62259 DIREC TIONS IN DE VELOPMENT Countries and Regions From Farm to Firm Rural-Urban Transition in Developing Countries Nora Dudwick, Katy Hull, Roy Katayama, Forhad Shilpi, and Kenneth Simler From Farm to Firm From Farm to Firm Rural-Urban Transition in Developing Countries Nora Dudwick, Katy Hull, Roy Katayama, Forhad Shilpi, and Kenneth Simler © 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 staff of the International Bank for Reconstruction and Development / The World Bank. The findings, interpretations, and conclusions expressed in this volume do not necessarily reflect the views of the Executive Directors of The World Bank or the governments they represent. The World Bank does not guarantee the accuracy of the data included in this work. The bound- aries, colors, denominations, and other information shown on any map in this work do not imply any judgment on the part of The World Bank concerning the legal status of any territory or the endorsement or acceptance of such boundaries. Rights and Permissions The material in this 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: 978-0-8213-8623-1 eISBN: 978-0-8213-8640-8 DOI: 10.1596/978-0-8213-8623-1 Cover photo: copyright © Ariadne Van Zandbergen. Library of Congress Cataloging-in-Publication Data From farm to firm : rural-urban transition in developing countries / Nora Dudwick ... [et al.]. p. cm. Includes bibliographical references and index. ISBN 978-0-8213-8623-1 — ISBN 978-0-8213-8640-8 (electronic) 1. Rural-urban migration—Developing countries. I. Dudwick, Nora. HB2160.F76 2011 307.2'4091724—dc22 2011008873 Contents Foreword xiii Acknowledgments xv Overview 1 WDR 2008 and WDR 2009: Rural and Urban Perspectives on Transformation 1 Building on the WDRs: Our Work Program 7 Main Findings 9 References 11 Part I Global, National, and Local Perspectives on Urbanization and Inclusive Growth: Examples from Africa 13 Chapter 1 Urbanization: Essential for Development 15 Urbanization and Inequality 17 Why Africa? 18 Organization of Part I 21 Notes 22 References 23 v vi Contents Chapter 2 Global Level: Recent Patterns and Trends in Rural and Urban Welfare Inequality 25 Sources of Data 28 Methods 30 Results 32 The Elusive Quest for Patterns 44 Conclusion 47 Notes 48 References 49 Chapter 3 National Level: Three Country-Level Perspectives on Rural-Urban Transitions 51 Structural and Spatial Transformations in Ghana, Mozambique, and Uganda 52 Rural and Urban Poverty Trends 57 Sources of Rural-Urban Inequalities 64 Methodology 65 Results 66 Implications for Policy 73 Notes 76 References 77 Chapter 4 Local Level: Intraurban Welfare Disparities and Implications for Development 79 Growing Cities 80 Lower Poverty, Better Services 81 Welfare Inequalities 82 Welfare Differences: Linked to Location 88 Understanding Spatial Inequalities 90 The Colonial Legacy 91 Failure to Strengthen Systems for Land Delivery 92 Conclusion 100 Notes 102 References 102 Part II Managing the Rural-Urban Transformation: Examples from South Asia 107 Chapter 5 Pathways Out of Poverty: Managing the Rural-Urban Transformation in South Asia 109 Contents vii Note 111 References 112 Chapter 6 Patterns of Rural-Urban Transformation in South Asia 113 Poverty Incidence and Geographical Distribution 114 Contribution of Agriculture to Growth and Employment 115 Urbanization, Nonagricultural Activities, and the Rural-Urban Transformation 118 Rural Nonfarm Activities and the Rural-Urban Transformation 120 Conclusion 121 Notes 122 References 123 Chapter 7 Policies, Institutions, and Initial Conditions 125 Economic Policies 125 Factor Markets: Land and Labor 128 Human Capital Endowments 131 Connective Infrastructures 134 Conclusion 138 Notes 139 References 139 Chapter 8 Land Market Institutions and the Rural-Urban Transformation 141 Impact of Land Inequality: The Existing Evidence 141 Impact of Land Market Restrictions: The Existing Evidence 147 Summary of Case Studies and Policy Implications 154 Notes 156 References 156 Chapter 9 Geographical Linkages and the Rural-Urban Transformation 159 Poverty, Urbanization, and Employment Structure along the Rural-Urban Continuum 159 Case Study 1: The Rural Employment Pattern and Urban Linkages 163 viii Contents Case Study 2: Agglomeration, Spillover, and Location of Nonfarm Activities 165 Notes 171 References 172 Chapter 10 Labor Mobility and the Rural-Urban Transformation 175 Internal Migration in South Asia 177 Internal Work Migration in Nepal 178 Destination Choice of Migrants 184 Notes 188 References 189 Chapter 11 Summary and Conclusions 191 Three Levels of Perspectives on Urbanization and Shared Growth: Sub-Saharan Africa 191 Managing the Rural-Urban Transformation: Examples from South Asia 195 Note 200 References 200 Appendix A Countries and Surveys Used for Consumption and Poverty Analysis 201 Appendix B Data, Methodology, and Results for Decomposition Analysis 205 Data 205 Model Specification 206 Oaxaca-Blinder Decomposition 207 Quantile Decomposition 208 Note 221 References 221 Index 223 Boxes 1 Country-Level Policy Recommendations Derived from the WDRs 5 1.1 Defining Urban 16 2.1 Are Rural-Urban Welfare Differences Shrinking or Growing? 27 Contents ix 2.2 Comparing Welfare Differences and Changes 31 4.1 Poverty and Slums 89 4.2 Postindependence Attitudes toward Urban Development 93 4.3 Corruption in Kampala 97 4.4 Slum Upgrading in Accra 99 4.5 Promoting Agglomeration through Regional Integration 101 6.1 Economic Density versus Density of the Poor 116 8.1 Inequality and the Evolution of Occupational Structure: Theoretical Insights 143 8.2 Evidence on the Effects of Land Reforms 149 10.1 The Recent Evidence on Migration: A Selected Review 176 10.2 Migrating to Economic Density in Brazil: Rational Decisions or Bright Lights? 188 Figures 1 WDR 2008’s Classification of Agriculture-Based, Transforming, and Urbanized Countries 2 2 Divergence, Then Convergence, in Rural-Urban Gaps 4 1.1 Urbanization and Per Capita GDP across Countries, 2000 17 1.2 Proportion of Urban Populations Living in Slums by Region, 2005 20 2.1 Cross-Sectional Data from WDR 2009 on Urban-Rural Consumption Differences 26 2.2 Rural-Urban Welfare Differences in Initial Survey Year, Selected Countries 34 3.1 Shares of GDP and Labor from Agriculture, Selected Countries 53 3.2 WDR 2008’s Classification of Agriculture-Based, Transforming, and Urbanized Countries 54 3.3 Sectoral Shares of GDP, Value Added: Ghana, Mozambique, and Uganda, 1980–2007 55 3.4 Population and Urbanization: Ghana, Mozambique, and Uganda, 1960–2007 56 3.5 Nonmonetary Welfare, Rural versus Urban: Ghana, Mozambique, and Uganda 62 3.6 Oaxaca-Blinder Decomposition of Urban-Rural Welfare Differences: Ghana, Mozambique, and Uganda, Selected Years 67 x Contents 3.7 Cumulative Distributions of Welfare (Household Consumption per Adult Equivalent): Ghana, Mozambique, and Uganda, Selected Years 71 4.1 Population Growth: Accra, Kampala, and Maputo, 1970–2009 80 4.2 Access to Services in Capitals versus Nationwide: Ghana, Uganda, and Mozambique, Selected Years 83 4.3 Access to Infrastructure Services by Wealth Quintile: Accra, Kampala, and Maputo 85 4.4 Stunting: Nationwide versus Capital City 87 6.1 Poverty Headcount Ratio in Lagging and Leading Regions, South Asia 115 6.2 Contribution of Agriculture to GDP Growth, South Asia 117 6.3 Agriculture’s Contribution to GDP and Employment, Sri Lanka 119 6.4 Contribution of Rural Nonfarm Sectors to Rural Income and Employment, South Asia 121 7.1 Average Tariff Rate (Unweighted), South Asia 126 7.2 Average Tariff Rate on Agricultural Products, South Asia 127 7.3 Nominal Rate of Assistance, South Asia 128 7.4 Difficulty of Hiring and Redundancy Costs, South Asia 130 7.5 Access to Health Care Facilities, Sri Lanka 134 7.6 Road Density, South Asia 135 7.7 Travel Times, Nepal 137 7.8 Travel Times, Sri Lanka 138 8.1 Area under Land Development Ordinance (LDO) Restrictions, Sri Lanka 150 8.2 Agriculture’s Share in Total Employment, Selected Countries in Asia 151 9.1 Poverty Headcount Ratio: Bangladesh, 2004–05 160 9.2 Poverty Headcount Ratio: Pakistan, 2004–05 160 9.3 Access to Services: Bangladesh, 2000 161 9.4 Access to Services: Pakistan, 2004–05 162 9.5 Percentage of Urban Population in Small, Medium, and Large Cities, Bangladesh and Pakistan 162 9.6 Shares of Nonfarm Employment and Enterprises by Area: Bangladesh, 2006 166 Contents xi 9.7 Annual Percentage Growth in Nonfarm Employment and Enterprises: Bangladesh, 2000–06 167 10.1 Out-migration of Workers by District: Nepal, 2001 179 10.2 In-migration of Workers: Nepal, 2001 180 10.3 Migrants and Nonmigrants by Age Category, Nepal 180 10.4 Education Level of Migrants and Nonmigrants, Nepal 181 10.5 Employment of Migrants and Nonmigrants, Nepal 181 10.6 Effects of Regressors on Destination Choice (Absolute Values), Nepal 187 Tables 2.1 Changes over Time in Absolute Levels of Welfare Measures, Selected Countries 37 2.2 Changes over Time in Welfare Measures (Kakwani Improvement Index), Selected Countries 41 2.3 Significance of Relationship between Convergence and Log GDP Per Capita 46 3.1 Growth Rates: Ghana, Mozambique, and Uganda, 1990s and 2000s 52 3.2 Ten Largest Cities: Ghana, Mozambique, and Uganda, Selected Census Years 58 3.3 Poverty, Inequality, and Population Distribution: Ghana, Mozambique, and Uganda, Selected Years 59 3.4 Sectoral Decomposition of Poverty: Ghana, Mozambique, and Uganda, Selected Years 60 4.1 Poverty in Capitals versus Nationwide: Ghana, Uganda, and Mozambique 81 4.2 Access to a Flush Toilet: Accra, Kampala, and Maputo, Selected Years 86 4.3 Access to a Shared Toilet: Accra, Kampala, and Maputo, Selected Years 86 4.4 Welfare Measures in Subareas of Greater Accra 90 6.1 Poverty, Inequality, and Economic Growth, South Asia 114 6.2 Selected Indicators for Contribution of Agriculture to Economy and Rural Poverty, South Asia 116 6.3 Selected Urbanization Indicators, South Asia 118 7.1 Selected Human Development Indicators, South Asia 132 7.2 Access to Education and Health Care Facilities in Rural Areas: Nepal, 2003 133 xii Contents 7.3 Selected Infrastructure Indicators, South Asia 136 8.1 Employment Choice of Individuals, Pakistan: Marginal Effects from Multinomial Logit Regressions 145 8.2 Real Wages, Pakistan: Individual-Level Instrumental Variable Regressions, Labor Force Survey, 2005–06 147 8.3 Sectoral Employment Shares in Sri Lanka, 2002 152 8.4 Effects of the Land Development Ordinance (LDO) on Employment Choice, Sri Lanka 152 8.5 Market Access, Land Tenure Arrangement, and Income, Sri Lanka 153 9.1 Urban Access, Cash Crop Potential, and Nonfarm Employment in Bangladesh: Multinomial Logit Results 164 9.2 Agglomeration Economies and Start-up Firms’ Location, Bangladesh 168 9.3 Spillover and Start-up Firms’ Location, Bangladesh 169 9.4 Productivity and Spatial Externality in Bangladesh: Regression Results from Non-Metro Investment Climate Survey Sample 170 10.1 Ethnic Profile of Migrants and Nonmigrants (Adult Males), Nepal 182 10.2 Comparing Actual Destination and Alternative Destinations of Migrants, Nepal 183 10.3 Consumption and Choice of Migration Destination: Male Migrants, Nepal 185 A.1 Countries and Surveys Used for Consumption and Poverty Analysis 202 B.1 Regression Results, Ghana 209 B.2 Regression Results, Mozambique 210 B.3 Regression Results, Uganda 212 B.4 Average Household Characteristics: Ghana, Mozambique, and Uganda, Selected Years 214 B.5 Decompositions of Rural-Urban Welfare Differences: Ghana, Mozambique, and Uganda, Selected Years 215 B.6 Decompositions of Welfare Differences over Time for Rural and Urban Areas: Ghana, Mozambique, and Uganda, Selected Years 218 Foreword The World Bank’s World Development Report 2009: Reshaping Economic Geography (WDR 2009) represented a turning point in the Bank’s thinking about the links among geography, poverty reduction, and growth. In pre- senting this new thinking, the book sparked a rich intellectual debate within the development community from which emerged the following paradigm- shifting notion: Spreading investment across a territory does not necessar- ily promote convergence in welfare across regions; rather, improving the welfare of poor people might be better achieved by allowing capital investment to flow to where it can be the most productive, while enabling people to move to those areas offering economic opportunities. Their mobility could be supported through a combination of human capital investment and connective infrastructure investments. In the same way that it pushed economists and policy makers alike in a new direction, WDR 2009 also generated a great deal of curiosity about how this conceptual framework could be applied to different regions, dif- ferent types of countries, and different situations. The two studies that make up this volume, one featuring Sub-Saharan Africa and the other South Asia, represent an effort to provide texture to the propositions and models set out in WDR 2009. Relying on development professionals from various Bank departments—those with experience in geography and xiii xiv Foreword urban and rural development policy, along with experts in poverty and inequality analysis and policy—these studies tested analytical methodolo- gies adapted to relatively data-poor economies, as well as those that could benefit from more sophisticated data foundations. The studies benefited from the guidance and advice of a group of experienced economists and policy experts both within and outside the Bank. Possibly more important from the viewpoint of their contribution to development policy, these studies focus attention and resources on the links between geography and poverty in the world’s poorest regions—Sub-Saharan Africa and South Asia—adding new value to the knowledge base created by WDR 2009. We are pleased to present this collection to the economists, social sci- entists, and policy makers concerned with issues of geography and devel- opment in poor countries, and we hope it will continue to spark useful and evidence-based debates aimed at reshaping economic geography. Jaime Saavedra Marisela Montoliu Acting Director Adviser and Former Head Poverty Reduction and Equity Spatial and Local Development Poverty Reduction and Economic Sustainable Development Management Network Network The World Bank The World Bank Acknowledgments The World Bank work program on rural-urban transformation, co-sponsored by the Poverty Reduction and Economic Management (PREM) Network and Sustainable Development (SD) Network vice presidencies, spanned two years, from mid-2008 to mid-2010, and benefited from the contribu- tions of development analysts and practitioners from the World Bank and other international organizations and academic institutions. The authors would like to thank the Poverty Reduction and Equity Group, led by Ana Revenga, sector director, Jaime Saavedra, sector man- ager, and Louise Cord, former sector manager, and the Finance, Economics, and Urban Department, led by Zoubida Allaoua, sector direc- tor, for their ongoing support of this work program. Special thanks go to Marisela Montoliu, who initiated the cross-sectoral work program as head of the Spatial and Local Development team in the SD Network. Her intellectual leadership and hands-on engagement throughout the research and writing process—including her own contribution of the overview chapter—enhanced the relevance and impact of the work program. The authors also gratefully acknowledge the substantial contributions of the panel of peer reviewers—Christopher Delgado, Paul Dorosh, Antonio Estache, Maria Emilia Freire, Chorching Goh, Steve Haggblade, Peter Lanjouw, and Stephen Mink—who provided substantive feedback xv xvi Acknowledgments and guidance throughout the process. The guidance of Louise Cord and Jaime Saavedra, and the feedback provided at different stages of the process by Judy Baker, Luc Christiaensen, Gaurav Datt, Marianne Fay, Gershon Feder, Louise Fox, Stephen Karam, Alexander Kremer, Somik Lall, Emmanuel Skoufias, Robert Townsend, and Hassan Zaman, con- tributed to the progress and quality of the studies. The input of country team staff was also important, and included comments from Sébastien Dessus, Chris Jackson, Dino Merotto, Antonio Nucifora, Uri Raich, and Rachel Sebudde. Shaohua Chen, Kalpana Mehra, Toru Nishiuchi, Prem Sangraula, Daan Struyven, Wei Xiao, and Haomiao Yu supported the data analysis and provided figures and graphs. Brian Blankespoor, Dana Thomson, and Emily Schmidt conducted the geographic information sys- tem (GIS) analyses, including production of maps. Paul Cahu contributed to the analysis and the overview chapter. Finally, the authors would like to acknowledge the organizational and logistical support provided by Alexander Arenas, Ramu Bishwakarma, Lanto Ramanankasina, and Grace Sorensen. Overview The transition from predominantly rural to increasingly urban economies is one of the greatest development challenges of these times. Urbanization spurs growth and reduces poverty, but also can engender inequalities. Managing the rural-to-urban transition in a way that ensures shared growth is therefore a major concern of developing country policy makers and the development community. Rural-urban transformations have been the focus of two consecutive World Development Reports (WDRs): WDR 2008, Agriculture for Development and WDR 2009, Reshaping Economic Geography (World Bank 2007, 2008). Our research agenda picks up where the WDRs left off by providing more finely tuned insights into the transformation process across rural and urban spaces. This overview distills some of the principal messages of the WDRs. It then outlines the main objectives of our work program before summarizing our key findings and their policy implications. WDR 2008 and WDR 2009: Rural and Urban Perspectives on Transformation The two most recent WDRs provided a starting point for our comple- mentary studies of the rural-urban transformation. WDR 2008, 1 2 From Farm to Firm Agriculture for Development, studies the structural transformation of the economy from a rural development perspective. WDR 2009, Reshaping Economic Geography, addresses the spatial transformations that accom- pany those structural changes—but it does so primarily from an urbaniza- tion perspective. WDR 2008 notes that developing countries are at very different stages in the transition from predominantly rural to largely urban-based economies. It classifies countries into three different groups: “agriculture- based,” “transforming,” and “urbanized” (see box 1 later in this overview). These classifications are far from static, although invariably agriculture contributes to a declining share of growth as countries develop. For exam- ple, over the course of two to three decades China and India changed from agriculture-based to transforming countries, Indonesia became increasingly urbanized, and Brazil developed into a full-fledged urbanized country (figure 1). The speed at which countries make the transition from largely agriculture-based to urbanized economies varies greatly. Likewise, the pace of change within countries may vary dramatically. Large countries, in particular, have regional disparities that replicate “the three worlds of Figure 1 WDR 2008’s Classification of Agriculture-Based, Transforming, and Urbanized Countries 80 agriculture-based BDI countries agriculture’s contribution to growth, 1990–2005, % RWA 60 CMR MWI SDN PRY BEN NER NGA CMR INDIA PNG 40 (1965–94) LAO BGR TGO GHA AZE CIV SYR MDG KEN BFA ETH NPL ZMB MOZ GINMLI TCD UGA 20 HND PAK IRN YEM KHM SEN DZA GTM CHINA BRA PHL BOL AGO IDN EGY BGD VNM UKR RUS DOM PER IND (1981–2001) ARG TUN VEN SVK TUR MAR THA LKA CHN CHL BRAZIL CZE SLV ECU INDONESIA ZAR 0 (1970–96) MEX POL ZAF MYS ROM (1970–96) TJK COL HUN BLR transforming countries urbanized countries ZWE –20 0.0 0.2 0.4 0.6 0.8 1.0 rural poor/total poor, 2002 actual poverty data predicted poverty data poverty data over time Source: World Bank 2007, fig. 1.3. Overview 3 agriculture.” Thus India, a “transforming” country, has agriculture-based states such as Uttar Pradesh and Bihar and urbanized states such as Goa. WDR 2009 further explores geographically uneven development, whether it is local, national, or international. It classifies areas according to whether their principal challenges are related to “density,” “distance,” or “division” (see box 1). According to these categories, China is best described as a “1-D” country, where the principal challenge is how to increase the density of economic activity. Brazil is characterized as a “2-D” country, where an additional challenge rests in increasing the con- nectivity between “leading” (mostly urban) and “lagging” (typically rural) regions. India is described as a “3-D” country, where internal divisions rep- resent a further challenge. The 1-D, 2-D, and 3-D distinctions can be applied as well on both narrower and broader geographical scales. For example, a 1-D town is one with insufficient agglomeration, and a 3-D continent is one facing the challenges of inadequate economic density, lagging regions, and major economic barriers. According to WDR 2009, urbanization is initially associated with a divergence of living standards between leading and lagging areas. Divergence occurs because the concentration of capital and labor brings with it productivity and cost advantages, resulting in higher wages in urban areas. Concentration also enables the provision of better infrastruc- ture and social services in towns and cities. WDR 2009 finds that rural- urban welfare disparities begin to converge once countries reach upper-middle-income levels (figure 2). Various phenomena explain this expected convergence: out-migration from rural areas reduces surplus labor; rural productivity benefits from technological advances; and gov- ernments become more sophisticated, expanding their fiscal bases and investing in rural infrastructure and services. Essential household con- sumption is predicted to converge first, followed by access to basic public services and, finally, wages and incomes. Within-city spatial disparities, manifested in slums, are predicted to last longer than rural-urban dispari- ties, persisting until countries reach high-income status. Both WDRs argue that when governments and others are considering how best to manage the rural-urban transition, they must tailor policies to the country’s (or area’s) stage of transformation. According to WDR 2008, in agriculture-based economies policies should focus on boosting produc- tivity in smallholder farming. In transforming economies, policies should also promote growth of the rural nonfarm sector. And in urbanized economies, policies should help to link farmers to modern food markets, agroindustry, and environmental services. WDR 2009 asserts that in 1-D 4 From Farm to Firm Figure 2 Divergence, Then Convergence, in Rural-Urban Gaps a. Locally, first divergence, then convergence, in rural-urban gaps ratio of urban to rural per capita 3 consumption 