brunei Darussalam Romania UMC Libya UMC Swaziland LMC Cayman Islands Russian Federation UMC Morocco LMC Tanzania LIC Channel Islands Serbia UMC Syrian Arab Republic LMC Togo LIC Croatia Tajikistan LIC Tunisia LMC Uganda LIC Cyprus Turkey UMC West Bank and Gaza LMC Zambia LIC Equatorial Guinea Turkmenistan LMC Yemen, Rep. LMC Zimbabwe LIC Faeroe Islands Ukraine LMC French Polynesia Uzbekistan LMC South Asia Gibraltar Greenland Afghanistan LIC Guam Bangladesh LIC Hong Kong SAR, China Bhutan LMC Isle of Man India LMC Kuwait Maldives LMC Latvia Nepal LIC Liechtenstein Pakistan LMC Macao SAR, China Sri Lanka LMC Malta Monaco Netherlands Antilles New Caledonia Northern Mariana Islands Oman Puerto Rico Qatar San Marino Saudi Arabia Singapore Trinidad and Tobago Turks and Caicos Islands United Arab Emirates Virgin Islands (U.S.) Source: World Bank data. Note: OECD = Organisation for Economic Co-operation and Development. This table classifies all World Bank member-economies and all other economies with popula- tions of more than 30,000. Economies are divided among income groups according to 2009 gross national income per capita, calculated using the World Bank Atlas method. The groups are low income (LIC), $995 or less; lower-middle income (LMC), $996–$3,945; upper-middle income (UMC), $3,946–$12,195; and high income, $12,196 or more. GLOBAL MONITORING REPORT 2011 APPENDIX 1 173 Notes Kaufmann D, A. Kraay, and M. Mastruzzi. 2010. “The Worldwide Governance Indicators: 1. Statistical analysis is based on available data Methodology and Analytical Issues.” Policy as of January 2011. The maps in the report Research Working Paper 5430, World Bank, were created with updated data, as of end- Washington, DC. March 2011. Lall, S. 2000. “The Technological Structure and 2. Go and Quijada forthcoming. Performance of Developing Country Manufac- 3. Lofgren and Rodarte forthcoming. tured Exports, 1985–1998.” Working Paper 4. Long and Freese 2006. An alternative to the 44, Queen Elizabeth House, University of multinomial logit model would be the general- Oxford, Oxford, UK. ized ordered logit model, specifically propor- Lofgren, H., and I. Rodarte. Forthcoming. “Macro tional and partial proportional odds models. Analysis of Health and Education MDGs: Brief 5. World Bank 2004. Review and Country–Level Diagnosis.” Back- 6. Small and Hsiao 1985. ground note for Global Monitoring Report 7. The null hypothesis is rejected in only two 2011, World Bank, Washington, DC. cases, when testing the independence of out- Long, J. S., and J. Freese. 2006. Regression Mod- come 2 (off target and above average) in equa- els for Categorical Dependent Variables Using tion 4 (primary education) and outcome 3 (on Stata. 2nd edition. College Station, TX: Stata target) in equation 8 (access to clean water). Press. Our test results do not reject the assump- Marshall, M., and B. Cole. 2010. “Global Report tion of independence in the seven remaining 2009: Conflict, Governance, and State Fragil- specifications. ity.” Center for Global Policy, Washington, 8. Long and Freese 2006. DC. 9. McFadden 1973, p. 113. McFadden, D. 1973. “Conditional Logit Analysis 10. Marshall and Cole 2010. of Qualitative Choice Behavior.” In Frontiers 11. Kaufmann, Kraay, and Mastruzzi 2010. of Econometrics, ed. P. Zarembka, 105–42. New York: Academic Press. Small, K., and C. Hsiao. 1985. “Multinomial Logit Specification Tests.” International Eco- References nomic Review 26 (3): 619–27. Go, D., and A. Quijada. Forthcoming. “Assess- World Bank. 2004. Global Monitoring Report: ing the Odds of Achieving the MDGs.” Back- Policies and Actions for Achieving the Millen- ground paper for Global Monitoring Report nium Development Goals and Related Out- 2011, World Bank, Washington, DC. comes. Washington, DC. Appendix 2 Policies to Reduce Biodiversity Loss and Enhance Human Development Require a Concerted Approach B oth poverty and economic development use expansion and more intense use, global affect global biodiversity and the pro- biodiversity loss will continue unabated or vision of ecosystem goods and services. accelerate without additional policies.1 Larg- More food, water, and