2 1 0 2 4 6 8 GDP per capita (2000 US$, thousands) b. Nationally, divergence, then convergence, in incomes between leading and lagging areas 0.6 coefficient of variation of regional 0.5 wages or income 0.4 0.3 0.2 0.1 0 0 2 4 6 8 10 12 14 16 GDP per capita (constant international Geary-Khamis $, thousands) Sweden, 1920–61 United States, 1840–1960 Spain, 1860–1975 Habsburg Empire, 1756–1910 Japan, 1955–83 United Kingdom, 1871–1955 c. Internationally, divergence, then convergence—but only in growing regions 2.0 coefficient of variation of 1.9 1.8 GDP per capita 1.7 1.6 1.5 1.4 1.3 1.2 1950 1960 1970 1980 1990 2000 2010 year Source: World Bank 2008. Overview 5 Box 1 Country-Level Policy Recommendations Derived from the WDRs WDR 2008 classifies countries according to the contribution of agriculture to eco- nomic growth and the share of the poor in the rural sector. In “agriculture-based” economies, agriculture contributes 20 percent or more to growth, and more than half of the poor live in rural areas. In “transforming” economies, agriculture con- tributes on average 7 percent to growth, but poverty remains overwhelmingly rural. In “urbanized” economies, agriculture contributes even less to overall growth, and poverty is mostly urban. WDR 2009 classifies countries according to the relative importance of “density,” “distance,” and “division.” Density refers to the economic output per unit of land area and is the principal challenge in “1-D” countries. Distance is mani- fested in densely populated lagging regions in “2-D” countries. Divisions—for example, between ethnic groups—are an additional obstacle in “3-D” countries. The seven countries in Sub-Saharan Africa and South Asia featured in our study would fit into the categories shown in the table. WDR 2009 1-D countries 2-D countries 3-D countries WDR 2008 Agriculture- based Nepal economies Ghana Mozambique Uganda Transforming Bangladesh economies Sri Lanka Pakistan Urbanized economies Based on these classifications, the WDRs might suggest the following policy responses. For Ghana, an agriculture-based economy, WDR 2008 might recom- mend public investments to promote agricultural growth and food security. And because the principal challenge on a national level is one of insufficient density, WDR 2009 might recommend instigating spatially neutral institutions such as improved land policies or country-wide education and health initiatives. (continued next page) 6 From Farm to Firm Box 1 (continued) In Nepal, another agriculture-based country, a relatively large share of the total population is in lagging and poorly connected regions. It thus faces issues of both distance and density. In addition to the policy menu proposed for Ghana, WDR 2009 might therefore recommend increased investments in connective infrastructure. Bangladesh, Mozambique, Pakistan, and Uganda are all on the brink of entering the transforming economy category. Thus WDR 2008 might call for interventions to boost rural nonfarm productivity through investments in human capital and nonagricultural activities in small- and medium-size towns. Free from major cultural divides but with significant barriers to mobility, Bangladesh and Mozambique qualify as 2-D countries that, as WDR 2009 might suggest, would benefit from connective infrastructure in addition to spatially unbiased institutions. By contrast, the ethnic, cultural, or political divisions present in Pakistan and Uganda might justify territorially targeted programs, in addition to aspatial institutions and connective infrastructure. For Sri Lanka, the only country in our sample that fits squarely in the trans- forming economy category, WDR 2008 might recommend programs to promote nonfarm activities in small and secondary towns. WDR 2009 might call for sound, spatially blind policies and investments in connective infrastructure. Because of the different perspectives of these two WDRs, some of the broad policy implications emerging from them may seem at odds. For example, for low- density, agriculture-based economies WDR 2008 might recommend targeted investments in rural areas, whereas WDR 2009 might suggest that spatially neutral investments be the main priority. But both WDRs would concur that a detailed country-level analysis should guide the specifics of policy design and that a nar- rower lens—for example, a focus on a particular region or city within each country—could unearth more complex challenges. countries policies should focus on the provision of “spatially blind” institu- tions, such as sound macroeconomic policies, broadly distributed social services, and good land laws. In 2-D countries, governments should also implement connective infrastructure—such as good roads—so that the benefits of agglomeration economies can be shared as widely as possible. And in 3-D countries, institutions and infrastructure should be supple- mented with spatially targeted interventions in order to overcome ethnic and cultural divisions. Box 1 indicates how the policy recommendations of the two WDRs might be applied to the countries featured in our studies. Overview 7 Building on the WDRs: Our Work Program Our work program builds on the findings of the WDRs on the rural- urban transformation, focusing specifically on Sub-Saharan Africa and South Asia—two regions that pose particularly urgent issues for policy makers. Although each region is still largely agricultural, each is under- going rapid rates of urbanization. In fact, cities within the two regions have the highest prevalence of slums in the world: UN-HABITAT (2008) estimates that more than 6 out of 10 city dwellers in Sub- Saharan Africa and 4 out of 10 in South Asia live in slums. These slums are a powerful reminder that even though the rural-urban transforma- tion may be good for national development, it can incur inequitable out- comes for households. Our country-level focus helps to add nuance to the broad typologies provided in each WDR. For example, with agriculture contributing 35 percent to its gross domestic product (GDP), Nepal is classified squarely as an agriculture-based country. But a narrower lens indicates that the country is in fact a patchwork of largely urbanized areas, such as the Kathmandu Valley, and almost entirely rural areas, such as the eastern hills region. Similarly, although Mozambique as a whole is making the transition from an agricultural to an urban economy, the pace of change is by no means even. Its capital, Maputo, in the south, is growing at a very fast pace, but a significant share of the poor population remains in the rural northern regions. WDR 2008 has a rural focus and WDR 2009 an urban one. Although by virtue of their subject matter both reports acknowledge the interde- pendence of rural and urban areas, they sometimes overlook distortions and spillovers that flow from rural to urban spaces, and vice versa. We aim to look across the rural-urban continuum to consider the interaction between rural and urban transformations. For example, what are the push and pull factors shaping migration decisions? And how do institu- tions such as land laws affect transformation in both rural and urban areas? Our approaches are interdisciplinary, demonstrating how analytical methods can be tailored to the quantity and quality of data available. In relatively data-poor countries, such as many of those of Sub-Saharan Africa, it makes more sense to analyze patterns and trends in rural-urban dispari- ties using decomposition tools rather than attempt to address causality. In more data-rich countries, such as those of South Asia, regression-based tools can provide indications of causality. 8 From Farm to Firm Our studies, as reported in this volume, use three different levels of analysis—global, national, and local—to look at the patterns and trends in the ongoing rural-urban transformation and the reasons for the uneven economic landscapes. The first level of analysis is global, encompassing more than 40 low- and middle-income countries. Taking a global perspec- tive (chapter 2), we examine rural-urban welfare differences, using both monetary and nonmonetary measures. We then use cross-sectional sur- veys to assess the degree of convergence or divergence in rural and urban living standards as national income and the urban share of the population increase over time. Finally, we look at covariates to help explain why spa- tial inequalities are increasing in some countries and decreasing in others. We then move to the national level of analysis (chapter 3), homing in on rural-urban differences in three African countries—Ghana, Mozambique, and Uganda. For the national-level analysis, we use Oaxaca- Blinder decomposition techniques to determine whether rural-urban inequalities stem from differences in endowments, such as levels of educa- tion, or from differences in returns to those endowments. Put simply, this technique helps us to understand whether rural-urban welfare disparities in these three African countries stem from people or from places. Moving on from the national-level analysis, we address the rural-urban transformation from the perspective of intraurban and intraregional dif- ferences. After the discussion of the rural-urban transformation in Sub- Saharan Africa, we consider patterns and trends in welfare within three African cities—Accra, Kampala, and Maputo—investigating the relation- ship between inequalities in living standards and location within these cities (chapter 4). We then look at the causes of urban inequalities, focus- ing especially on colonial legacies, land policies, and weak governance. In part II of this report, we return to the national level of analysis but use examples of countries in South Asia. Chapter 5 introduces the regional perspective on managing the rural-urban transformation, and chapter 6 looks at the patterns in South Asia—specifically Bangladesh, Nepal, Pakistan, and Sri Lanka. This case study approach, employing both regression-based tools and qualitative analysis, enables us to look beyond the question of “people or places” to uncover in chapter 7 the effects of policies and institutions—whether land distribution and property rights, or access to infrastructure and services—on rural-urban transformations in the four countries. We then move to describing land market institutions (chapter 8), geographical links (chapter 9), and labor mobility (chapter 10) in relation to the rural-urban transformation. Chapter 11 presents our summary and conclusions. Overview 9 Main Findings Many of the key findings in the studies described here are complemen- tary. First, they confirm that, on average, urban areas enjoy a higher stan- dard of living than rural areas. But beyond this broad (and far from surprising) finding, they discover a high level of variation. For example, chapter 2 finds that the size of the rural-urban gap is highly variable across the 41 developing countries, especially among the poorest coun- tries. And even in countries with the same levels of urbanization or GDP per capita, the rural-urban differences may be enormous or relatively small. Chapter 6 reveals that within-country differences in economic structure and well-being can be even larger than the differences between countries. For example, even in Sri Lanka, the most urbanized country in South Asia, the poorer regions are more than 15 hours away from the nearest medium-size city. Like the two WDRs, our studies confirm the existence of bumpy eco- nomic landscapes. But they also suggest that patterns of convergence and divergence may not be neatly tied up with a country’s level of develop- ment, at least within the observable time horizon for countries in Africa and South Asia. For example, our cross-country analysis finds little evi- dence to suggest that welfare differences are either consistently expand- ing or consistently diverging as GDP per capita grows. And both our Africa and Asia studies suggest that more than the level of development, the kinds of policies and institutions that a country has in place as it develops will determine the texture of the economic landscape. Our report echoes the call by WDR 2009 for spatially neutral institu- tions as a sine qua non of the effective management of the rural-urban transformation. Land institutions emerge as particularly important. For example, the case study on Sri Lanka shows how restrictions on the sale of rural land increase the cost of migration, thereby impeding rural-urban transitions. Our study of three African cities also reveals how inadequate legal frameworks, poor land management, and corruption have combined to create extremely tight markets for formal land and housing, contribut- ing to the precipitous growth of slums. But our studies also indicate that land is a very difficult area to reform from a political economy perspec- tive. According to the case studies on Pakistan and Sri Lanka, if political economy factors make full-fledged land policy reform impossible, inter- mediate measures—such as greater access to credit and improved effi- ciency of land markets—can mitigate the negative impact of land tenure conditions on the rural-urban transformation. 10 From Farm to Firm Like WDR 2009, our studies also highlight the importance of spatially unbiased access to services, in particular basic education. In the African countries analyzed, human capital disparities play a bigger role than dif- ferences in returns to education in accounting for the observed gaps in rural-urban welfare. Similarly, regression results for the South Asia coun- tries confirm that people with higher levels of education can reap consid- erable positive returns in the labor market. Thus, the studies of both Sub-Saharan Africa and South Asia suggest that nationwide human capi- tal standards could be powerful instruments for narrowing rural-urban disparities. Similarly, uneven access to services is found to underpin spatial divides, whether they are between rural and urban areas, smaller and larger towns, or different parts of the same city. Our study of three African countries finds that rural-urban migration choices are based not only on income prospects but also on prospective access to schools and health care cen- ters. Meanwhile, our study of three African cities reveals that the uneven living standards across income quintiles are closely correlated with the level of services people can access in their neighborhood. Combined, these findings suggest that policies to broaden access to basic services would both remove pressures on congested metropolises and contribute to more geographically balanced development. Both WDRs suggest that policies should be carefully sequenced to a country or area’s level of development. But our studies underscore that the appropriate policy mix will depend on the unit of analysis. For exam- ple, at the country level Ghana is indeed a 1-D country that would ben- efit, first and foremost, from spatially blind institutions (see box 1). But at the city level, Accra faces vast socioeconomic divides that might call for the kind of spatially targeted interventions recommended for 3-D areas. Moreover, our studies suggest that because of observed interactions across policies, a combination of measures can often have a much stronger impact on the rural-urban transformation than a sequence of policies. For example, we find that the benefits of infrastructure investments are larger in areas with better agricultural potential and fewer restrictions in factor markets. These studies serve as a reminder that, although the rural-urban dis- tinction is helpful from an analytical perspective, a rural-urban continuum may be more useful from a policy perspective, because interventions are likely to have second-round impacts in areas beyond their immediate geographical target. For example, in Pakistan the impact of rural land inequality is felt in the higher pace of migration and lower wages for Overview 11 unskilled workers in towns and cities. In Sri Lanka, although public investment in irrigation has satisfied an immediate goal of boosting agri- cultural productivity, it has resulted in less employment diversification, slower migration, and lower overall living standards. Meanwhile, our study of three African cities suggests that not all policies to address urban inequalities should target urban areas—for example, introducing better services in rural areas would reduce the push factors that drive people to cities in the first place. The implications that emerge from our work are only some of the many considerations facing policy makers. Above all, leaders often face restrictive political environments that limit the likelihood that many of the “first best” solutions will be implemented—political economy issues we only touched on in our studies. We have, however, demonstrated the high cost of inaction: a failure to institute reforms is impeding, and will continue to impede, the extent to which everyone can share in the eco- nomic benefits of rural-to-urban transformations. We hope that these studies will help to tip the political economy in favor of change. References UN-HABITAT (United Nations Human Settlements Programme). 2008. State of the World’s Cities 2008/2009: Harmonious Cities. Nairobi: UN-HABITAT. World Bank. 2007. World Development Report 2008: Agriculture for Development. Washington, DC: World Bank. ———. 2008. World Development Report 2009: Reshaping Economic Geography. Washington, DC: World Bank. PA R T I Global, National, and Local Perspectives on Urbanization and Inclusive Growth: Examples from Africa CHAPTER 1 Urbanization: Essential for Development The world is becoming increasingly urbanized.1 Indeed, half of humanity is now living in urban areas, and more than 70 percent of the populations of Latin America, North America, and Europe are living in cities (see box 1.1 for definitions of the term urban). Although 60 percent of the people in Sub-Saharan Africa still live in rural areas, it is the fastest- urbanizing region of the world. In fact, the United Nations predicts that by 2030 Africa will be a predominantly urban continent. Urbanization stems in part from rural-to-urban migration, in part from natural increase, and in part from the reclassification of urban boundaries as cities expand outward and small population centers grow and are designated as urban areas. Urbanization is therefore happening, and it will continue to happen. This trend is inevitable because urban development is an integral part of economic development. Economic growth is invariably accompanied by a transition from a predominantly agrarian economy to an economy dom- inated by the production of nonagricultural goods and services. Although some of this transformation can take place in situ as the rural nonfarm economy grows and diversifies, the overriding pattern is one of increasing urbanization. Firms take advantage of agglomeration economies (sharing of infrastructure, better matching of workers to jobs, and knowledge 15 16 From Farm to Firm Box 1.1 Defining Urban There is no uniform global definition of urban. The United Nations (2004) argues that “given the variety of situations in the countries of the world, it is not possible or desirable to adopt uniform criteria to distinguish urban areas from rural areas.” Statistical definitions of urban areas vary from country to country (or even within countries) and can be based on administrative boundaries, size, level of services, or population density (World Bank 2008). This report delineates rural and urban areas following the classification used by the Population Division of the UN Secretariat in its regular report World Urban- ization Prospects (WUP)—see United Nations (2005). Thus rural areas and urban areas, and rural and urban populations, are defined according to the criteria used by each country. Details on data sources and methods used for specific countries are available on the WUP Web site (http://esa.un.org/unup/). This classification, although admittedly imperfect, is an attempt to balance competing demands. In choosing it, we give considerable weight to the repre- sentativity of the household survey data that are the basis for our empirical analysis. All of these surveys were designed to be nationally representative, as well as representative of rural and urban areas, according to the classification sys- tem used in each country. In principle, it is possible to reclassify areas using more rigorous and consistent criteria, such as the agglomeration index developed for the 2009 World Development Report (World Bank 2008). Such a reclassification would have many benefits, but in the context of this report those benefits would be offset by an enormous cost: the survey data would no longer be statistically representative of rural or urban areas, which in turn casts doubt on inferences made from the observed rural-urban welfare differences. In addition to consistency issues across countries, simple categorization within a country inevitably involves a loss of information. In reality, there is a rural- to-urban continuum, ranging from sparsely populated isolated settlements to small towns to secondary cities to megacities. Thus, in any given country there is heterogeneity within areas that are classified as rural or urban. Although a few countries stratify their household surveys into three or four levels of the rural- urban continuum (e.g., rural, urban, and metropolitan in Brazil), they are the exceptions. Urbanization: Essential for Development 17 Figure 1.1 Urbanization and Per Capita GDP across Countries, 2000 100 80 urban (%) 60 40 20 10 0 10,000 20,000 30,000 40,000 GDP per capita (US$) Source: Annez and Buckley 2008. Note: GDP = gross domestic product. spillovers), which leads to what Arthur (1990) describes as “positive feedback”—a mutually reinforcing relationship between increases in pro- ductivity and the concentration of firms. Similarly, people concentrate to take advantage of higher-paying employment opportunities, better prices because of denser markets, and improved amenities. The suggestion that urbanization is a requirement for economic devel- opment is borne out by the empirical evidence: few countries have realized income levels of US$10,0002 per capita before reaching about 60 percent urbanization (see figure 1.1). And simple bivariate regressions, though no indication of causality, suggest that urbanization is a very strong indicator of productivity growth over the long run (Annez and Buckley 2008). Urbanization and Inequality During the early stages of development, economic growth and urbaniza- tion tend to increase spatial inequalities. Agglomeration economies create leading areas, characterized by economic dynamism and rising standards of living. Most often, those who are already better-off—that is, have more education and more assets to fall back on—are better placed to take advantage of these new opportunities. But the flip side of leading areas is lagging areas—areas that have not experienced structural transformation and where standards of living are stagnating or even declining. Spatial inequalities exist at many levels. At the international level, early industrialization has concentrated production in Western Europe and North America, leaving entire regions of the world behind. At the country level, large portions of countries such as northern Ghana or northern 18 From Farm to Firm Uganda are considered lagging areas. Disparities also appear within such regions. For example, low living standards have been observed in rural areas of southern Mozambique, only 100–200 kilometers from the boom- ing capital city of Maputo. Drilling down further, even very prosperous cities have squalid slums. At the country level, because of the nature of structural transforma- tion, inequalities between leading and lagging areas often correspond to rural-urban divides. World Development Report 2009: Reshaping Economic Geography (WDR 2009) argues that rural-urban living standards diverge as countries develop and become more urbanized, converging only once they reach a relatively high development threshold. Specifically, it finds that “urban-to-rural gaps in consumption levels rise until countries reach upper-middle-income levels” (World Bank 2008). But policy makers cannot afford to sit back and wait for their countries to pass a hypothesized development threshold before spatial inequalities begin to converge, especially when that threshold lies far in the future. Spatial inequalities entail less inclusive growth processes—that is, less poverty reduction for a given rate of growth—and are often associated with social tensions and even conflict, particularly when spatial inequali- ties are aligned with ethnic, linguistic, or religious divides (Kanbur and Venables 2005). Therefore, a key question emerges: How can developing countries manage the transition from a predominantly rural economy to a more urbanized economy in a way that produces acceptable growth and equity outcomes in the short and medium term? Part I of this report aims to help the development community gain insights into this question of how to manage the urbanization process in a way that preserves growth and promotes equity. It complements recent and forthcoming publications on spatial inequalities at the global and regional levels (the regional level, using examples from South Asia, is cov- ered in part II of this report). In contrast to the long-term view provided by WDR 2009, part I presents a more medium-term picture of patterns and trends in spatial inequalities. In painting that picture, part I focuses on Sub-Saharan Africa. Why Africa? The focus on Africa in part I is intended in part to address questions of African exceptionalism. The notion that Africa, contrary to other regions, has experienced “urbanization without growth” (Fay and Opal 2000) is probably overstated, holding true only for small African countries at low Urbanization: Essential for Development 19 levels of urbanization or failed states that are experiencing “push” as opposed to “pull” rural-to-urban migration (Annez and Buckley 2008). Even though the “urbanization without growth” paradigm is misleading, the region has experienced urbanization with lower levels of shared growth. Ravallion, Chen, and Sangraula (2007) point out that urbaniza- tion in Sub-Saharan Africa has been associated with less poverty reduc- tion than in other regions. And, alongside Latin America, Sub-Saharan Africa is experiencing the highest levels of urban inequality in the world.3 Thus higher average levels of well-being in urban areas may mask high levels of deprivation among poor urban residents. According to UN- Habitat (2008) estimations, a higher proportion of city dwellers in Sub-Saharan Africa live in slums than anywhere else in the world (figure 1.2). Although the slum classification says nothing about the depth of deprivation, examination of conditions across regions by UN-Habitat’s Global Urban Observatory suggests that slum dwellers in Sub-Saharan Africa experience a higher level of multiple shelter deprivations compared with slum dwellers in other regions (UN-Habitat 2008). Meanwhile, African policy makers appear to be particularly ambivalent about the process of rural-urban transformation: a UN survey of popula- tion policies indicates that 83 percent of African governments are imple- menting policies to reduce rural-to-urban migration, and 78 percent are intervening to reduce migration to large cities in particular. These percent- ages are notably higher than the world averages (United Nations 2008a). Because the urban population of Africa will almost double over the next two decades, growing from about 290 million in 2007 to 540 million in 2025 (United Nations 2008b), the question “Is Africa’s urbanization dif- ferent, and if so, why?” is of growing urgency from a policy perspective. Why focus here on three particular countries in Sub-Saharan Africa— Ghana, Mozambique, and Uganda? These countries are not intended to be representative of the entire region, but they do bear certain common characteristics that enable us to draw some indicative conclusions about processes of structural and spatial transformation. First, they are all at an early stage of economic development and structural transformation, albeit to slightly different degrees. Second, all three countries have experienced urbanization over the course of the last three decades. And third, each has experienced robust growth in its gross domestic product (GDP) over the past 15–20 years, which enables us to observe the trends in spatial inequality associated with economic development.4 After looking at rural-urban disparities in Ghana, Mozambique, and Uganda, we narrow the lens of analysis further to examine intraurban 20 Figure 1.2 Proportion of Urban Populations Living in Slums by Region, 2005 EASTERN ASIA SOUTHERN ASIA 36.5 NORTHERN AFRICA 42.9 14.5 WESTERN ASIA 24.0 SOUTH-EASTERN ASIA 27.5 OCEANIA LATIN AMERICA AND SUB-SAHARAN 24.1 OCEANIA AFRICA 24.1 THE CARIBBEAN 27.0 62.2 PERCENTAGE OF THE URBAN POPULATION LIVING IN SLUMS AS OF 2005 0 − 15 15 − 25 25 − 35 AUGUST 2010 IBRD 38032 35 − 45 > 45 Data source: UN-Habitat 2008. Urbanization: Essential for Development 21 inequalities in the capitals of these countries: Accra, Maputo, and Kampala. Three observations informed a decision to focus on these capi- tal cities. First, across Sub-Saharan Africa, large cities (of between 1 and 5 million) have been growing at a faster rate than medium-size and small cities and thus account for a rising share of Africa’s urban population. Growth in the primary city, it seems, will represent the dominant trend for the foreseeable future. Second, spatial inequalities in big cities tend to be larger than in smaller cities (Kilroy 2007). And third, the greater quan- tity of data available for capital cities as opposed to secondary cities in each of these countries allow a more thorough investigation of patterns and trends across income and nonincome welfare measures. Organization of Part I The chapters that follow investigate urbanization and rural-urban welfare inequalities on three geographic scales: global, national, and local. At the global level, we examine in chapter 2 cross-country evidence of rural-urban welfare inequalities in 41 low- and middle-income countries. Specifically, we look at differences in rural and urban welfare based on both monetary and nonmonetary measures. We then use household sur- vey data to uncover patterns of divergence or convergence over time. Finally, we investigate whether these patterns are associated with other covariates, such as per capita GDP. The findings indicate that investiga- tors should be cautious in predicting welfare convergence or divergence based on a country’s level of development alone. Although urban welfare is consistently higher than rural welfare, the magnitude of the gap is extremely variable, even in countries with similar levels of urbanization or per capita GDP. Over time, few clear patterns of convergence or diver- gence emerge. Rather than reinforcing the broad prediction in WDR 2009 of convergence-then-divergence as countries develop, the findings indicate that deeper analysis of the institutional and policy environment is required to understand the dynamics of rural-urban transitions at the country level. Chapter 3 therefore takes the analysis to the national scale as we try to uncover the sources of rural-urban consumption inequalities. Focusing on Ghana, Mozambique, and Uganda, we investigate why urban welfare is higher than rural welfare in these three countries. We use decomposi- tion techniques to determine whether these inequalities stem from dif- ferences in endowments (such as levels of education) or differences in returns to those endowments. In other words, would the rural poor, by 22 From Farm to Firm virtue of their household characteristics, remain poor if they lived in urban areas? Or would a move to the city entail higher returns, all other things being equal? Differences in endowments emerge as the principal explanation of rural-urban inequalities, particularly in Uganda and Mozambique. These results indicate that barriers to mobility are not the principal explanation of welfare inequalities between rural and urban areas and suggest that measures to boost rural education would have a significant impact on spatial inequalities. To gain a yet more finely tuned picture, we consider in chapter 4 dis- parities at the local level in Accra, Maputo, and Kampala, relying on data on demographic and physical changes in these three capital cities. We then look at income and nonincome welfare in the capitals versus the countries as a whole before examining the extent of intraurban inequali- ties. In the final section, we investigate the principal causes of spatial inequalities in these cities, highlighting colonial legacies, dysfunctional land markets, and weak governance. Concluding chapter 11 of this report pulls together the findings of each of the levels of analysis to reflect on policy implications and out- lines an agenda for future research. The findings reinforce the call in WDR 2009 for spatially blind institutions to promote equality of oppor- tunities in rural and urban areas. But the findings also question the extent to which such institutions can be effective without concurrent investments in connective infrastructure and carefully targeted interven- tions, even in relatively low-income environments. Avenues for future research include more spatially sensitive poverty analysis and further investigation into the impact of land tenure on rural-urban transitions. Urbanization may be part and parcel of development, but concerted pol- icy making—based on sound analytics—is required to ensure that it incorporates shared growth and opportunities for all. Notes 1. Part I of this report was produced by a team led by Kenneth Simler and Nora Dudwick of the Poverty Reduction and Equity Group (PRMPR) under the Poverty Reduction and Economic Management (PREM) vice presidency. Team members were Paul Cahu, Katy Hull, Roy Katayama, and Kalpana Mehra. The chapters were prepared under the supervision of Louise Cord and Jaime Saavedra. This work is part of a joint work program between PRMPR and FEUSE (Finance, Economics and Urban Department, Spatial and Economics Unit). The FEUSE team was led by Forhad Shilpi and supervised by Marisela Montoliu. Urbanization: Essential for Development 23 2. All dollar amounts are U.S. dollars unless otherwise indicated. 3. According to UN-Habitat (2008), the two regions exhibit “exceptionally high levels of urban inequality”: a subset of 26 cities in Africa and 19 in Latin America have average Gini coefficients of 0.54 and 0.55, respectively. 4. An additional reason for choosing these countries is the availability of repeated cross sections of good-quality household survey data. References Annez, P. C., and R. M. Buckley. 2008. “Urbanization and Growth: Setting the Context.” In Urbanization and Growth, ed. M. Spence, P. C. Annez, and R. M. Buckley. Commission on Growth and Development. Washington, DC: World Bank. Arthur, B. W. 1990. “Positive Feedbacks in the Economy.” Scientific American 262: 92–99. Fay, M., and C. Opal. 2000. “Urbanization without Growth: A Not So Uncommon Phenomenon.” Policy Research Working Paper 2412, World Bank, Washington, DC. Kanbur, R., and A. Venables. 2005. “Rising Spatial Disparities and Development.” Policy Brief No. 3, United Nations University, World Institute for Development Economics Research, Helsinki. Kilroy, A. 2007. “Intra-Urban Spatial Inequality: Cities as ‘Urban Regions.’” World Development Report 2009 Background Paper, World Bank, Washington, DC. Ravallion, M., S. Chen, and P. Sangraula. 2007. “New Evidence on the Urbanization of Global Poverty.” Policy Research Working Paper 4199, World Bank, Washington, DC. UN-Habitat. 2008. State of the World’s Cities 2008/9: Harmonious Cities. Nairobi: United Nations Human Settlements Programme. United Nations. 2004. World Urbanization Prospects: The 2003 Revision. Department of Economic and Social Affairs, Population Division. New York: United Nations. ———. 2005. World Urbanization Prospects: The 2004 Revision. Department of Economic and Social Affairs, Population Division. New York: United Nations. ———. 2008a. Urban Indigenous Peoples and Migration Factsheet. http://www .un.org/esa/socdev/unpfii/documents/factsheet_migration_final.pdf. ———. 2008b. World Urbanization Prospects: The 2007 Revision. Department of Economic and Social Affairs, Population Division. New York: United Nations. World Bank. 2008. World Development Report 2009: Reshaping Economic Geography. Washington, DC: World Bank. CHAPTER 2 Global Level: Recent Patterns and Trends in Rural and Urban Welfare Inequality In almost every country in the world, average living standards in urban areas are superior to those in rural areas. This pattern is observed whether welfare is measured by average income, consumption, poverty indexes, infant mortality, health, access to services, or numerous other variables. Likewise, the superiority of average urban living standards is the norm regardless of national income levels and tends to be maintained during the development process, even as countries move from predominantly rural agrarian economies to more urbanized economies with larger indus- trial and service sectors. However, the size of rural-urban welfare gaps varies a great deal across countries, and some argue that the evolution of these gaps follows a pre- dictable pattern as a country grows and develops. According to the World Development Report 2009: Reshaping Economic Geography (WDR 2009) (World Bank 2008), the rural-urban gap in consumption increases until countries reach upper-middle-income levels, or $4,000–$12,000 in gross national income per capita. After that point, the rural-urban gap is expected to diminish as the urban share of the population grows, the domestic product and factor markets become more efficient, and the delivery of public services becomes more equal over space. As described 25 26 From Farm to Firm in WDR 2009, in industrial economies, rural-urban welfare differences are small, but not completely eliminated. The development trajectories of today’s industrialized countries con- form well to the paradigm of rural-urban divergence-then-convergence in living standards. The current low-income countries may follow the same path, and yet the available data also pose some critical questions, as illus- trated in figure 2.1 taken from WDR 2009. The figure shows the ratios of urban-rural per capita consumption as revealed by 120 cross-sectional household surveys conducted in 75 countries. On the right-hand side of the distribution, the urban-rural consumption ratio drops from about 2.2:1 to about 1.6:1. The left-hand side of the distribution in figure 2.1 is almost L-shaped. The bottom part of the L reveals greater urban-rural inequality accompanying higher levels of gross domestic product (GDP) (in the cross section), as described in WDR 2009. However, along the ver- tical part of the L are the poorest countries, all with per capita GDP of less than $1,000 (not adjusted for purchasing power parity), in which average urban consumption is anywhere from 150 percent to over 300 percent of rural consumption. What accounts for this large range of urban-rural inequalities among countries at roughly similar stages of development? In this chapter, we seek to complement WDR 2009, and other recent studies summarized in box 2.1, by examining more closely rural-urban Figure 2.1 Cross-Sectional Data from WDR 2009 on Urban-Rural Consumption Differences ratio of urban to rural per capita 3 consumption 2 1 0 2,000 4,000 6,000 8,000 GDP per capita (2000, US$) Source: World Bank 2008. Global Level: Recent Patterns and Trends in Rural and Urban Welfare Inequality 27 Box 2.1 Are Rural-Urban Welfare Differences Shrinking or Growing? The evidence on the evolution of rural-urban welfare inequalities over time is mixed. Considerable literature exists on both the theoretical and empirical aspects of the convergence or divergence of rural and urban living standards as countries develop. Taking a long-term view, WDR 2009 asserts that an economic divide will emerge between rural and urban areas as the processes of urbaniza- tion and development proceed (World Bank 2008). This view is consistent with the notion of an industrializing core undergoing a virtuous circle of development, with positive feedback between productivity and the concentration of firms, while a rural periphery suffers from slower growth or even stagnation. According to WDR 2009, once countries reach the upper-middle-income level, rural-urban disparities tend to narrow. Essential household consumption is hypothesized to converge first, followed by access to basic public services and, finally, wages and incomes. The World Institute for Development Economics Research of the United Nations University (UNU/WIDER) research project found divergence of rural- urban living standards not only in low-income countries but also in middle- income countries such as the Czech Republic, Hungary, Mexico, Poland, and the Russian Federation. For low-income countries, divergence was stronger for income measures than nonincome measures. Limited convergence was found among nonincome measures, with many cases of neither convergence nor divergence over time. By contrast, there is evidence that rural-urban differences in poverty rates are narrowing, but only in some regions. Ravallion, Chen, and Sangraula (2007) find that since the early 1990s, rural poverty rates have, in the aggregate, been falling faster than urban poverty rates. They attribute this finding to poor peo- ple “urbanizing even more rapidly than the population as a whole.” In other words, the narrowing of rural and urban poverty rates is largely the product of a compositional effect, with rural-urban population shifts accounting for much of the more rapid poverty reduction in rural areas. The authors note pro- nounced regional differences in the differences in rural and urban poverty rates. For example, in China and in the Europe and Central Asia region, the rural-urban poverty rates diverged over the period 1993–2002, whereas in Sub-Saharan Africa and South Asia, there is no clear trend in the differences between rural and urban poverty rates. 28 From Farm to Firm welfare inequalities among the low- and middle-income countries that make up the L-shaped scatter on the left side of figure 2.1. These coun- tries represent the bulk of the World Bank’s clients. In addition to com- paring mean consumption, we analyze more distribution-sensitive welfare indicators such as the poverty headcount and poverty gap. We also address two nonmonetary welfare indicators, undernutrition and school enrollment. In addition to examining the cross-sectional evidence, we use repeated surveys in these countries to better understand the dynamics of changes in rural-urban welfare gaps and whether they are increasing or decreasing over time as these countries become richer and more urbanized. If the dynamics look anything like the cross-sectional information in figure 2.1, one would expect divergence of rural-urban living standards in the coun- tries with a low rural-urban gap and convergence in the countries with a large gap. An alternative hypothesis would be that in all cases the rural- urban divergence accompanies growth, which would imply extremely large rural-urban inequalities in the future for those low-income coun- tries that already have urban-rural ratios of 2:1 or greater. Finally, in this chapter we use cross-country regressions to explore possible country characteristics or typologies associated with rural-urban welfare diver- gence or convergence. Sources of Data The data used in the cross-country analysis presented in this chapter were drawn from two main sources. For the monetary-based welfare measures, we use an updated version of the data set constructed by Ravallion, Chen, and Sangraula (2007) for their study of the rural-urban distribution of poverty. The data set is compiled from dozens of household living stan- dards surveys that use broadly similar methods for defining a monetary measure of individual welfare based on consumption or income,1 includ- ing both cash transactions and imputed values for items produced and consumed by the household. The welfare measure and the national poverty lines are converted to internationally comparable terms by apply- ing purchasing power parity (PPP) exchange rates to the local currency values. The purpose of PPP exchange rates is to adjust for spatial differences across countries in the purchasing power of a nominal unit of currency; they make no adjustment for spatial differences in purchasing power within countries. Such differences can be large, especially between rural Global Level: Recent Patterns and Trends in Rural and Urban Welfare Inequality 29 and urban areas and in low-income countries with poorly integrated markets. For example, Ravallion, Chen, and Sangraula (2007) report that urban poverty lines are frequently 40–50 percent higher than rural poverty lines, with the difference reaching as high as 79 percent. If the higher cost of basic needs in urban areas is ignored, then urban poverty is underestimated and rural-urban differences in poverty are overstated in countries where urban poverty is lower than rural poverty, which is the vast majority. In their data set, Ravallion and his colleagues correct this deficiency by setting the rural extreme poverty line in each country at $1.25 per day (in PPP terms), and then using the ratio of urban to rural poverty lines observed in each country to set an urban extreme poverty line (also in PPP terms) for each country that reflects its prevailing rural- urban differences in the cost of basic needs. A similar procedure is used for the $2.00 per day poverty line. The data set distinguishes rural