fi rewood are needed est losses are expected in Central and South to sustain population growth, especially in America, South Asia, and Sub-Saharan the poorer parts of the world. At the same Africa (map A2.1). time, expected rising levels of affluence in The importance of biodiversity for devel- emerging economies will add to the demand opment is recognized by Millennium Devel- for products like meat, construction tim- opment Goal 7, which includes targets to ber, and paper. When current technologies “reverse the loss of environmental resources” and consumption patterns prevail, increased and “reduce biodiversity loss.” However, global consumption by a larger and richer there is scant quantitative evidence on ways population will drive: in which especially poorer people depend on biodiversity, although patterns can be distin- • expansion of agriculture, forestry, and bio- guished. In general, it seems that poor people energy production; depend more on renewable natural resources • intensification of production, leading to than on biodiversity. However, the value of overexploitation and pollution from excess biodiversity aspects in terms of risk insurance, nutrients and contaminants; and ecosystem resilience, and larger area ecosys- • higher exploitation of remaining natural tem services (such as temperature regulation) ecosystems. still represents a large knowledge gap.2 In the context of rising demand for food, Traditionally, policies to reduce biodiver- water, and firewood, growing populations sity loss focus on area protection measures. that lack the capabilities to minimize environ- However, these measures have no effect on mental impacts of production will be more the unprotected surrounding areas. Further- susceptible to overexploiting and degrading more, expanding protected areas and reduc- their livelihoods. As a consequence of land ing deforestation would impose limits on 175 176 APPENDIX 2 GLOBAL MONITORING REPORT 2011 MAP A2.1 Average change in species populations, relative to the intact situation (mean species abundance or MSA), 2000 and 2050 IBRD 38470 APRIL 2011 2000 Greenland (Den) Faeroe Norway Iceland Islands Finland (Den) Sweden The Netherlands Russian Federation Estonia Canada Isle of Man (UK) Denmark Russian Latvia Fed. Lithuania United Ireland Kingdom Germany Poland Belarus Channel Islands (UK) Belgium Czech Rep. Luxembourg Slovak Rep. Ukraine Kazakhstan Austria Hungary Moldova Mongolia Liechtenstein France Slovenia B&HSerbia Romania Switzerland Croatia Andorra Montenegro Kos. Bulgaria Georgia Azer- Uzbekistan Kyrgyz D.P.R. Rep. United States Portugal Spain Monaco Albania FYRM Turkey Armenia baijan Turkmenistan Tajikistan of Korea Japan Italy Greece Syrian Gibraltar (UK) Cyprus Rep. of Tunisia Malta Lebanon Arab Islamic Rep. Afghanistan China Korea Israel Rep. Iraq of Iran Bermuda Morocco Kuwait (UK) West Bank and Gaza Jordan Bahrain Pakistan Nepal Bhutan Former Algeria Libya Arab Rep. Qatar The Bahamas St. Kitts and Nevis Spanish of Egypt Saudi Cayman Is. (UK) Dominican Antigua and Barbuda Sahara Arabia United Arab Bangladesh Mexico Cuba Rep. Dominica Emirates India Myanmar Lao Haiti St. Lucia Mauritania Oman P.D.R. Belize Jamaica St. Vincent and the Grenadines Cape Verde Mali Niger Rep. of Thailand Vietnam N. Mariana Islands (US) Guatemala Honduras Barbados Senegal Chad Eritrea Yemen Guam (US) El Salvador Nicaragua The Gambia Burkina Sudan Marshall Grenada Faso Djibouti Cambodia Philippines Islands Trinidad and Tobago Guinea-Bissau Guinea Benin Federated States of Costa Rica Panama Ghana Nigeria Ethiopia Sri Micronesia R.B. de Guyana Sierra Leone Côte Central Lanka Venezuela Suriname Liberia D’Ivoire African Rep. Brunei Cameroon Somalia Palau Colombia French Guiana (Fr) Togo Malaysia Equatorial Guinea Uganda Maldives Ecuador São Tomé and Príncipe Gabon Congo Rwanda Kenya Singapore Nauru Dem. Rep. of Burundi Seychelles Papua Solomon Congo Indonesia New Guinea Islands Tanzania Comoros Peru Brazil Timor-Leste Angola Zambia Malawi Mayotte (Fr) Vanuatu Bolivia Zimbabwe Mauritius Namibia Madagascar Paraguay Botswana Mozambique Réunion (Fr) New Swaziland MSA (%) Australia