and urban areas using the definitions given in World Urbanization Prospects (WUP), a regular publication of the Population Division of the United Nations Secretariat (United Nations 2005). WUP largely follows defini- tions used by national statistical offices. The analysis in this section employs a subset of these data that has been updated to use the 2005 PPP exchange rates (Ravallion, Chen, and Sangraula [2007] used the 1993 PPP exchange rates). The data in this subset cover 41 countries, with two surveys for each country. The subset of countries was chosen on the basis of four criteria: (1) interval between the oldest and most recent available Living Standards Measurement Study (LSMS) surveys (longer intervals preferred), (2) range of national income levels, (3) geographic distribution, and (4) availability of Demographic and Health Surveys (DHS) data for nonmonetary welfare measures. The 41 countries in the data set are listed in appendix A, along with the year of the surveys and the GDP per capita (in 2005 interna- tional PPP dollars) corresponding to the survey year. The second data source is the Demographic and Health Surveys col- lected by ICF Macro. The DHS, which collects extensive information on nutrition, education, and access to services, has an extremely high degree of comparability across countries and survey years. Our analysis focuses on the nutritional status of children and school enrollment rates. DHS data are available for 27 of the 41 countries with LSMS data (the 27 countries in the DHS sample and years of the surveys are listed in appendix A). A significant limitation of both data sets is the time span covered. The surveys in the data sets were conducted between 1990 and 2006. The intervals between the two surveys in any given country range from 30 From Farm to Firm 4 to 16 years, which is short when compared with the rural-urban and structural transformations that typically take decades to achieve. Noise in the form of interannual fluctuations, such as from a particularly good or bad agricultural year, is likely to be high relative to the signal of long-term trends of divergence or convergence. But they are the best data available. Also noteworthy, because the surveys are repeated cross sections, little can be said about the changes in rural-urban welfare gaps caused by com- positional changes, such as the artificial “convergence” that would occur if a group of poor people moved from a poor rural area to a less poor urban area. However, this factor is addressed in chapter 3 for the three Sub-Saharan African countries featured in that chapter. Methods The comparisons in this chapter are intended to be exploratory and descriptive; they are not designed to establish causal inferences. The approach is straightforward. First, we compare the initial levels of wel- fare measures in rural and urban areas within each country. This compar- ison is carried out only for the first year of survey data, and so provides a baseline against which one can assess the convergence or divergence of welfare levels. In the second set of analyses, we examine whether welfare—as cap- tured by the selected measures—is improving more rapidly in urban areas or rural areas. Because the time period between the two surveys is relatively short, the results are unlikely to support statistically robust statements about convergence or divergence of the kind posited in WDR 2009. Instead, we use an approach similar to that adopted by Sahn and Stifel (2003) in their analysis of nonmonetary welfare measures in Sub-Saharan Africa. Changes in welfare between the two surveys are calculated separately for rural and urban areas in each country for each welfare measure. This calculation is made first using the absolute changes in levels between the two periods (e.g., mean consumption in 2004 minus rural mean consumption in 1996 in a given country). A similar comparison is then made using the Kakwani improvement index, which is based on the idea that for finite measures of welfare an incremental improvement is more difficult to achieve when the initial level is closer to the maximum (see box 2.2). Finally, we inves- tigate patterns across countries in the convergence or divergence of wel- fare measures, looking specifically at national income levels and regional groupings as potential correlates. Global Level: Recent Patterns and Trends in Rural and Urban Welfare Inequality 31 Box 2.2 Comparing Welfare Differences and Changes Multiple methods are used for comparing rural and urban performance on the various welfare measures. A commonly used approach is to express the differ- ence as a ratio, such as the ratio of mean urban consumption to mean rural consumption (e.g., World Bank 2008). The ratio is extremely intuitive, with val- ues closer to 1 indicating greater equality between groups on the welfare measure. This approach may be appropriate for comparing means, such as mean consumption, but it can be misleading for the five other welfare meas- ures we consider, which are based on proportions of the population above or below a certain threshold. For example, consider a country whose poverty headcount of 60 percent in rural areas and 30 percent in urban areas later fell to 6 percent and 3 percent, respectively. In relative terms, the difference is unchanged, because the probability of being in poverty is twice as high in rural areas in each period. However, in absolute terms, the rural-urban gap has been narrowed tremendously—a fact that is missed by the ratio measure. Therefore, for the poverty, nutritional status, and school enrollment measures, we use the absolute differences between rural and urban areas. However, absolute differences can be misleading in assessing development progress over time, either within a country or between countries. As Sen (1981) and others have argued, the relationship between achievement and the values of many common welfare indicators is nonlinear. In particular, as the standard of living reaches higher levels, incremental improvement becomes more difficult. It is arguably easier to reduce poverty by three percentage points if the initial level is 60 percent than if the initial level is 6 percent. Kakwani’s achievement index compares levels of welfare measures to their minimum and maximum possible levels (Kakwani 1993). For an increasing welfare measure x (i.e., higher levels imply greater welfare) recorded at time t, the Kakwani achievement index takes the form In (M − M0 ) − In (M − x t ) f ( x t ,M0 ,M ) = , In (M − M0 ) where M0 and M are the lower and upper bounds of the welfare measure, respec- tively. Because poverty and stunting are measures of negative outcomes—that is, lower values represent higher welfare—for those variables we define xt as equal to 100 – wt, where wt is the conventional poverty index or undernutrition (continued next page) 32 From Farm to Firm Box 2.2 (continued) prevalence. If the welfare measures in periods 1 and 2 are defined as x1 and x2, it is possible to define an improvement index that is equal to the difference in the achievement indexes for the two periods: Q( x1, x 2 , M0 , M ) = f ( x 2 , M0 , M ) − f ( x1, M0 , M ) In (M − x1) − In (M − x 2 ) = . In (M − M0 ) The improvement index satisfies several important axioms of welfare compar- isons; for details, see Kakwani (1993). In our study, we use six different welfare measures, four monetary and two nonmonetary.2 The comparisons are made using aggregate rural and urban statistics for each survey. The monetary measures, based on con- sumption per capita, are mean consumption (in 2005 international PPP dollars), the poverty headcount index using both the $1.25 per day and $2.00 per day thresholds, and the poverty gap index (FGT1) for $2.00 per day. By employing two poverty lines and two poverty indexes, we are able to detect changes in the distribution of welfare that often are not apparent from comparison of mean consumption alone. The two nonmonetary welfare measures used are school enrollment and child undernutrition. The education measure is the proportion of children between the ages of 6 and 15 (inclusive) who were attending school, at any level, at the time of the survey. For child undernutrition, we use the incidence of stunting (measured by height more than two stan- dard deviations below the median for the child’s age and sex) among chil- dren under 3 years of age. Stunting is generally the result of a combination of insufficient nutrient intake and repeated illness and can therefore be interpreted as a measure of chronic deprivation. Results The results that follow aim to answer two principal sets of questions, one static and the other dynamic. The first set of questions relates to welfare levels: Is welfare consistently higher in urban areas than in rural areas? Are these disparities higher or lower in middle-income countries Global Level: Recent Patterns and Trends in Rural and Urban Welfare Inequality 33 than in low-income countries? The second set of questions relates to changes in welfare: Are rural-urban differences shrinking or growing over time? Do trends vary according to countries’ levels of development? Rural-Urban Inequalities: The Static Picture In almost all countries (40 of 41), mean consumption is higher in urban areas than in rural areas. Panel (a) of figure 2.2 plots the ratio of urban to rural mean consumption per capita against log national GDP per capita.3 All countries except Armenia have a ratio greater than 1.0. The ratio is extremely variable among the low-income countries, ranging from 1.2 in Tanzania and Madagascar to over 3.5 in Burkina Faso. For most of the lower-middle- and upper-middle-income countries, mean consumption in urban areas is two to three times larger than that in rural areas. Although a few low-income, high urban-rural ratio countries appear in the upper left-hand corner of panel (a), overall the plot shows a strong positive cross-sectional correlation between the urban-rural consumption ratio and GDP per capita, indicating larger welfare gaps in middle-income countries than in low-income countries.4 In almost all cases, the poverty headcount rate is higher in rural areas, although the extent of the disparity varies across countries. Panel (b) of figure 2.2 presents the differences in rural and urban poverty using the extreme poverty line of $1.25/day (in 2005 PPP dollars). The rural-urban differences are highly variable among the low-income countries, ranging from just a few percentage points to over 40 percentage points in Burkina Faso and India. These differences are smaller in the middle-income coun- tries, because for middle-income countries the percentage of the popula- tion below the extreme poverty line is very low in both rural and urban areas. By contrast, when the $2.00 per day poverty line is used (panel (c), the average difference in poverty rates appears to be about the same for both low- and middle-income countries. Consistent with the mean consumption and poverty headcount results, the poverty gap, which captures both the proportion of the pop- ulation below the poverty line and the average depth of their poverty, is higher in rural areas in almost all of the 41 countries in the sample. Panel (d) of figure 2.2 shows the rural-urban differences in the poverty gap index for the initial period using the $2.00 per day poverty line. The rural-urban differences in the poverty gap range from 0 to 30 percentage points. Overall, the rural-urban differences in the poverty gap are smaller in the middle-income countries, dropping to 12 percentage points or less for countries in the sample with annual GDP per capita greater than 34 Figure 2.2 Rural-Urban Welfare Differences in Initial Survey Year, Selected Countries (a) Ratio of urban-rural mean consumption per capita (b) Absolute difference in $1.25/day poverty headcount (c) Absolute difference in $2.00/day poverty headcount difference in poverty headcount index BFA IND 40 ratio of urban-rural mean consumption difference in poverty headcount index 40 BFA IDN BFA NPL 3.5 MLI UGA NPLVNM UGA KHM BRA MLI 30 VNM GTM 3 PRY 30 IND BOL MEX IDN UGA NPL GTM NIC HND EGY PER BOL CMR ($1.25/day) (2005 PPP $) SLV UZB TUNCOL ($2.00/day) COL ZAF ZMB GTM 20 SEN CIV PRY ECU BRA 2.5 KHM SEN CMR HND 20 SEN NIC BOL PER UZB HND COL MLI ZMB KGZ ZAF MEX MOZ VNM KHM CMR NIC PRY ECU LKA SLV ZMB ECU GHA CIV PER BRA 10 GHA PAK DOMKAZ 2 MAR TUN 10 MOZ SLV MOZ MAR IND IDN DOM NGA LKA TUN ZAF MDG MDA CI MDG KGZ EGY KAZ NGA GHA LKA EGY MDA DOM MEX TZA BGD UZB TZA 0 BGD 1.5 PAK MAR TZA KGZ PAK NGA 0 BGD MDG MDA KAZ 1 –10 ARM –10 ARM ARM 500 1,000 2,500 5,000 10,000 500 1,000 2,500 5,000 10,000 500 1,000 2,500 5,000 10,000 GDP per capita, 2005 PPP $ (log scale) GDP per capita, 2005 PPP $ (log scale) GDP per capita, 2005 PPP $ (log scale) (d) Absolute difference in $2.00/day poverty gap (e) Absolute difference in stunting prevalence (f) Absolute difference in school enrollment rates urban-rural difference in enrollment rates 30 PER 50 MLI BFA rural-urban difference in stunting IND 20 BFA difference in poverty gap index UGA IDN MAR VNM GTM 40 NPL prevalence (<–2 SDs) 20 BOL SEN UGABGD PAK BOL MLI (6–15 years old) ZMB 15 GHA SEN CIV MAR KAZ UZB GTM ($2.00/day) HND NPL ZMB IND 30 SEN COL CMR MDG KHM CMR NIC LKA PRY ECU MOZ BFA NIC MOZ 10 PER BRA MLI NGA GHA CIV SLV 10 MDG NPL ZMB GTM NGA NIC DOM COL MOZ KGZ EGYTUN ZAF NGA EGY 20 CIV TZA MDAPAK DOMKAZ DOM EGY MDG MAR MEX COL GHA CMR 0 BGD TZA 5 ARM 10 BOL UGA PER KHM TZA ARM –10 0 0 BGD KAZ 500 1,000 2,500 5,000 10,000 500 1,000 2,500 5,000 10,000 500 1,000 2,500 5,000 10,000 GDP per capita, 2005 PPP $ (log scale) GDP per capita, 2005 PPP $ (log scale) GDP per capita, 2005 PPP $ (log scale) Source: Authors’ calculations using LSMS and DHS data (see appendix A). Note: SD = standard deviation. See appendix A for country abbreviations. Global Level: Recent Patterns and Trends in Rural and Urban Welfare Inequality 35 $5,000 (PPP). Taken together, panels (c) and (d) indicate that, among the poor, the depth of poverty tends to be greater among the rural poor, which is somewhat surprising in view of the higher inequality that gener- ally characterizes urban areas. The undernutrition (stunting) rates in the initial survey year are higher in rural areas in all 27 countries in the DHS sample. As shown in panel (e) of figure 2.2, in the initial survey years the stunting rates are 1–22 percentage points higher in rural areas than in urban areas, with most countries exhibiting differences of 10–15 percentage points. The rural-urban differences in stunting rates do not diminish in the middle- income countries, and in fact, the largest differences are in some of the richest countries in the sample in strong contrast with the mean con- sumption and poverty differences on view in panels (a)–(d). Several explanations are possible for this finding. One is that basic nonfood liv- ing expenses are much higher in the more urbanized middle-income countries, leaving a smaller share of the budgets of poor households to meet food needs. Another possibility is that the more urbanized middle-income countries have greater problems with congestion and inadequacy of public health and sanitation in poor areas, contributing to urban undernutrition through the morbidity pathway. The rural-urban differences in school enrollment rates, especially among low-income countries, vary a great deal. Panel (f) of figure 2.2 shows the difference in the percentage of children aged 6–15 who are enrolled in school.5 For example, in Bangladesh and Kazakhstan the rural and urban enrollment rates in this age group are almost equal, whereas in Burkina Faso the difference in enrollment rates is 44 percentage points (63 percent in urban areas versus 19 percent in rural areas). The rural-urban enrollment gaps are generally smaller among the lower- middle-income countries in the sample, although even in the lower- middle-income countries the difference in enrollment rates is generally between 10 and 20 percentage points. To summarize, three observations emerge from the data on initial levels of welfare in rural and urban areas. First, in both low- and middle- income countries urban areas are almost always better off than rural areas. This not surprising observation is consistent with the conventional wisdom about the average rural-urban welfare differences described at the outset of this chapter. Second, the size of the rural-urban gaps is highly variable, especially among the poorest countries. This variability was observed previously for mean consumption, but it also holds for poverty measures, nutritional status, and school enrollment. Third, the 36 From Farm to Firm size of rural-urban differences in mean consumption, the $2.00 per day poverty headcount, and undernutrition do not appear to vary systemati- cally with GDP per capita. The cross-sectional evidence presented thus far provides little, if any, support for the hypothesis of diverging rural- urban welfare levels during the early stages of development. In the sec- tions that follow, we address the question of spatial convergence more adequately by using two surveys for each country. Rural-Urban Inequalities: The Dynamic Picture Turning to the data on changes in welfare, the uniformity of the finding that initial welfare levels are almost always higher in urban areas than in rural areas allows us to conclude that, with only a few exceptions, a more rapid improvement (or less dramatic decline) in welfare in rural areas would represent convergence. Because GDP per capita grew over the interval between surveys in 38 of the 41 countries in the sample,6 we also can conclude that whatever patterns that emerge can be associ- ated with growing and developing economies, with minimal noise from countries that are economically struggling or in crisis during the period under consideration. Table 2.1 summarizes the results for the absolute changes in the six welfare measures. Similar to the approach of Sahn and Stifel (2003), the entries in the table refer to the question, did rural welfare improve more than urban welfare?—that is, is there rural-urban welfare convergence? A “yes” entry indicates that the absolute improvement in the welfare meas- ure was larger in the rural sector than in the urban sector, or that the wel- fare measure improved in the rural sector and declined in the urban sector. A “(yes)” entry indicates that the welfare measure declined in both rural and urban areas, but that the deterioration was less in rural areas. Conversely, a “no” entry indicates that the welfare measure improved more in the urban area than in the rural area, or improved in the urban area and declined in the rural area. A “(no)” entry indicates that the wel- fare measure declined in both areas, but that the decline was less in urban areas. Therefore, entries of “yes” and “(yes)” may be interpreted as indica- tors of rural-urban convergence in living standards, and entries of “no” and “(no)” are indicators of welfare divergence.7 The bottom rows of the table summarize instances of convergence and divergence via simple counts. The data on changes in welfare levels do not yield clear patterns of either convergence or divergence. Among the 41 countries in the sam- ple, mean consumption and $2.00 per day poverty diverge in about twice as many countries as they converge. By contrast, extreme poverty Table 2.1 Changes over Time in Absolute Levels of Welfare Measures, Selected Countries Did rural welfare improve more than urban welfare? Mean Headcount Headcount Poverty gap School Region Country consumption ($1.25/day) ($2.00/day) ($2.00/day) Stunting enrollment AFR Burkina Faso Yes Yes Yes Yes (No) No Cameroon No No No No (Yes) No Côte d’Ivoire No Yes Yes Yes Yes Yes Ghana No No No No (Yes) (No) Madagascar No No No No Yes Yes Mali Yes Yes No Yes (Yes) Yes Mozambique No Yes No Yes No Yes Nigeria No No No No No Yes Senegal No Yes No Yes Yes Yes South Africa (Yes) (No) (No) (No) — — Tanzania (No) (No) (No) (No) No No Uganda No No No No Yes Yes Zambia Yes Yes Yes Yes (Yes) (Yes) EAP Cambodia No No Yes No No — Indonesia No Yes No Yes — — Vietnam No Yes No Yes — — 37 (continued next page) 38 Table 2.1 (continued) Did rural welfare improve more than urban welfare? Mean Headcount Headcount Poverty gap School Region Country consumption ($1.25/day) ($2.00/day) ($2.00/day) Stunting enrollment ECA Armenia No No No No Yes — Kazakhstan No Yes (No) (No) Yes No Kyrgyz Republic (Yes) No (No) No — — Moldova (No) No (No) No — — Uzbekistan (Yes) (Yes) (No) (Yes) — — LAC Bolivia (No) (No) (No) (No) No (No) Brazil Yes Yes Yes Yes — — Colombia No (Yes) (Yes) (Yes) No Yes Dominican Republic Yes (Yes) Yes (Yes) Yes Yes Ecuador No (Yes) (No) (Yes) — — El Salvador No No No No — — Guatemala Yes Yes Yes (Yes) Yes Yes Honduras No Yes Yes Yes — — Mexico Yes Yes Yes Yes — — Nicaragua (Yes) Yes Yes Yes No No Paraguay Yes (Yes) (Yes) Yes — — Peru No No No No No Yes MNA Egypt, Arab Rep. No Yes Yes Yes Yes Yes Morocco (Yes) No (No) (No) Yes Yes Tunisia No Yes Yes Yes — — SAR Bangladesh No No No No Yes Yes India No Yes No Yes (Yes) — Nepal No Yes No No No Yes Pakistan (No) (No) (No) (No) — — Sri Lanka Yes Yes Yes No — — Number in which rural welfare increased more 14 24 15 21 16 16 Yes 9 19 13 17 11 15 (Yes) 5 5 2 4 5 1 Number in which urban welfare increased more 27 17 26 20 10 7 No 23 13 16 14 9 5 (No) 4 4 10 6 1 2 Source: Authors’ calculations using LSMS and DHS data (see appendix A). Note: “Yes” indicates that welfare improved more in rural areas, or improved in rural and decreased in urban areas; “(yes)” indicates that welfare decreased in both, but urban decreased more; “no” indicates that welfare improved more in urban areas, or improved in urban and decreased in rural; “(no)” indicates that welfare decreased in both, but rural decreased more; — = not available; AFR = Africa; EAP = East Asia and the Pacific; ECA = Europe and Central Asia; LAC = Latin America and the Caribbean; MNA = Middle East and North Africa; SAR = South Asia. 39 40 From Farm to Firm ($1.25 per day) rates and the $2.00 per day poverty gap converge in slightly more countries than they diverge. That convergence is more common for $1.25 per day poverty is to be expected, because as coun- tries get richer the extreme poverty rate eventually approaches zero. As a result, many countries experience convergence of rural-urban poverty rates even while their mean consumption levels and $2.00 per day poverty rates are diverging. The Latin America and Caribbean (LAC) countries have a much higher prevalence of rural-urban welfare convergence on consumption- based measures than other regions. Convergence occurs in poorer coun- tries such as Honduras and Nicaragua as well as richer countries such as Mexico. This finding may be related to the higher levels of urbanization in the LAC region. Because DHS surveys are available for only 6 of the 12 LAC countries in the sample, it is more difficult to assess the region’s performance on the