Caledonia (Fr) 100 South Lesotho Africa 90 Uruguay Chile Argentina 80 70 New Zealand 60 50 40 30 20 10 0 not included 2050 Greenland (Den) Faeroe Norway Iceland Islands Finland (Den) Sweden The Netherlands Russian Federation Estonia Canada Isle of Man (UK) Denmark Russian Latvia Fed. Lithuania United Ireland Kingdom Germany Poland Belarus Channel Islands (UK) Belgium Czech Rep. Luxembourg Slovak Rep. Ukraine Kazakhstan Austria Hungary Moldova Mongolia Liechtenstein France Slovenia B&HSerbia Romania Switzerland Croatia Andorra Montenegro Kos. Bulgaria Georgia Azer- Uzbekistan Kyrgyz D.P.R. Rep. United States Portugal Spain Monaco Albania FYRM Turkey Armenia baijan Turkmenistan Tajikistan of Korea Japan Italy Greece Syrian Gibraltar (UK) Cyprus Rep. of Tunisia Malta Lebanon Arab Islamic Rep. Afghanistan China Korea Israel Rep. Iraq of Iran Bermuda Morocco Kuwait (UK) West Bank and Gaza Jordan Pakistan Bhutan Bahrain Nepal Former Algeria Libya Arab Rep. Qatar The Bahamas St. Kitts and Nevis Spanish of Egypt Saudi Cayman Is. (UK) Dominican Antigua and Barbuda Sahara Arabia United Arab Bangladesh Mexico Cuba Rep. Dominica Emirates India Myanmar Lao Haiti St. Lucia Mauritania Oman P.D.R. Belize Jamaica St. Vincent and the Grenadines Cape Verde Mali Niger Rep. of Thailand Vietnam N. Mariana Islands (US) Guatemala Honduras Barbados Senegal Chad Eritrea Yemen Guam (US) El Salvador Nicaragua The Gambia Burkina Sudan Marshall Grenada Faso Djibouti Cambodia Philippines Islands Trinidad and Tobago Guinea-Bissau Guinea Benin Federated States of Costa Rica Panama Nigeria Ethiopia Sri Micronesia R.B. de Guyana Sierra Leone Côte Ghana Central Lanka Venezuela Suriname Liberia D’Ivoire African Rep. Brunei Cameroon Somalia Palau Colombia French Guiana (Fr) Togo Malaysia Equatorial Guinea Uganda Maldives Ecuador São Tomé and Príncipe Gabon Congo Rwanda Kenya Singapore Nauru Dem. Rep. of Burundi Seychelles Papua Solomon Congo Indonesia New Guinea Islands Tanzania Comoros Peru Brazil Timor-Leste Angola Zambia Malawi Mayotte (Fr) Vanuatu Bolivia Zimbabwe Mauritius Namibia Madagascar Paraguay Botswana Mozambique Réunion (Fr) New Swaziland MSA (%) Australia Caledonia (Fr) 100 South Lesotho Africa 90 Uruguay Chile Argentina 80 70 New Zealand 60 50 40 30 20 10 This map was produced by the Map Design Unit of The World Bank. 0 The boundaries, colors, denominations and any other information not included shown on this map do not imply, on the part of The World Bank Group, any judgment on the legal status of any territory, or any endorsement or acceptance of such boundaries. Source: PBL 2010, http://www.pbl.nl/en/publications/2010/Rethinking_Global_Biodiversity_Strategies. GLOBAL MONITORING REPORT 2011 APPENDIX 2 177 agricultural land expansion, pushing land nature conservation. International policy and food prices upward. This would espe- domains, including agricultural and for- cially affect urban people who depend on the estry sector policies, development assistance market for their food. Therefore, additional (including the role of international fi nancial measures should include smarter and better- institutions) and international trade provide managed land use, alternative consumption clear opportunities to integrate biodiversity patterns, and the development and applica- and ecosystem goods and services in their tion of technologies to increase production policies in ways that can support poverty efficiency per hectare. Measured by food reduction as well as sustainable use and con- prices, these routes would provide relief for servation of natural resources.3 urban poor, increasing food security and affordability. However, rural poor people are oftentimes particularly vulnerable to changes Notes in land values and uses. Potential increases 1. PBL 2010. in natural rents could affect rural poor for 2. CBD 2010. whom land entitlements are not adequately 3. Kok et al. 2010. arranged, making them vulnerable for dis- placement by larger landholders with access to technology and markets. References In conclusion, policies to reduce biodiver- CBD (Convention on Biological Diversity). 