nonmonetary welfare measures. For those LAC coun- tries for which data on the nonmonetary measures are available, there is an equal number of instances of convergence and divergence. Convergence is slightly more common for absolute changes in non- monetary welfare measures. The last two columns of table 2.1 reveal that rural-urban stunting rates converged in 16 of 26 countries, and school enrollment rates converged in 16 of 23. Multiple (and nonmutually exclusive) explanations are possible for the higher frequency of conver- gence for these measures. One explanation might be that children’s nutri- tional status and school enrollment are input measures that lead to changes in output measures—that is, reduced spatial inequality in the formative years may lead to relatively better productivity and labor mar- ket outcomes in later years, which in turn would reduce spatial inequal- ity in consumption and poverty in those years. Another possible explanation is related to migration and the changing composition of rural and urban populations over time. If better-nourished and better-educated people are more likely to migrate to urban areas, greater convergence on the nonmonetary welfare input indicators may not translate into greater convergence in monetary indicators in later periods. When the Kakwani improvement index is applied, the majority of the countries exhibit divergence on the consumption-based welfare meas- ures. Table 2.2 reveals that between one-half and two-thirds of the coun- tries in the sample have increasing spatial inequalities in mean consumption and poverty. In general, the Kakwani index tends to show less convergence than the absolute measures presented in table 2.1, because for a given level of absolute change it gives greater weight to Table 2.2 Changes over Time in Welfare Measures (Kakwani Improvement Index), Selected Countries Did rural welfare improve more than urban welfare? Mean Headcount Headcount Poverty gap School Region Country consumption ($1.25/day) ($2.00/day) ($2.00/day) Stunting enrollment AFR Burkina Faso Yes Yes Yes Yes (No) No Cameroon No No No No (Yes) No Côte d’Ivoire No Yes Yes Yes Yes Yes Ghana No No No No (Yes) (Yes) Madagascar No No No No Yes Yes Mali Yes No No No (Yes) Yes Mozambique No Yes No Yes No No Nigeria No No No No No Yes Senegal No No No No No Yes South Africa (Yes) (No) (No) (No) — — Tanzania (No) (No) (No) (No) No No Uganda No No No No Yes No Zambia Yes Yes Yes Yes (Yes) (Yes) EAP Cambodia No No No No No — Indonesia No No No No — — Vietnam No No No No — — (continued next page) 41 42 Table 2.2 (continued) Did rural welfare improve more than urban welfare? Mean Headcount Headcount Poverty gap School Region Country consumption ($1.25/day) ($2.00/day) ($2.00/day) Stunting enrollment ECA Armenia No No No No Yes — Kazakhstan No Yes (No) (No) Yes No Kyrgyz Republic (Yes) No (No) No — — Moldova (No) No (No) No — — Uzbekistan (Yes) (Yes) (Yes) (Yes) — — LAC Bolivia (No) (No) (No) (No) No (Yes) Brazil Yes Yes Yes Yes — — Colombia No (Yes) (Yes) (Yes) No No Dominican Republic Yes (Yes) Yes (Yes) Yes Yes Ecuador No (Yes) (Yes) (Yes) — — El Salvador No No No No — — Guatemala Yes Yes Yes Yes No Yes Honduras No Yes No Yes — — Mexico Yes Yes Yes Yes — — Nicaragua (Yes) Yes Yes Yes No No Paraguay Yes (Yes) (Yes) Yes — — Peru No No No No No No MNA Egypt, Arab Rep. No Yes Yes Yes Yes Yes Morocco (Yes) No (Yes) (No) Yes No Tunisia No Yes No Yes — — SAR Bangladesh No No No No Yes Yes India No No No No (Yes) — Nepal No No No No No Yes Pakistan (No) (No) (No) (No) — — Sri Lanka Yes Yes Yes Yes — — Number in which rural welfare increased more 14 18 15 17 14 13 Yes 9 13 10 13 9 10 (Yes) 5 5 5 4 5 3 Number in which urban welfare increased more 27 23 26 24 12 10 No 23 19 19 18 11 10 (No) 4 4 7 6 1 0 Source: Authors’ calculations using LSMS and DHS data (see appendix A). Note: See table 2.1 for meaning of yes/(yes)/no/(no) responses and region abbreviations. — = not available. 43 44 From Farm to Firm improvements in the group or sector that has a higher initial welfare level. That said, the instances of divergence in mean consumption and the $2.00 per day headcount are the same whether the absolute differences or the Kakwani index is used. The consistency of the direction of change happens to be particularly strong when the Kakwani index is used on these data. Fully 70 percent of the countries record a uniform divergence (“no”) or convergence (“yes”) across all four consumption and poverty indicators. Again, the LAC region shows a much stronger tendency to rural-urban welfare convergence of mean consumption and poverty than any of the other regions. The Kakwani index reveals slightly more instances of convergence than divergence for the nutrition and education welfare measures. In the last two columns of table 2.2, convergence on stunting and school enroll- ment rates is observed in slightly more than half of the countries. It is striking that in many countries the direction of the convergence or diver- gence of the nonmonetary measures is the opposite of what is observed for the consumption-based measures. For example, despite the wide- spread rural-urban convergence of consumption and consumption poverty in the LAC region, there is slightly more divergence on under- nutrition and school enrollment. The opposite pattern is observed in Sub-Saharan Africa and South Asia. In summary, the data on changes in absolute or relative welfare pro- vide few clear patterns of convergence or divergence. The consumption- based measures diverge over time in about two-thirds of the countries, and the undernutrition and education measures converge in slightly more than half. For any given country, rural-urban convergence or diver- gence on the consumption-based welfare measures is a poor predictor of whether convergence or divergence is observed on the nonmonetary measures. A close look at tables 2.1 and 2.2 also reveals very few clear trends for the common characteristics across countries where conver- gence (or divergence) has occurred in recent years. The Elusive Quest for Patterns In this section, we use simple econometric models to detect patterns between rural-urban convergence or divergence on these welfare meas- ures and other covariates, in particular, national income per capita. For these regressions, we chose two different sets of dependent variables, cor- responding to the measures of convergence and divergence in the preced- ing section. In one set of regressions, the dependent variable is the Global Level: Recent Patterns and Trends in Rural and Urban Welfare Inequality 45 rural-urban difference in the absolute level of the welfare measure. In the second set of regressions, it is the rural-urban difference in the Kakwani achievement index for that welfare measure. We use five different speci- fications on the right-hand side of the regression equation: 1. A basic pooled data model that regresses the welfare levels on log GDP per capita, log GDP per capita squared, and a set of six control variables related to the level of sectoral and spatial transformation (population density, rate of urban population growth, urban popula- tion share, agriculture value added per worker, agriculture share of GDP, and GDP growth rate)8 2. Model 1 with a time trend variable added to control for the different years in which the surveys were conducted 3. Model 1 with time trend and region-fixed effects 4. Model 1 with time trend and country-fixed effects 5. Model 1 with time trend, region-fixed effects, and country-fixed effects. Cross-country regressions on pooled data, such as models 1 and 2, effectively trace the relationship between national income and rural- urban welfare differences using data from diverse countries at different stages of development. Although the set of six variables measuring sec- toral and spatial transformation control for the differing stages of devel- opment to some degree, meaningful inference from such regressions is problematic and, some would argue, impossible. They are included here largely as a basis of comparison with the fixed effects models, and any interpretations of the results from the pooled cross-country data should be made with caution. The regression results are summarized in table 2.3, which shows the significance levels for the joint tests of significance of the coefficients on log GDP per capita (linear and quadratic). When estimated on pooled data (models 1 and 2), the regressions show a statistically strong relation- ship between national income and rural-urban differences in the consumption-based welfare measures, especially when measured by the Kakwani achievement index. The coefficients from the pooled models indicate a U-shaped relationship, with rural-urban inequalities among this group of 41 low- and middle-income countries decreasing and then increasing with the rising levels of GDP per capita. This pattern corre- sponds roughly with the L-shaped (or “tilted L”) pattern in the scatter plots described earlier in this chapter. When region-fixed effects are 46 Table 2.3 Significance of Relationship between Convergence and Log GDP Per Capita Pooled + time Time trend + Time trend + Time trend + region- Pooled trend region-fixed country-fixed and country-fixed Dependent variable (model 1) (model 2) effects (model 3) effects (model 4) effects (model 5) Absolute differences Log mean consumption ** ** ** n.s. n.s. Poverty headcount ($1.25/day) n.s. n.s. n.s. n.s. n.s. Poverty headcount ($2.00/day) *** *** * n.s. n.s. Poverty gap ($1.25/day) n.s. n.s. n.s. n.s. n.s. Poverty gap ($2.00/day) n.s. * n.s. n.s. n.s. Stunting n.s. n.s. n.s. n.s. n.s. School enrollment ** ** * n.s. n.s. Kakwani achievement index Log mean consumption *** *** *** n.s. n.s. Poverty headcount ($1.25/day) *** *** *** n.s. n.s. Poverty headcount ($2.00/day) *** *** *** n.s. n.s. Poverty gap ($1.25/day) ** ** ** n.s. n.s. Poverty gap ($2.00/day) *** *** *** n.s. n.s. Stunting n.s. n.s. n.s. n.s. n.s. School enrollment * * n.s. n.s. n.s. Source: Authors’ calculations using LSMS and DHS data (see appendix A). Note: All dependent variables are rural-urban differences. *p < .10. **p < .05. ***p < .01. n.s. = not significant. Global Level: Recent Patterns and Trends in Rural and Urban Welfare Inequality 47 introduced (model 3), the significance and the shape of the relationship between rural-urban inequalities and GDP per capita are essentially the same as with the pooled models. When country-fixed effects are included (models 4 and 5), the coeffi- cients on log GDP per capita are not significant for any of the welfare measures. This result is most likely attributable to data limitations, because for country-fixed effects the regression depends entirely on within-country variations in the variables. When limited to only two observations per country and a relatively short time interval between those observations (from 4 to 16 years), this variation is insufficient to fit a good model with log GDP per capita against any of the rural-urban dif- ferences in welfare. Whether more data over a longer time period would uncover a systematic relationship remains an open question. Conclusion This chapter has used recent survey data to examine the nature and magnitude of welfare inequalities between rural and urban areas in 41 low- and middle-income countries. The welfare measures are both monetary and nonmonetary aspects of well-being, including mean con- sumption, several poverty measures, child undernutrition, and school enrollment rates. According to the first set of analyses described in this chapter, for nearly every country over the time period considered, the standard of liv- ing was higher in urban areas across almost all six of the variables used in our study. However, the magnitude of rural-urban inequalities was extremely variable across welfare measures and across countries. Even in countries with the same levels of urbanization or GDP per capita, the rural-urban differences may be enormous or relatively small. For exam- ple, rural-urban differences are large in Burkina Faso but small in Tanzania. And Bolivia, which is higher up the national income scale, has a much larger rural-urban gap than the Dominican Republic. The welfare gap is not consistently wider in low-income countries than in middle- income countries. In fact, when measured by consumption and stunting, the average rural-urban welfare difference is roughly the same, irrespective of a country’s level of development. We then compare the improvements in welfare in each country during the interval between two surveys to determine whether the rural-urban welfare gap expanded or contracted. The total number of divergences in rural-urban welfare is slightly greater than the number of convergences. 48 From Farm to Firm There is a higher rate of divergence, or increasing rural-urban inequality, for consumption-based welfare measures, whereas the nonmonetary measures show a slightly higher rate of convergence over time. In the final section, we look for evidence of a relationship between changes in the rural-urban welfare gap and changes in national income. The cross-country results from our sample of low- and middle-income countries indicate a U-shaped relationship, with rural-urban inequalities tending to be largest in the poorest countries in the sample, such as Nepal and Uganda, and also in the richest countries, such as Brazil and South Africa. However, after controlling for country-fixed effects we detect no statistically significant relationship for any of the variables, although the short time span covered by the survey data is likely a major reason for this result. Our findings, along with those of previous studies described in box 2.1, reveal a wide array of welfare convergence or divergence patterns accom- panying growth and urbanization. At least at the low- and middle-income levels, there seems to be little evidence that spatial inequalities automat- ically diminish as countries develop. These results imply that country- level policies and institutions, rather than particular thresholds of per capita GDP or urbanization, are driving the processes of rural-urban divergence and convergence. This observation in turn suggests the hazards of painting with brushstrokes that are too broad, and the importance of in-depth analysis of each country’s “stylized facts” before proceeding to policy prescriptions. In the next chapter, we turn to a more in-depth analysis, examining the sources of inequality in three rapidly growing African countries. Notes 1. In 10 of the 41 countries used for this report, welfare is measured by income rather than consumption expenditure. All of these countries are in Latin America. It is not clear a priori whether using income in some countries and consumption in others introduces a bias in the comparison of rural-urban wel- fare gaps, or the direction of the bias if there is one. For the analysis of changes over time, there is unlikely to be any bias, because none of the countries changed the welfare measure from income to consumption, or vice versa, from one survey to the next. 2. Ten welfare measures (five monetary and five nonmonetary) were analyzed for this study, but only six are reported here because of length considerations. The patterns for the other four are similar to the six presented here. Full details are available from the authors upon request. Global Level: Recent Patterns and Trends in Rural and Urban Welfare Inequality 49 3. We produced similar versions of these plots showing the urban population share and the agriculture sector share of GDP on the horizontal axis, because these shares are also indicative of a country’s stage of development. The alter- native plots look almost identical to the plots in figure 2.2 because of the high intercorrelation of the three variables. 4. Note that although the patterns in figure 2.2, panel (a), are similar to those in figure 2.1, the positive correlation in figure 2.2, panel (a), continues up to $10,000, whereas in figure 2.1 the correlation turns negative at about $3,000. This result likely stems from the different base years used (2000 versus 2005) and the PPP adjustment that is incorporated in figure 2.2 but not figure 2.1. 5. In this case, we do not differentiate by the level of schooling, nor do these fig- ures convey any information about the quality of schooling. 6. The three exceptions are Colombia, Côte d’Ivoire, and Paraguay. 7. This interpretation is reversed for the small minority of cells (8 of 209) in which initial welfare was higher in rural areas. These include the following countries and welfare measures: Armenia (mean consumption and all three poverty measures), Bangladesh ($2.00 per day poverty headcount and poverty gap), Kazakhstan (school enrollment), and Morocco ($1.25 per day poverty headcount). 8. Additional regressions without these control variables yielded results that are essentially the same as those reported here. References Kakwani, N. 1993. “Performance in Living Standards: An International Comparison.” Journal of Development Economics 41 (2): 307–36. Ravallion, M., S. Chen, and P. Sangraula. 2007. “New Evidence on the Urbanization of Global Poverty.” Policy Research Working Paper 4199, World Bank, Washington, DC. Sahn, D., and D. Stifel. 2003. “Urban-Rural Inequality in Living Standards in Africa.” Journal of African Economies 12 (4): 564–97. Sen, A. K. 1981. “Public Action and the Quality of Life in Developing Countries.” Oxford Bulletin of Economics and Statistics 43 (November): 287–319. United Nations. 2005. World Urbanization Prospects: The 2004 Revision. Department of Economic and Social Affairs, Population Division. New York: United Nations. World Bank. 2008. World Development Report 2009: Reshaping Economic Geography. Washington, DC: World Bank. CHAPTER 3 National Level: Three Country-Level Perspectives on Rural-Urban Transitions Looking at levels and changes in both income and nonincome measures of welfare, we sought in chapter 2 to determine patterns and trends in rural-urban inequalities across a sample of 41 low- and middle-income countries. It was not possible, however, to identify a consistent trend of either convergence or divergence across such a broad set of countries. This chapter, then, explores in greater depth the rural-urban inequal- ities in three countries in Sub-Saharan Africa—Ghana, Mozambique, and Uganda—with the aim of gaining more finely textured insights into country-specific patterns and trends and the mechanics driving the rural-urban divide. We begin by describing for these three countries the present structural and spatial transformations and the accompanying welfare trends. Despite their roughly similar levels of development, these countries present an interesting mix of similarities and differences in urbanization rates, spatial inequalities, and other characteristics. Next, we evaluate the role of differences in household characteristics and the returns to those characteristics in explaining rural-urban inequalities. We find that in Mozambique and Uganda, rural-urban inequalities largely stem from low levels of human capital and other assets, whereas in Ghana, inequalities are attributable to characteristics and returns in almost equal proportions. 51 52 From Farm to Firm Structural and Spatial Transformations in Ghana, Mozambique, and Uganda The low-income countries of Ghana, Mozambique, and Uganda have made significant progress in growth and poverty reduction since the 1990s. The growth rates for both gross domestic product (GDP) and GDP per capita during the 1990s and 2000s exceeded the average rates for Sub-Saharan Africa (table 3.1). In fact, Ghana, Mozambique, and Uganda are among only eight Sub-Saharan African countries that have attained or outperformed regional averages during these two periods.1 In each of these countries, economic growth has also been accompa- nied by significant structural and spatial transformations. However, a large share (and in Mozambique and Uganda, a clear majority) of the labor force is still in agriculture, and the agricultural share of GDP remains significant. From 1990 to 2005, the average share of labor employed in agriculture and its contribution to GDP were 60 percent (labor) and 40 percent (GDP) in Ghana, 80 percent (labor) and 30 percent (GDP) in Mozambique, and 80 percent (labor) and 40 percent (GDP) in Uganda (figure 3.1). Agricultural output increased over the same period, primarily because of expansion of the areas under cultivation, contributing to overall economic growth. From 1990 to 2005, agriculture accounted for an estimated 36 percent of economic growth in Ghana, 22 percent in Mozambique, and 25 percent in Uganda (figure 3.2). Even though under the World Development Report 2008: Agriculture for Development (WDR 2008) (World Bank 2007) typology the three coun- tries can still be classified as “agriculture-based economies,” major structural changes have taken place over the last two decades. The relative share of agricultural output in GDP has decreased, and the industrial and service sectors of the economy are growing much more rapidly (figure 3.3). These structural transformations suggest that these countries are moving toward becoming “transforming” countries under the WDR 2008 typology. Table 3.1 Growth Rates: Ghana, Mozambique, and Uganda, 1990s and 2000s Per capita GDP growth rate GDP growth rate 1990–2000 2000–08 1990–2000 2000–08 Sub-Saharan Africa –0.2 3.0 2.6 5.6 Ghana 1.5 3.2 4.3 5.6 Mozambique 2.9 5.6 6.1 8.0 Uganda 3.8 4.0 7.1 7.5 Source: Authors’ calculations based on World Development Indicators, World Bank (October 2009). Figure 3.1 Shares of GDP and Labor from Agriculture, Selected Countries 1.0 NPL BFA BDI NER RWA ETH MLI GIN MWI MOZ UGA TZA 0.8 MDG LAO PNG TCD KEN KHM ZMB SEN AGO CHN share of labor and GDP in agriculture VNM ZAR SDN ZWE CMR China 1961–2003 TGO IND THA 0.6 BEN BGO ZAR GHA BDI CIV LAO YEM IDN TUR ETH PAK UGA TZA BOL GTM Nigeria 1961–2003 MLI RWAGHA KHM LKA Brazil 1961–2003 0.4 NPL PHL MAR NER BEN EGY TJK TGO NGA AZE HND SDN CMR PRY NGA PER TCD TJK MDG PNG SYR SLV MOZ BFA UZB UZB AZE PRY IRN MWI VNM ECU TUN KEN CIV HND DZA BGD COL IND PAK SYR IRN GTM MYS POL MEX 0.2 GIN YEM LKA PHL DOM BLR EGY ROM BRA SEN ZWE CHN UKR MAR COL CHL IDN ROM TUN TUR ZMB AZE BOL BLR BGR DZA SLV ZAF MYS HUN ARG VEN UKR SVK CZE AGO ECU DOM BRA CHL BGR PER THA MEX SVK ARG VEN ZAF POL HUN CZE 0 90 150 245 400 670 1,100 1,800 3,000 4,900 8,100 GDP per capita, constant 2000 US$ (log scale) share of labor in agriculture (1990–2005, average) share of GDP from agriculture (1990–2005, average) trajectories of the share of labor in agriculture, 1961–2003 53 Source: World Bank 2007, fig. 1.2. Note: To interpret country abbreviations, see World Bank (2007, p. xviii). 