2010. sity loss and enhance human development “Linking Biodiversity Conservation and require a concerted approach. Some of these Poverty Alleviation: A State of Knowledge options are already being discussed at the Review.” Technical Series No. 55, Montreal, international level; others appear to be more Canada. contentious (changing consumption pat- Kok, M. T. J., S. Tyler, A. G. Prins, L. Pintér, terns), requiring careful consideration at the H. Baumüller, J. Bernstein, E. Tsioumani, national level, and will be dependent on vol- H. David Venema, and R. Grosshans. 2010. “Prospects for Mainstreaming Ecosystem untary approaches. This requires broadening Goods and Services in International Policies.” the scope of biodiversity policies on all levels Biodiversity 1 (1–2): 45–51. of decision making. Although most manage- PBL (Netherlands Environmental Assessment ment decisions affecting biodiversity and eco- Agency). 2010. Rethinking Global Biodiver- system goods and services are made at a local sity Strategies: Exploring Structural Changes level, these local decisions are conditioned by in Production and Consumption to Reduce national and international policies beyond Biodiversity Loss. Bilthoven, Netherlands. ECO-AUDIT Environmental Benefits Statement The World Bank is committed to pre- Saved: serving endangered forests and natural • 32 trees resources. The Office of the Publisher has • 10 million BTU of chosen to print Global Monitoring Report total energy 2011: Improving the Odds of Achieving • 3,075 pounds of CO2 the MDGs on recycled paper with 30 per- equivalent of cent post-consumer waste, in accordance greenhouse gases with the recommended standards for paper • 14,809 gallons of usage set by the Green Press Initiative, a wastewater nonprofit program supporting publishers • 899 pounds of in using fiber that is not sourced from solid waste endangered forests. For more information, visit www.greenpressinitiative.org. ow many countries are on track to achieve the Millennium socially excluded groups that are still very poor and often well H Development Goals (MDGs) by 2015? How many countries are off target, and how far are they from the goals? And what behind in reaching the goals. Moreover, progress could stall without stronger global growth, expanded access to export factors are essential for improving the odds that off-target markets for developing countries, and adequate assistance countries can reach the goals? This year’s Global Monitoring from donors. Report 2011: Improving the Odds of Achieving the MDGs This year’s report also presents findings and lessons from examines these questions. It takes a closer look at the diversity of impact evaluations in health and education to better country progress, presents the challenges that remain, and understand results on the ground. Such evaluations often assesses the role of growth, policy reforms, trade, and donor show that the quantity of services devoted to health and policies in meeting the MDGs. education has increased—but not the quality. This may be one Two-thirds of developing countries are on target or close to reason that progress toward those MDGs measured by being on target for all the MDGs. Among developing countries outcomes (as in health) is slower than it is for those MDGs that are falling short, half are close to getting on track. For those measured by access (as in education). Enhancing the efficiency, countries that are on track, or close to it, solid economic growth incentives, and accountability in service delivery is essential to and good policies and institutions have been the key factors in improving outcomes. their success. With improved policies and faster growth, many Global Monitoring Report 2011 is prepared jointly by the countries that are close to becoming on track could still achieve World Bank and the International Monetary Fund. It reviews the targets in 2015 or soon after. progress toward the MDGs and sets out priorities for policy Yet challenges abound. Even the middle-income countries responses, both for developing countries and for the on track to achieve the MDGs are home to indigenous and international community. ISBN 978-0-8213-8700-9 SKU 18700