54 From Farm to Firm Figure 3.2 WDR 2008’s Classification of Agriculture-Based, Transforming, and Urbanized Countries 80 agriculture-based BDI countries agriculture’s contribution to growth, 1990–2005 (%) RWA 60 CMR MWI SDN PRY BEN NER NGA CMR INDIA PNG 40 (1965–94) LAO BGR TGO GHA AZE CIV SYR MDG KEN BFA ETH NPL ZMB MOZ GINMLI TCD UGA 20 HND PAK IRN YEM KHM SEN DZA GTM CHINA BRA PHL BOL AGO IDN EGY BGD VNM UKR RUS DOM PER IND (1981–2001) ARG MAR TUN THA LKA VEN SVK TUR CHN CHL BRAZIL CZE SLV ECU INDONESIA ZAR 0 (1970–96) MEX POL ZAF MYS ROM (1970–96) TJK COL HUN BLR transforming countries urbanized countries ZWE –20 0.0 0.2 0.4 0.6 0.8 1.0 rural poor/total poor, 2002 actual poverty data predicted poverty data poverty data over time Source: World Bank 2007, fig. 1.3. Note: To interpret country abbreviations, see World Bank (2007, p. xviii). Spatial transformations have accompanied structural transforma- tions. One outcome has been the increasing regional divides that largely track agricultural versus industrial (rural versus urban) distinctions. Coincidentally, in all three countries the regional division has broadly manifested itself in a north-south divide. In Ghana, the mostly rural northern regions have lagged behind the more dynamic southern regions. In Mozambique, industrial growth has mainly occurred around Maputo in the south, whereas the northern and central areas of the country have remained agricultural centers. In Uganda, firms have emerged mostly in the areas bordering Kampala, and the northern regions have been left behind, in part because of conflict and instability. Another spatial out- come has been the rural-urban divides that are the focus of the rest of this chapter. The extent of urbanization varies in Ghana, Mozambique, and Uganda. According to each country’s definitions, half of Ghana’s popula- tion, just over a third of Mozambique’s population, and about 13 percent of Uganda’s population are now urban.2 Starting from a relatively high level of urbanization of about 23 percent in 1960, the urban share of the National Level: Three Country-Level Perspectives on Rural-Urban Transitions 55 Figure 3.3 Sectoral Shares of GDP, Value Added: Ghana, Mozambique, and Uganda, 1980–2007 a. Ghana 100 90 80 share of GDP (%) 70 60 50 40 30 20 10 0 1980 1985 1990 1995 2000 2005 b. Mozambique 100 90 80 share of GDP (%) 70 60 50 40 30 20 10 0 1980 1985 1990 1995 2000 2005 c. Uganda 100 90 80 share of GDP (%) 70 60 50 40 30 20 10 0 1980 1985 1990 1995 2000 2005 agriculture industry services Source: World Development Indicators, World Bank (July 6, 2009). 56 From Farm to Firm population in Ghana grew rapidly until the 1970s, when it slowed, and then resumed rapid growth in the mid-1980s.3 Mozambique has been rapidly urbanizing since the early 1970s. In Uganda, although overall population growth is extremely high, the level of urbanization has only gradually increased since the 1960s.4 As seen in figure 3.4, for the last Figure 3.4 Population and Urbanization: Ghana, Mozambique, and Uganda, 1960–2007 a. Total and urban population 35 30 Uganda 25 population (millions) Ghana 20 total Mozambique 15 Ghana 10 urban 5 Mozambique Uganda 0 1960 1965 1970 1975 1980 1985 1990 1995 2000 2005 b. Urbanization 50 urban population (% of total) 40 30 20 10 0 1960 1970 1980 1990 2000 Ghana East Asia SSA South Asia Mozambique Uganda Source: World Development Indicators, World Bank (July 2009). SSA = Sub-Saharan Africa. National Level: Three Country-Level Perspectives on Rural-Urban Transitions 57 20–30 years Ghana and Mozambique have been urbanizing at a rate close to that of East Asia, and much more rapidly than Sub-Saharan Africa or South Asia. The size distribution of the 10 largest cities reflects the level of urbanization across the three countries. In line with the urbanization pattern across countries, the estimated total population of the largest 10 cities in 2009 is 6.0 million in Ghana, 4.0 million in Mozambique, and 2.5 million in Uganda (table 3.2). Rural and Urban Poverty Trends Changes in the rural and urban poverty profiles have accompanied struc- tural and spatial transformations. According to national poverty lines as opposed to the standard international ones used in chapter 2, poverty rates have fallen in both rural and urban areas in each country (table 3.3). In fact, all three of the Foster-Greer-Thorbecke (FGT) poverty measures (poverty headcount, poverty gap, and poverty gap squared) decreased from the 1990s to the early to mid-2000s (from 1991–92 to 2005–06 in Ghana, from 1996–97 to 2002–03 in Mozambique, and from 1992–93 to 2005–06 in Uganda).5 Because the large majority of the population is rural, much of the overall national poverty reduction has been driven by the gains in the rural sector, and so the rural-urban poverty divide is converging.6 Thus, although the majority of the poor are still in rural areas, poverty is urbanizing in Mozambique. The fall in the proportion of the poor in rural areas of Mozambique is somewhat overstated in the official statis- tics shown in table 3.3 because of an administrative reclassification of areas from rural to urban between 1996–97 and 2002–03. Nevertheless, the narrowing of rural-urban differences in poverty combined with a gen- uine increase in the urban population share points to the same qualitative result of an increase in share of the poor in urban areas. In Uganda, the rural share of the poor is essentially unchanged at 93–94 percent between the two survey periods, whereas the proportion of the poor in rural areas of Ghana increased slightly, from 82.3 to 85.7 percent between 1991–92 and 2005–06, indicating a ruralization of the poor (table 3.3). Not surprisingly, because these countries can still be classified as agriculture-based economies, poverty reduction in rural areas contributed the most to aggregate poverty reduction. Sectoral decompositions of poverty reduction, as in Ravallion and Huppi (1991), indicate that the rural sector accounted for 70 percent of the aggregate headcount poverty reduction in Ghana, 85 percent in Mozambique, and 90 percent in Uganda. The total intrasectoral effects (i.e., poverty reduction within rural 58 Table 3.2 Ten Largest Cities: Ghana, Mozambique, and Uganda, Selected Census Years Ghana Mozambique Uganda Census Estimate Census Estimate Census Estimate City 1984 2000 2009 City 1997 2007 2009 City 1991 2002 2009 Accra 867,459 1,659,136 2,365,018 Maputo 989,386 1,099,102 1,120,245 Kampala 774,241 1,208,544 1,560,080 Kumasi 496,628 1,171,311 1,852,449 Matola 440,927 675,422 729,469 Gulu 38,297 113,144 208,268 Tamale 135,952 293,879 447,349 Nampula 314,965 477,900 515,320 Lira 27,568 89,871 174,630 Takoradi 85,280 175,438 308,266 Beira 412,588 436,240 441,957 Jinja 65,169 86,520 101,604 Ashiaman 50,918 150,312 271,850 Chimoio 177,608 238,976 253,259 Mbale 53,987 70,437 84,215 Tema 100,052 141,479 175,717 Nacala 164,309 207,894 217,479 Mbarara 41,031 69,208 93,969 Cape Coast 57,224 118,105 175,710 Quelimane 153,187 192,876 200,788 Masaka 49,585 61,300 70,273 Obuasi 60,617 115,564 166,950 Tete 104,832 152,909 164,201 Entebbe 42,763 57,518 70,052 Sekondi 70,214 114,157 171,032 Lichinga 89,043 142,253 155,277 Kasese 18,750 53,446 91,906 Koforidua 58,731 87,315 109,489 Pemba 88,149 141,316 153,900 Njeru 36,731 52,514 75,380 Total 1,983,075 4,026,696 6,043,830 Total 2,934,994 3,764,888 3,951,895 Total 1,148,122 1,862,502 2,530,377 Sources: Census data and estimates from World Gazetteer, http://www.world-gazetteer.com/. Table 3.3 Poverty, Inequality, and Population Distribution: Ghana, Mozambique, and Uganda, Selected Years Ghana Mozambique Uganda 1991–92 2005–06 1996–97 2002–03 1992–93 2005–06 Poverty headcount (SE) Rural 0.636 (0.010) 0.393 (0.009) 0.705 (0.008) 0.549 (0.010) 0.603 (0.008) 0.342 (0.008) Urban 0.278 (0.015) 0.108 (0.007) 0.609 (0.014) 0.521 (0.014) 0.288 (0.015) 0.137 (0.012) Total 0.517 (0.009) 0.286 (0.007) 0.686 (0.007) 0.540 (0.008) 0.564 (0.007) 0.311 (0.007) Poverty gap (SE) Rural 0.240 (0.005) 0.135 (0.004) 0.291 (0.005) 0.205 (0.005) 0.226 (0.004) 0.097 (0.003) Urban 0.074 (0.005) 0.031 (0.003) 0.257 (0.008) 0.190 (0.007) 0.087 (0.005) 0.035 (0.003) Total 0.185 (0.004) 0.096 (0.003) 0.284 (0.004) 0.200 (0.004) 0.209 (0.004) 0.087 (0.002) Poverty gap squared (SE) Rural 0.117 (0.004) 0.066 (0.002) 0.153 (0.003) 0.104 (0.004) 0.112 (0.003) 0.039 (0.001) Urban 0.029 (0.003) 0.013 (0.002) 0.138 (0.006) 0.091 (0.004) 0.037 (0.003) 0.014 (0.002) Total 0.088 (0.003) 0.046 (0.002) 0.150 (0.003) 0.100 (0.003) 0.103 (0.002) 0.035 (0.001) Rural share of poverty (% of total) Poverty headcount 82.3 85.7 81.8 69.0 93.7 93.0 Poverty gap 86.8 87.9 81.6 69.6 94.7 94.3 Poverty gap squared 89.2 89.3 81.2 70.6 95.3 94.3 Gini Rural 0.342 0.408 0.358 0.365 0.328 0.363 Urban 0.347 0.374 0.456 0.463 0.396 0.432 Total 0.373 0.425 0.385 0.404 0.365 0.408 Rural population (%) 66.9 62.4 79.6 67.9 87.6 84.6 Urban population (%) 33.1 37.6 20.4 32.1 12.4 15.4 Source: Authors’ calculations based on survey data. Note: Because of differences in definition, the rural and urban population shares differ from those shown in figure 3.4. SE = standard error. 59 60 From Farm to Firm and urban sectors) accounted for 95 percent or more of the drop in aggre- gate headcount poverty in each of these countries. Similar patterns are also observed for the poverty gap and poverty gap squared measures (see table 3.4). Population shifts from rural to urban areas account for only a relatively small share of the poverty reduction. According to the sectoral decompo- sitions, the population shift effects account for 6.9 percent of poverty reduction in Ghana, 7.8 percent in Mozambique, and 3.7 percent in Uganda. Although population shares have been shifting from rural areas with high poverty rates to urban areas with lower rates, the negative val- ues of the population shift effects suggest that rural-urban population movements have been beneficial for overall poverty reduction (table 3.4). And yet despite progress in poverty reduction, inequality has been on the rise in all three countries. The Gini index has increased within both the rural and urban sectors and for the entire population for the periods evaluated (table 3.3). As explored in much greater detail in chapter 4, urban inequality tends to be greater than rural inequality, as is true for both Mozambique and Uganda. However, in Ghana rural inequality, as measured by the Gini index, was higher than urban inequality in 2005–06. This finding may stem in part from the growing inequalities between the northern and southern regions of the country. Table 3.4 Sectoral Decomposition of Poverty: Ghana, Mozambique, and Uganda, Selected Years Ghana Mozambique Uganda 1991/92–2005/06 1996/97–2002/03 1992/93–2005/06 Abs. diff. % Abs. diff. % Abs. diff. % Poverty headcount –23.1 100.0 –14.6 100.0 –25.3 100.0 Intrasectoral effect –21.9 94.6 –14.2 97.7 –24.7 97.6 Population shift effect –1.6 6.9 –1.1 7.8 –0.9 3.7 Interaction effect 0.3 –1.4 0.8 –5.5 0.3 –1.3 Poverty gap –8.9 100.0 –8.4 100.0 –12.2 100.0 Intrasectoral effect –8.4 94.8 –8.3 98.0 –12.0 98.5 Population shift effect –0.7 8.3 –0.4 4.8 –0.4 3.4 Interaction effect 0.3 –3.1 0.2 –2.8 0.2 –1.9 Poverty gap squared –4.2 100.0 –5.1 100.0 –6.8 100.0 Intrasectoral effect –3.9 94.4 –4.9 97.0 –6.7 98.9 Population shift effect –0.4 9.4 –0.2 3.5 –0.2 3.3 Interaction effect 0.2 –3.7 0.0 –0.5 0.1 –2.1 Source: Authors’ calculations based on survey data. Note: Abs. diff. = Absolute difference. National Level: Three Country-Level Perspectives on Rural-Urban Transitions 61 Although consumption measures of poverty have improved for all three countries, trends for nonmonetary indicators of welfare—school attendance, stunting, and access to water—are mixed (figure 3.5). For example, school attendance rates for children aged 6–15 years have increased, and rural-urban disparities are decreasing in both Mozambique and Uganda. This development may be due in part to the Universal Primary Education initiatives in each country. By contrast, school atten- dance rates in Ghana have fallen, and rural-urban inequalities have remained fairly constant over the 10-year period between surveys. Similarly, progress in improving the health of children under 3, as measured by the prevalence of stunting, has varied across the three countries. In Uganda, the rural-urban gap has narrowed, with improve- ments in rural areas. In Ghana, the rural-urban gap also has narrowed, but unfortunately mainly because of increases in urban areas. In Mozambique, we see a slight divergence between the rural and urban stunting prevalence, with a small increase in rural areas and a small decrease in urban areas. As for indicators of access to improved water sources (such as piped water, tube wells or boreholes, and protected dug wells), in Ghana and Uganda a substantial expansion of access in rural areas has underpinned a narrowing of rural-urban inequalities. But in Mozambique, the rural-urban convergence has been brought about by a decline in access in urban areas. The decline in urban access between 1995 and 2000 may stem in part from a change, following Mozambique’s 1997 census, in the definition of urban to include small towns. However, the further decline between 2000 and 2006 is indicative of a failure of urban infrastructure to expand at a pace commensurate with population growth. In all three countries, large rural-urban inequalities in access to improved water sources remain, with a 21 percentage point difference in Ghana, a 45 percentage point differ- ence in Mozambique, and a 30 percentage point difference in Uganda. In summary, not unlike our cross-country observations in chapter 2, the picture for the three countries examined here is mixed. Even though each country experienced significant economic growth and var- ious degrees of structural and spatial transformation, no consistent pic- ture of rural-urban divergence or convergence emerges. Consumption poverty has fallen in both rural and urban areas of Ghana, Mozambique, and Uganda, but progress on reducing rural-urban inequalities in nonmon- etary measures of poverty—school attendance, stunting, and access to improved water sources—varies across countries. Uganda, which has seen a much slower pace of urbanization than either Ghana or Mozambique, 62 From Farm to Firm has shown progress across all three of the nonmonetary measures, with gains in rural areas exceeding those in urban areas. However, not all indicators in Ghana and Mozambique have improved. In Ghana, school attendance rates dropped in both rural and urban areas, and stunting preva- lence worsened, in particular in urban areas. In Mozambique, stunting in Figure 3.5 Nonmonetary Welfare, Rural versus Urban: Ghana, Mozambique, and Uganda a. School attendance rates in rural and urban areas for 6- to 15-year-olds, selected years 100 90 84 85 87 83 76 75 78 75 80 72 % of rural population 70 67 70 62 60 55 49 50 40 30 20 10 0 1993 1998 2003 1997 2003 1995 2000 Ghana Mozambique Uganda b. Stunting prevalence in rural and urban areas of children under 3 years of age, selected years 50 41 39 40 37 37 % of rural population 30 31 29 30 27 26 27 23 22 22 19 20 16 11 10 0 1993 1998 2003 1997 2003 1995 2000 2006 Ghana Mozambique Uganda rural urban (continued next page) National Level: Three Country-Level Perspectives on Rural-Urban Transitions 63 Figure 3.5 (continued) c. Access to improved water source in rural and urban areas, 1990–2006 1. Rural areas 100 90 80 70 % of population 60 50 40 30 20 10 0 1990 1995 2000 2006 2. Urban areas 100 90 80 70 % of population 60 50 40 30 20 10 0 1990 1995 2000 2006 Ghana Mozambique Uganda Sources: Demographic and Health Surveys; World Development Index, World Bank (July 2009). rural areas increased slightly, and access to improved water sources in urban areas decreased. Despite substantial progress in poverty reduction and the signs of early structural transformation, the agriculture sectors of these countries are still large, and the manufacturing and service sectors are still in an early stage of development. Any convergence or divergence of different living stan- dards indicators over the last 15 years is small relative to the rural-urban gaps in these indicators. 64 From Farm to Firm Sources of Rural-Urban Inequalities What drives rural-urban inequalities? One way to approach that ques- tion is to consider two polar explanations for why some areas of a country are poorer or richer than others. At one end of the spectrum is the “concentration effect”—that is, persistently poor areas emerge from the spatial concentration of people with low levels of productive assets, including human, physical, financial, and social capital. According to this view, poor people in poor areas would most likely be poor even if they lived in richer areas. At the other end of the spectrum is the “geography effect,” which attributes persistent geographic differ- ences in living standards to differences in spatial characteristics such as endowments of local public goods and services (e.g., transport, electric- ity, and water). In other words, differences in living standards are caused by spatially determined differential returns to assets, so that two households with identical observable assets would have different standards of living because the returns to those assets would be higher in the better-endowed area. This line of analysis has been conducted for several countries in other regions, but to our knowledge not in Sub-Saharan Africa. In a study of Bangladesh, Ravallion and Wodon (1999) found that the geography effect dominated rural-urban welfare differences, with pronounced differences in returns to household characteristics such as education. A subsequent study of Bangladesh by Shilpi (2008) examined not only rural-urban dif- ferences, but also differences between a region that is integrated with urban growth poles and a more isolated region that is cut off from the growth centers by major rivers. This study found large geographic differ- ences in returns between integrated and isolated regions. The differential returns were particularly pronounced among higher-income households, which are related to differential public capital and market access. Among poorer households, the geography effect is important in explaining inequalities between integrated and isolated regions, but not inequalities between rural and urban areas within each of these regions. The policy implication is that investments in the human capital of poor people can help mitigate rural-urban disparities within regions, but that investments in connective infrastructure are more important for reducing inequalities between integrated and isolated regions. The investments in connective infrastructure facilitate the flow of goods and migrants between regions, which also stimulates the development of growth poles in urban areas of the more isolated region. National Level: Three Country-Level Perspectives on Rural-Urban Transitions 65 A recent regional study by Skoufias and Lopez-Acevedo (2008) looks at these issues in eight countries in Latin America. They find that the concentration effect is the dominant source of inequalities between rural and urban areas within regions, indicating a need for targeted poli- cies to improve the human capital of poor households. However, like Shilpi (2008), their study reveals that large differences in returns (i.e., a dominant geography effect) are responsible for most of the inequali- ties between regions. Thus, policies to remove impediments to labor mobility—such as connective infrastructure and improving the effi- ciency of credit, land, and labor markets—are appropriate. Methodology To better understand the sources of rural-urban welfare inequalities in the three Sub-Saharan countries of interest here, we examine whether welfare differences are attributable primarily to concentration effects or geography effects. We decompose differences in average household wel- fare (measured in log welfare ratios, which is the log of household con- sumption per adult equivalent as a proportion of local poverty lines) into components attributable to household endowments and the returns to those endowments using Oaxaca-Blinder decompositions (details on the data and methodology appear in appendix B).7 For this analysis, we limit ourselves to inequalities in consumption per adult equivalent for several reasons. First, it is a more comprehensive measure of welfare than the other measures examined in the previous section such as school enroll- ment, stunting, or access to water. Second, unlike stunting or school enrollment, consumption permits a complete ordering of all households in the data set. Third, consumption is a continuous measure, allowing finer distinctions in the welfare levels of households. According to the classic Lewis and Harris-Todaro models, rural-to- urban migration flows are a function of wage or expected income differ- entials between rural and urban areas. Similarly, rural-urban differences in the marginal welfare benefits (returns) to certain household endow- ments may be considered an indicator of the incentives that might exist for rural inhabitants to migrate to urban areas. Expectations of marginal welfare gains associated with moving from rural to urban areas may be partially represented by the differences in coefficients for relevant house- hold endowments in our regressions.8 Such differences indicate that the geography effect dominates. By contrast, the absence of differences in 66 From Farm to Firm coefficients suggests that the concentration effect is the primary expla- nation for rural-urban welfare inequalities. Results Our findings indicate that in Mozambique and Uganda rural-urban inequalities in average household welfare are primarily attributable to the concentration effect—or differences in household endowments. In Ghana, the sources of rural-urban inequalities are divided almost equally between the concentration and geography effects, or endowments and returns to those endowments.9 In Mozambique, endowments accounted for over 100 percent of the average rural-urban welfare difference in both 1996–97 and 2002–03.10 In Uganda, endowments accounted for 77 and 73 percent of the rural-urban difference in average welfare in 1992–93 and 2005–06, respectively. In Ghana, returns accounted for 67 percent of the difference in 1991–92 but dropped to 47 percent in 2005–06, so that differences in endowments and returns accounted for roughly equal shares (see table B.5 in appendix B). By decomposing rural-urban welfare differences, we can identify fac- tors that play a relatively large role in accounting for these welfare inequalities. Using the Oaxaca-Blinder decompositions, the endowments and returns components can each be further disaggregated by categories of variables (see figure 3.6 for the results). The vertical axis of the figure represents the rural-urban difference in mean (log) welfare ratios, and each component of the bars represents a contribution to the overall rural- urban difference. Therefore, the sum of the positive and negative subcom- ponents of both the endowment and return (coefficients) components will equal the overall rural-urban difference in mean welfare. Education and the sector of employment of household heads are typically the two endowment factors that contribute substantially to rural-urban welfare inequalities. Interpretation of the sector of employment variable is ambiguous because it is to some extent tied to location (e.g., the agriculture sector is predominantly rural) and because it is a characteristic that is more amenable to change than other characteristics. The sector of employ- ment variable could appear to capture a particular set of skills, a com- bination of innate ability and accumulated experience. Alternatively, it may be viewed as a realization of human capital and a characteristic that the individual can change in response to the different opportuni- ties available. By including sector of employment as a regressor we are National Level: Three Country-Level Perspectives on Rural-Urban Transitions 67 Figure 3.6 Oaxaca-Blinder Decomposition of Urban-Rural Welfare Differences: Ghana, Mozambique, and Uganda, Selected Years 1.0 urban-rural difference in (log) welfare ratios 0.8 0.6 0.4 0.2 0 –0.2 –0.4 –0.6 –0.8 –1.0 coef. end. coef. end. coef. end. coef. end. coef. end. coef. end. 1991–92 2005–06 1996–97 2002–03 1992–93 2005–06 Ghana Mozambique Uganda education sector of employment age household size and composition community region constant terms Source: Authors’ calculations. Note: coef. = coefficient; end. = endowment. interpreting sector of employment as a skill, reflecting the individual’s skills and experience. In all three countries, despite higher returns to education in urban areas, human capital inequalities play a bigger role in accounting for overall rural-urban welfare inequalities. As shown in figure 3.6, the con- tribution of educational attainment differences in the endowments com- ponent is consistently larger than the contribution of education in the coefficients component for each of the three countries and across years. This indicates that the differences in the level of educational attainment of household heads account for more of the mean rural-urban welfare inequalities than do the differences in returns to education. The scarcity of nonfarm employment in rural areas rather than rural- urban differentials in returns for the same sector explains a large share of rural-urban welfare inequalities. Although differences in returns exist between rural and urban areas, these differences are often negligible in 68 From Farm to Firm explaining rural-urban welfare inequalities. For example, in figure 3.6 the sector of employment component of the returns effect is very small in most cases. By contrast, differences in sectors of employment make up a relatively large share of the endowment effect for all three coun- tries. The returns differential is marginalized by the fact that employ- ment in agriculture dominates in rural areas. For example, 88 percent of rural household heads in Mozambique were employed in agriculture in 2002–03. In principle, the higher returns in urban areas to education and in the secondary and tertiary sectors in these countries should draw labor from rural areas to urban areas until returns are equalized. However, rural- urban differentials in returns persist, although these differentials are small relative to the differences in endowments in Mozambique and Uganda. Because migration decisions are based on expected net benefits, not just wage differentials, the low expected probability of finding a good job in the city in view of lower rural education levels and high urban unemployment levels may be reducing the incentives for rural-to- urban migration. Other factors such as migration costs, insufficient land rights in both rural and urban areas, and social segmentation may also be barriers to migration. Across all three countries, differences in returns to age contribute to much of the rural-urban welfare difference. The positive differentials in figure 3.6 indicate that the returns to age are more favorable in urban areas. Because there are no substantial differences in the mean age between rural and urban areas, the contribution of age to the endow- ments component is negligible. In urban areas, it is plausible that wage earners earn more as they gain seniority and accumulate skills with age, and that those in the informal sector accumulate assets and social cap- ital with age. However, in rural areas the output of farmers may fall with age as the ability to work hard or improve farming techniques diminishes with age. Not surprisingly, farmers tend to be better off in rural areas than in urban areas, where little land is devoted to agriculture. Although some high-value agriculture is carried out in or near cities, more often agricul- ture is a sector of last resort in urban areas, and often represents under- employment. In Mozambique and Uganda, a negative difference for the constant term indicates that household heads with no formal education engaged in agriculture in rural areas of the reference region were better off than their urban counterparts (i.e., uneducated household heads engaged in agriculture in urban areas). In Ghana, uneducated household National Level: Three Country-Level Perspectives on Rural-Urban Transitions 69 heads in agriculture in the rural forest zone were slightly worse off in 1991–92 than their urban counterparts in the forest zone, but this trend had reversed by 2005–06. Negative coefficient differentials for the regional variables indicate that, on average, the marginal benefit of rural households being in non- reference regions (relative to the reference region) tend to be greater than the marginal benefits of urban households being in nonreference regions (relative to the reference region) controlling for other factors. For example, in Mozambique the fact that the breadbasket of the country is the central and northern regions may explain why rural households in those regions tend to be better off than rural households in the southern region, controlling for other factors. For urban areas, because Maputo, the capital city and largest agglomeration, is located in the southern region, it makes sense that, on average, urban households in the central and northern regions tend to be not as well-off as urban households in the southern region. Decomposition of Welfare Differences over Time Increases in rural welfare over time stem primarily from the increasing returns in all three countries. The endowments used as regressors tend to change slowly over time—and slower than rising consumption in countries growing as rapidly as these three. Therefore, the share of the difference attributable to the coefficients (returns) will increase almost by construction. More interesting is seeing for which endowments the increase in returns is most significant. In rural areas, the mean welfare ratio (consumption per adult equivalent as a proportion of the poverty line) increased in Ghana from 1.27 to 1.98 between 1991–92 and 2005–06, in Mozambique from 1.08 to 1.29 between 1996–97 and 2002–03, and in Uganda from 1.15 to 1.84 between 1992–93 and 2005–06 (see table B.4 in appendix B). When these differences are decomposed, the returns component accounts for over 90 percent of increases in all three coun- tries. In Ghana and Uganda, greater returns to age of the household head account for much of the rural welfare increases, whereas in Mozambique increases in returns in the central region play a large role (see table B.5). As for improvements in average household endowments, progress in educational attainment occurs mostly in urban areas, although Uganda shows improvements in both rural and urban areas. Also, because of the high population growth rate in Uganda, the average number of adult equivalents per household increased from 3.6 to 4.0 in rural Uganda and from 3.1 to 3.6 in urban Uganda (see table B.4). 70 From Farm to Firm Similarly, increases in returns over time account for the majority of the increases in urban welfare. In urban areas, mean welfare ratios increased in Ghana from 2.22 to 3.46 between 1991–92 and 2005–06, in Mozambique from 1.39 to 1.77 between 1996–97 and 2002–03, and in Uganda from 2.07 to 3.79 between 1992–93 and 2005–06. When these differences were decomposed, the returns component accounted for 70 percent of the increase in Ghana, 79 percent of the increase in Mozambique, and all of the increase in Uganda. Similar to findings for rural areas, greater returns to age of the head of household accounted for much of the urban welfare increases in Ghana and Uganda, whereas increases in returns in the north- ern and central regions played a large role in Mozambique (see table B.6 in appendix B). Decomposition of Distributional Changes in Rural-Urban Inequalities Because the relative importance of returns and endowments in explain- ing rural-urban inequalities may vary considerably for different income groups, we explore the extent of these variations across the distribution by applying the quantile decomposition. This process involves construct- ing counterfactual welfare distributions and comparing them with the empirical rural and urban welfare distributions (see Machado and Mata 2005 for an example). The counterfactual distribution represents what a welfare distribution might look like if the rural population had obtained urban returns to their (rural) characteristics. If we then compare the counterfactual and rural distributions, the difference between the two distributions can be attributed to differences in returns because charac- teristics should be the same. Similarly, the difference between the coun- terfactual and urban distributions can be attributed to differences in characteristics. The cumulative distributions of welfare (in terms of household consumption per adult equivalent) in figure 3.7 indicate dif- ferent patterns for Ghana, Mozambique, and Uganda. In Ghana, for 1991–92 the counterfactual distribution runs very close to the actual urban distribution, which demonstrates that differences in the returns to household characteristics account for most of the rural- urban inequalities in welfare for that year. However, for 2005–06 the counterfactual distribution lies about midway between the actual rural and urban distributions, indicating that differences in household charac- teristics and returns to those characteristics account for about equal shares of the welfare inequalities for that year. The quantile regression approach allows the decomposition between endowments and returns to National Level: Three Country-Level Perspectives on Rural-Urban Transitions 71 Figure 3.7 Cumulative Distributions of Welfare (Household Consumption per Adult Equivalent): Ghana, Mozambique, and Uganda, Selected Years a. Ghana, 1991–92 b. Ghana, 2005–06 1.0 1.0 proportion of population 0.8 0.8 0.6 0.6 0.4 0.4 0.2 0.2 0.0 0.0 –2 –1 0 1 2 3 –2 –1 0 1 2 3 log welfare ratio c. Mozambique, 1996–97 d. Mozambique, 2002–03 1.0 1.0 proportion of population 0.8 0.8 0.6 0.6 0.4 0.4 0.2 0.2 0.0 0.0 –2 –1 0 1 2 –2 –1 0 1 2 log welfare ratio e. Uganda, 1992–93 f. Uganda, 2005–06 1.0 1.0 proportion of population 0.8 0.8 0.6 0.6 0.4 0.4 0.2 0.2 0.0 0.0 –1 0 1 2 –1 0 1 2 log welfare ratio rural counterfactual urban Source: Authors’ calculations from household survey data. Note: These are distributions of households rather than individuals, and therefore the poverty headcount cannot be discerned directly from these cumulative distributions. Because poorer households tend to be larger, headcount poverty rates will typically exceed the share of poor households. 72 From Farm to Firm vary across the income spectrum, and so it is striking that the proportions are so similar at all income levels in Ghana. This finding contrasts with those elsewhere, such as for Bangladesh where endowments account for a much larger share of the rural-urban welfare difference among the low- est income groups (Shilpi 2008). This result suggests that if the govern- ment seeks to reduce rural-urban inequalities, the same types of policies could be directed at all rural groups as opposed to focusing on human capital development for lower-income groups and interventions to increase mobility for higher-income groups. In Mozambique, there is very little difference in the actual rural and urban welfare distributions among the poor (the portions of the curves to the left of the vertical line) for both 1996–97 and 2002–03. The rural and urban distributions below the poverty line almost overlap for both 1996–97 and 2002–03, but they widen for the upper parts of the distri- bution. The lack of any dramatic changes in the lower half of the distri- bution is attributable in large part to the relatively short period between the surveys, only six years. Rural household heads in Mozambique are predominantly uneducated farmers, and the estimated returns for uneducated farmers are much lower in urban areas. Thus, simply obtaining urban returns without any change in employment sector or other endowments does not appear to benefit rural households. For this reason, the counterfactual distributions in both years lie to the left of the rural distributions and suggest that the rural households would be worse off if they received urban returns for their low endowment levels. This perhaps overly literal interpretation of the results highlights a limitation of this static decomposition: it does not account for the switching of employment sectors that usually occurs with rural-to-urban migration. In Uganda, differences in endowments account for most of the rural- urban inequalities throughout the distributions for both 1992–93 and 2005–06. A comparison of the counterfactual distributions and the urban distributions in figure 3.7 reveals that differences in endowments account for most of the observed rural-urban inequalities. The role of differences in endowments appears to be slightly larger at the lower end of the distribution. The role of endowments in accounting for the large majority of welfare inequalities persists from 1992–93 to 2005–06. In Uganda, although the standard of living has improved for both the rural and urban segments of the country, the degree of inequality between the rural and urban distributions is large and increases from 1992–93 to 2005–06, as illustrated by the wider gap between the rural and urban National Level: Three Country-Level Perspectives on Rural-Urban Transitions 73 distributions in 2005–06. Also, inequality within rural and urban areas increases over this period. Implications for Policy In this chapter, we have sought to gain deeper insights into rural-urban transitions in three African countries that have experienced robust levels of growth over the last two decades. For each country, economic growth and poverty reduction have been accompanied by significant, albeit vary- ing, levels of urbanization. And, as in almost every country of the world, urban areas are outperforming rural areas according to various measures of welfare. However, beyond these broad similarities each country has a different story to tell about changes in rural and urban welfare. When measured by national poverty lines, consumption poverty has fallen in both rural and urban areas, and at a faster rate in rural areas in Mozambique and Uganda, indicating a degree of convergence. But nonincome measures paint a much more complicated picture: convergence, where it has occurred, has resulted from strong improvements in rural areas (which appears to be the case for school attendance in Mozambique and Uganda) as well as deterioration in urban areas (which appears to be the case for stunting in Ghana or water supply in Mozambique). Thus, both the 41-country overview provided in chapter 2 and the more in-depth study here indi- cate that consistent trends are not to be expected, either across countries at similar stages of development or even within countries, across different welfare measures. The Oaxaca-Blinder decompositions and quantile regression extensions are useful for quantifying the relative contributions of the concentra- tion and geography effects to rural-urban inequalities in consumption. However, they are limited in that they are static decompositions that provide an accounting for inequalities, but are not based on a causal model of welfare outcomes and their determinants. The results should also be interpreted with caution because of the high potential for omitted-variable bias in the regressions. This effect biases the coefficient estimates and in turn the relative proportions of characteristics and returns to characteris- tics. This analysis is probably best viewed as a preliminary step that is sug- gestive of broad areas for policy and a source of guidance for more in-depth inquiry into specific policy areas such as factor markets or institutions. Of the three countries, the decomposition results in Uganda are the most straightforward: individual and household characteristics account 74 From Farm to Firm for a large proportion of rural-urban inequality. This finding parallels the results obtained for Latin America by Skoufias and Lopez-Acevedo (2008) and for low-income households in Bangladesh by Shilpi (2008). Returns to characteristics are not very different across rural and urban areas, but human capital stocks are—most notably in education. Education levels are low in Uganda and particularly so in rural areas. This situation suggests putting a high priority on human capital development, which is consistent with the government policy of universal primary edu- cation that was launched in 1997 and expanded to postprimary education and training in 2007. These policies take time to bear fruit, and school quality remains an issue. Although better rural education is likely to reduce rural-urban inequal- ities, it is not a panacea for Uganda’s growth and development. The pop- ulation is young and growing faster than 3 percent a year, and job creation will have to keep pace to sustain the current rate of poverty reduction. Complementary public and private investments to accelerate job cre- ation, and returns to employment, are needed as well. Uganda’s popula- tion is largely rural, and a recent computable general equilibrium (CGE) analysis by Dorosh and Thurlow (2009) points to the important role that greater agricultural productivity plays in generating broad-based welfare improvements in rural areas. Other work by Lall, Schroeder, and Schmidt (2009) indicates that, among urban areas, concentrating infrastructure investments in the Kampala corridor is likely to have the greatest impact on growth and employment. The decomposition analysis is the least informative for Mozambique, where consumption inequality between rural and urban areas is remark- ably low, especially among low-income groups. The decomposition results indicating concentration effects in extremis should not be taken at face value, because the small denominator leads to volatile results. Although the decompositions are not entirely trustworthy, based on results in other countries one would expect concentration effects to indeed dominate in a poor country such as Mozambique, which also has very low levels of education and other human capital. The data on nonmonetary rural- urban welfare gaps presented earlier in this chapter reveal that, despite similar consumption levels, rural areas of Mozambique trail urban areas significantly in school enrollments, child nutrition, and access to safe water. All these findings indicate that high priority should be given to human capital investments in rural areas of Mozambique. Perhaps as important as the policies suggested by these findings are those not suggested. In particular, the evidence that returns to education National Level: Three Country-Level Perspectives on Rural-Urban Transitions 75 are roughly equal in rural and urban areas suggests that labor market seg- mentation or barriers to migration are not significant factors in explaining the spatial divide. These results indicate that regulatory interventions in labor markets are not likely to ease rural-urban divides; indeed, such interventions could create more problems than they solve. The principal challenges appear to be twofold: (1) how to ensure that the rural labor force has sufficient endowments of human capital to take advantage of income-generating opportunities, and (2) how to create an enabling envi- ronment for job creation in both the rural and urban sectors. Ghana is the only one of the three countries that shows much differ- ence between the two survey years in the decomposition of rural-urban consumption inequalities. Differences in decompositions between sur- veys could arise from normal variability, such as a particularly good or bad crop year, or could reflect an evolution in the sources of rural-urban inequality. Many explanations are possible for the shift from a strong geography effect (returns) in 1991–92 to the almost equal balance between concentration and geography in 2005–06. For example, it would be consistent with a reduction in the inequality of returns, say from improved labor mobility. Unfortunately, that does not appear to be the case. Closer examination of the underlying data shows that during this period both endowments and returns increased in absolute terms in both rural and urban areas.11 Endowments and returns increased faster in urban areas, which is consistent with the increasing gap in mean con- sumption observed. Of the two components, the rural-urban gap in endowments increased much faster than the gap in returns, which accounts for the leftward shift of the counterfactual line in 2005–06. Because migration is significant in Ghana, it is important to understand how migration is related to this divergence in endowments. Could it be that rural educational services are lagging behind urban ones, or could it be that the rural youth moving to urban areas are disproportionately well-educated? Certain research areas lie beyond the scope of this chapter but are worthy of in-depth exploration. As noted earlier, spatial inequalities are not limited to rural-urban distinctions; they also have a north-south gra- dient in all three countries. We have not systematically investigated the nature and sources of regional welfare inequality in these countries, but that will be important for informing possible directions for policy. And rural land tenure systems, which have not been discussed here, almost certainly affect incentives to migrate from rural to urban areas. The cus- tomary land tenure systems that prevail in many Sub-Saharan African 76 From Farm to Firm countries are similar in many ways to the system in Sri Lanka, which has had a major impact on the pace of rural-urban transformation in that country, as described in chapter 8. Notes 1. The other five countries are Botswana, Equatorial Guinea, Ethiopia, Sudan, and Tanzania. 2. In Ghana, the 1960, 1970, 1984, and 2000 censuses defined urban as locali- ties with at least 5,000 inhabitants. In Mozambique, the 1980 census treated 12 cities as urban (Maputo, the nine provincial capitals, Nacala-Porto, and Chokwe), and the 1997 census expanded urban areas to include 23 cities and 68 towns. In Uganda, the 1980 and 1991 censuses defined urban as cities, municipalities, towns, town boards, and all trading centers with more than 1,000 inhabitants. The 2002 census changed the definition to gazette cities, municipalities, and towns with more than 2,000 inhabitants (United Nations 2008). Despite the low “urban” population threshold in Uganda, many observers note that Uganda is more urban than indicated by statistics based on administrative classifications. For example, the agglomeration index in the WDR 2009 classifies 28 percent of Uganda’s population as urban (World Bank 2008). 3. Ghana’s urban population share of 50 percent would make it an “intermedi- ate” rather than “incipient” urbanizer in the WDR 2009 classification, but it also shares important characteristics with other incipient urbanizing coun- tries, such as a sparsely populated lagging region in the north and the rela- tively large contribution of agriculture to GDP growth (figure 3.2). 4. These estimates may differ from household survey estimates. 5. Periods spanning multiple calendar years (for example, 1991–92 or 2005–06) are used for reporting these and other survey results because the household surveys took place over 12- to 14-month periods that were not confined to a single calendar year. 6. Note that the analysis in chapter 2 uses slightly different national poverty lines for cross-country comparisons and can provide slightly different results. Also, the analysis in this chapter is based on adult equivalents unless stated otherwise. 7. The application of Oaxaca-Blinder decompositions and quantile regression extensions is similar to the work by Machado and Mata (2005) on the counterfactual decomposition of changes in wage distributions; more recent works using the quantile regression decomposition include those by Nguyen et al. (2007), Skoufias and Lopez-Acevedo (2008), and Shilpi (2008). National Level: Three Country-Level Perspectives on Rural-Urban Transitions 77 8. When welfare is measured as consumption per adult equivalent in multiples of the local poverty line, spatial differences in preferences and prices are considered in the welfare measure. Although regional price differences may be taken into account in the welfare ratio measures, the static decompositions do not account for factors such as switching employment sectors, changes in the probability of finding employment, the costs associated with migration, support from social networks, or the presence of various push factors. 9. The regression results that underpin the decompositions for Ghana, Mozambique, and Uganda appear in appendix B (tables B.1, B.2, and B.3, respectively). Table B.4 summarizes characteristics of rural and urban house- holds in each of the three countries. 10. In Mozambique, the rural-urban difference in average welfare is very small. When decomposing such very small differences, the results should be inter- preted with some caution. To estimate the share of the welfare difference accounted for by either endowments or coefficients, one divides these com- ponents by the difference in average welfare. When the denominator is very small, as it is for Mozambique, it is quite easy to obtain percentages larger than 100 for one of the effects (concentration or geography) and offsetting negative percentages for the other effect. 11. This result from the 1991 and 2005 Ghana Living Standards Surveys (GLSS) is not consistent with the Demographic and Health Surveys (DHS) data on school attendance in figure 3.5. The reason for the discrepancy is not clear, but it could be attributed to the different reference populations: school-age children in the DHS and heads of household in the decomposition analysis. References Dorosh, P., and J. Thurlow. 2009. “Agglomeration, Migration, and Regional Growth: A CGE Analysis for Uganda.” IFPRI Discussion Paper 848, International Food Policy Research Institute, Washington, DC. Lall, S. V., E. Schroeder, and E. Schmidt. 2009. “Identifying Spatial Efficiency-Equity Tradeoffs in Territorial Development Policies: Evidence from Uganda.” Policy Research Working Paper 4966, World Bank, Washington, DC. Machado, J. A. F., and J. Mata. 2005. “Counterfactual Decomposition of Changes in Wage Distributions Using Quantile Regression.” Journal of Applied Econometrics 20: 445–65. Nguyen, B. T., J. W. Albrecht, S. B. Vroman, and M. D. Westbrook. 2007. “A Quantile Regression Decomposition of Urban-Rural Inequality in Vietnam.” Journal of Development Economics 83: 466–90. Ravallion, M., and M. Huppi. 1991. “Measuring Changes in Poverty: A Methodological Case Study of Indonesia during an Adjustment Period.” World Bank Economic Review 5 (1): 57–82. 78 From Farm to Firm Ravallion, M., and Q. Wodon. 1999. “Poor Areas or Poor People?” Journal of Regional Sciences 39 (4): 689–711. Shilpi, F. 2008. “Migration, Sorting and Regional Inequality: Evidence from Bangladesh.” Policy Research Working Paper 4616, World Bank, Washington, DC. Skoufias E., and G. Lopez-Acevedo. 2008. “Sources of Welfare Disparities within and between Regions in Latin America and Caribbean Countries. Volume 1: Synthesis.” Draft, World Bank, Washington, DC, December 1. United Nations. 2008. Urban Indigenous Peoples and Migration Factsheet. http://www.un.org/esa/socdev/unpfii/documents/factsheet_migration_final .pdf. World Bank. 2007. World Development Report 2008: Agriculture for Development. Washington, DC: World Bank. ———. 2008. World Development Report 2009: Reshaping Economic Geography. Washington, DC: World Bank. CHAPTER 4 Local Level: Intraurban Welfare Disparities and Implications for Development According to the World Development Report 2009 (WDR 2009), divisions between leading and lagging areas play out not only on a national scale but also within cities (World Bank 2008c). This chapter elaborates on this proposition using data on the capital cities of Accra, Kampala, and Maputo, each of which forms the largest urban agglomeration in its country. In the first section of this chapter, we consider the demographic and physical changes these cities have undergone over past decades. In the second section, we consider income and nonincome measures of wel- fare in the capitals versus the countries as a whole. Although the pri- mary cities do offer higher average levels of welfare and services than rural areas, we also find significant within-city disparities, as predicted by the WDR 2009. These disparities are the topic of the third section, where we look at income inequalities and inequality in access to services and health outcomes across income quintiles. In the fourth section, we discuss how welfare disparities are linked to location: the “bumpy” economic landscape of the city is manifested in the juxtapo- sition of well-serviced modern residences with slums. In the fifth sec- tion, we analyze why these slums exist, looking at the impact of colonial legacies, land policies, and weak governance on intraurban inequalities. A conclusion follows. 79 80 From Farm to Firm Growing Cities Each of the cities featured here has undergone profound demographic and physical changes over the last half-century. Figure 4.1 captures the precip- itous rate of population growth since the 1960s. At the time of Ghana’s independence, Accra’s population numbered about 260,000. The current population of the Accra agglomeration, by UN estimates, is 2.4 million, although the metropolitan government claims that the city accommodates up to 3 million people on any given day.1 In Uganda, Kampala’s population was an estimated 100,000 in the early 1960s (Nilsson 2006). With a cur- rent population of about 1.6 million, it is by far the largest city in Uganda (the next largest urban center has less than a tenth of Kampala’s popula- tion). Mozambique is rapidly urbanizing. In 2005, it was the fourth least urbanized country in southern Africa, but by 2025, it is projected to be the fourth most urbanized country. Maputo had a population of about 100,000 in the early 1960s and under 400,000 at the time of independence in 1975 (Grest 2006). The United Nations estimates the current population of the Maputo agglomeration to be about 1.8 million.2 As populations grow, cities must inevitably increase in size or density. Although all three of these cities have experienced an expansion of the urban periphery, Accra appears to have undergone the greatest spatial transformation over the last two decades, particularly to the west of the city center. This transformation may be in part a reaction to ambiguous land laws (discussed in more detail later in this chapter) that may encourage buyers to select land in an area that is unaffected by outstanding disputes (World Bank 2008a). Figure 4.1 Population Growth: Accra, Kampala, and Maputo, 1970–2009 3,000 population (thousands) 2,500 2,000 1,500 1,000 500 0 1970 1980 1990 2000 2005 2007 2009 Accra Kampala Maputo Source: United Nations 2010. Local Level: Intraurban Welfare Disparities and Implications for Development 81 Lower Poverty, Better Services Population booms in these cities are attributable in part to a reconfigura- tion of city boundaries, in part to natural increase, and in part to high rates of in-migration from rural areas. Rural migrants are drawn to cities with the expectation of higher standards of living, and average measures of income and nonincome welfare tend to bear out this expectation. For example, poverty indicators are better in the capital city than in the coun- try as a whole (see table 4.1). Over the period 1991–2006, Accra was the site of impressive levels of poverty reduction. At 10.6 percent, the poverty headcount in 2005–06 was less than half of what it was in 1991. But contrary to the trend in the rest of the country, where poverty has been steadily declining, in Accra it has been rising since the late 1990s. Indeed, the poverty headcount and poverty gap were higher in 2005–06 than in 1998–99. Nevertheless, the absolute poverty levels remain much lower in the capital than in the rest of the country. Likewise, in Kampala the poverty headcount and poverty gap declined dramatically over the course of the 1990s. Although rural areas also saw high rates of poverty Table 4.1 Poverty in Capitals versus Nationwide: Ghana, Uganda, and Mozambique Poverty headcount (%)a Gini coefficient a Accra-GAMAb (Ghana) 1991–92 23.1 (51.7) 0.364 (0.373) 1998–99 4.4 (39.5)c 0.323d (0.425),e 0.269f (0.388)g 2005–06 10.6 (28.5) 0.406 (0.425) Kampala (Uganda) 1992–93 20.8 (56.4) 0.394 (0.365) 2002–03 7.8 (38.8) 0.481 (0.428) 2005–06 5.5 (31.1) 0.392 (0.408) Maputo (Mozambique) 1996–97 47.8 (69.4) 0.444 (0.396) 2002–03 53.6 (54.1) 0.524 (0.415) Note: The first values in each column are for the capitals, and the nationwide values appear in parentheses. a. Authors’ calculations based on Living Standard Measurement Surveys except where otherwise noted. b. GAMA is the Greater Accra Metropolitan Area, which includes the Accra Metropolitan Area, Tema Municipal Area, and the urban areas in the Ga East and Ga West districts. c. Because survey data are not available, data were taken from Ghana Statistical Service 2007, table 2. d. Adjasi and Osei 2007, table X. e. Tuffour and Apallo 2005, table 2. f. Maxwell et al. 2000. g. Aryeetey and McKay 2007. 82 From Farm to Firm reduction, in 2005–06 absolute poverty was far lower in the capital (with a headcount of 5.5) than in rural areas (34.2) or the country as a whole (31.1). In terms of monetary welfare, Maputo, where urban poverty worsened between 1996–97 and 2002–03, is an exception to the largely positive story shown in table 4.1. In 1996–97, the poverty headcount of over 47 percent in the capital was significantly lower than that of 71.2 percent in rural areas. Poverty was also less severe, with a poverty gap of 16.5 in Maputo versus a gap of 29.9 in rural areas (Arndt, James, and Simler 2006). But by 2002–03, poverty in the capital had grown in both preva- lence and severity, while poverty in rural areas and the country as a whole had decreased significantly. As a result, Maputo had a level of poverty that was strikingly similar to that in the country as a whole in 2002–03. Like the poverty indicators, measures of nonincome welfare—such as access to services—tend to be much more favorable in urban areas than in rural areas. Because urban populations are dense, the per capita costs of many forms of infrastructure are lower, and competition among alter- native service providers can be a force for innovation and efficiency (Kessides 2006). As figure 4.2 indicates, infrastructure services are much better in the capital cities, and generally improved there during the peri- ods between the Demographic and Health Surveys used in this chapter. In all three countries, access to electricity, piped water, and toilet facil- ities is much higher in the capital city than in the country as a whole. In Uganda, for example, access to electricity is almost eight times higher in Kampala than across the whole country, and access to piped water is almost six times higher. In Maputo, access to all services is higher than across Mozambique, but lower than access in Kampala and Accra. In Accra, the vast majority of residents have access to electricity and water, whereas less than half of Ghana’s population as a whole benefits from similar access. Access to flush toilet facilities is scarcer even in capital cities. But here, too, people in the capitals have higher levels of access than in the countries as a whole. Welfare Inequalities These capital cities provide higher levels of income and nonincome wel- fare on average, but they also exhibit high levels of inequality. This find- ing is by no means atypical. In the developing world, income inequalities in cities are generally high: a subset of 26 cities in Africa and 19 in Latin America have average Gini coefficients of 0.54 and 0.55, respectively Local Level: Intraurban Welfare Disparities and Implications for Development 83 Figure 4.2 Access to Services in Capitals versus Nationwide: Ghana, Uganda, and Mozambique, Selected Years a. Access to infrastructure services, Accra and Ghana 100 80 60 percent 40 20 0 Accra Ghana Accra Ghana Accra Ghana electricity piped water flush toilet 1998 93.4 39.3 97.1 36.5 35.4 7.5 2008 90.6 43.9 59.4 40.7 46.7 14.1 100 b. Access to infrastructure services, Kampala and Uganda 80 60 percent 40 20 0 Kampala Uganda Kampala Uganda Kampala Uganda electricity piped water flush toilet 1995 55.1 7.1 57.2 6.1 12.0 1.6 2006 61.2 7.8 75.4 13.2 12.2 1.2 100 c. Access to infrastructure services, Maputo and Mozambique 80 60 percent 40 20 0 Maputo Mozambique Maputo Mozambique Maputo Mozambique electricity piped water flush toilet 1996 43.4 6.0 45.4 6.5 27.9 4.1 2002 48.0 8.9 62.5 16.9 35.4 4.7 Sources: Demographic and Health Surveys for Ghana and Uganda; Living Standards Survey for Mozambique. 84 From Farm to Firm (UN-Habitat 2008). Even in some parts of the developed world—in particular, the United States—some cities exhibit stark income inequal- ities. Although income inequality, as measured by the Gini coefficient, is relatively low in Accra (0.364 in 1991–92, and 0.269 in 1998–99), it is significantly higher—and increasing—in both Kampala and Maputo (table 4.1). In Uganda, inequality increased in both rural and urban areas between 1992–93 and 2002–03. The story is grimmer still for Mozambique, the poorest of the three countries. There, between 1996–97 and 2002–03 inequality increased sharply in Maputo, from 0.444 to 0.524, whereas it increased only slightly, from 0.396 to 0.415, in the country as a whole. Likewise, the higher level of average access to services in urban areas obscures significant within-city disparities. When access to basic infra- structure services is broken down by wealth quintile, each city exhibits major disparities (figure 4.3 and tables 4.2 and 4.3). Once more, these findings are typical of cities in general, and particularly those in develop- ing countries. For example, surveying access to services in the cities of Sub-Saharan Africa, Kessides (2006) finds large, statistically significant gaps between the access of the urban poor versus the urban nonpoor. These gaps are often greater than those between the urban poor and rural residents. In Accra, 68.5 percent of the poorest have access to piped water, but just over half of the poorest in Kampala and only two out of five among the poorest in Maputo have access to piped water. In Kampala, although access for the poorest is still much higher than the national average (56 percent versus 13 percent), congested urban living conditions make alter- natives to piped water, such as wells, less feasible than they would be in many rural settings. UN-Habitat estimates that only 17 percent of the res- idents of Kampala’s informal settlements have access to clean and safe piped water; the rest obtain water from “unhygienic and highly contami- nated sources.”3 Differences in sanitation facilities across quintiles are sharper still. In Maputo and Kampala, private flush toilet facilities are extremely rare for all but the richest quintile. Access to shared toilets tends to decrease over the wealthier quintiles as access to private toilets increases. In Accra, the figures suggest that the vast majority of people in all quintiles have access to either private or shared facilities. But in Kampala, and especially Maputo, access is far from complete. Eighty-four percent of the poorest quintile and over 70 percent of the second quintile of residents of Maputo are living without access to either flush or shared toilets. Figure 4.3 Access to Infrastructure Services by Wealth Quintile: Accra, Kampala, and Maputo a. Trends in urban access to piped water, Accra b. Trends in urban access to piped water, Kampala c. Trends in urban access to piped water, Maputo 100 100 100 80 80 80 percent percent 60 60 percent 60 40 40 40 20 20 20 0 0 0 poorest 2nd 3rd 4th richest poorest 2nd 3rd 4th richest poorest 2nd 3rd 4th richest 1998 92.9 97.4 100.0 98.4 96.8 1995 37.5 60.5 52.6 68.1 67.5 1996 24.8 26.8 51.8 50.5 73.1 2003 82.6 93.0 89.7 91.5 91.7 2006 55.6 70.1 73.7 81.3 96.8 2002 40.4 63.6 52.3 67.5 88.7 d. Trends in urban access to electricity, Accra e. Trends in urban access to electricity, Kampala f. Trends in urban access to electricity, Maputo 100 100 100 80 80 80 percent percent percent 60 60 60 40 40 40 20 20 20 0 0 0 poorest 2nd 3rd 4th richest poorest 2nd 3rd 4th richest poorest 2nd 3rd 4th richest 1998 71.5 97.2 100.0 100.0 100.0 1995 0.0 9.7 74.8 95.8 96.2 1996 10.2 20.5 47.3 60.6 78.8 2003 41.6 91.5 96.3 98.4 100.0 2006 10.1 35.1 68.3 93.8 99.3 2002 9.5 36.6 43.9 62.0 88.0 Sources: Demographic and Health Surveys for Ghana (Accra) and Uganda (Kampala); Living Standards Survey for Mozambique (Maputo). Note: The wealth index places individuals on a continuous scale of relative wealth. It is a composite measure of household living standards, calculated from data on ownership of assets. Assets include type of water access and sanitation facility, and so although one would expect to see a correlation between income or consumption quintile and asset ownership, these measures are not completely independent. 85 86 From Farm to Firm Table 4.2 Access to a Flush Toilet: Accra, Kampala, and Maputo, Selected Years Quintile of wealth index City (country) Poorest 2nd 3rd 4th Richest Accra (Ghana, 2008) 18.7 24.5 34.9 63.1 92.6 (0.0) (0.5) (2.9) (12.0) (55.0) Kampala (Uganda, 2006) 0.7 2.4 2.4 1.4 54.1 (0.0) (0.0) (0.0) (0.02) (5.82) Maputo (Mozambique, 2003) 0.0 0.0 1.8 12.0 75.6 (0.0) (0.0) (0.0) (0.0) (14.47) Source: Demographic and Health Surveys. Table 4.3 Access to a Shared Toilet: Accra, Kampala, and Maputo, Selected Years Quintile of wealth index City (country) Poorest 2nd 3rd 4th Richest Accra (Ghana, 2008) 96.6 93.4 80.9 65.2 19.5 (75.4) (89.4) (91.4) (89.8) (60.7) Kampala (Uganda, 2006) 88.8 83.8 79.1 70.0 20.3 (64.71) (42.54) (31.1) (34.44) (47.0) Maputo (Mozambique, 2003) 15.7 29.4 11.3 7.0 3.2 (11.85) (20.83) (17.27) (13.23) Source: Demographic and Health Surveys. Other sanitation services such as waste collection, drains, and sewers are also lacking. For example, Accra has the capacity to collect only 1,200 of the 1,500–1,800 tons of solid waste generated daily (UN-Habitat 2009), and Maputo’s Municipal Council collects just 30 percent of the waste generated in the city (World Bank 2006). Large swathes of popula- tion live beyond the reach of public water supplies and drainage facilities: Kampala’s main sewer line covers less than 3 percent of the city (Nuwagaba 2004), and an estimated 85 percent of Maputo’s residents live in areas where there is no rainwater drainage system (Maputo City Council 2005). What do the data reveal about trends over time? As the figures sug- gest, although urban infrastructure access generally improved during the period between the two most recent Demographic and Health Surveys, it did not become more equal. For example, access to electricity only mar- ginally improved for the poorest quintiles in Kampala and declined for the poorest in Maputo, with access remaining extremely unequal between top and bottom quintiles in those cities. Accra, which had rela- tively good electricity service for all quintiles in 1998, saw a drop for the poorest by 2003, which represented an increase in inequality. The one Local Level: Intraurban Welfare Disparities and Implications for Development 87 exception to the trend was drinking water in Maputo, where improved access for the poorest meant some narrowing of inequalities between the rich and poor. Unequal access to water and sanitation affects health. Figure 4.4 con- siders one indicator of ill health—stunting. Again, there is a clear urban advantage; on average, children in the capital cities are significantly less likely to be stunted than children nationwide. However, less stunting is found among the wealthiest quintiles nationwide than in the lowest (Maputo), bottom two (Accra), or even bottom three (Kampala) quin- tiles in the capital cities. There is also a considerable spread among wealth quintiles. In Accra, for example, the likelihood of stunting ranges from almost 20 percent among the second and third quintiles to no stunting at all for the top quintile. Ranges are similar in Kampala (almost a quarter of the poorest three quintiles compared with under 5 percent of the top quintile) and Maputo (about 30 percent among the poorest compared with under 10 percent of the wealthiest), although the existence of stunt- ing among the top urban quintiles suggests that, at least in Mozambique and Uganda, even the wealthiest are vulnerable to diseases and other health threats that affect their children’s well-being. Figure 4.4 Stunting: Nationwide versus Capital City 60 50 40 percent 30 20 10 0 poorest 2nd 3rd 4th richest quintile Accra Ghana Kampala Uganda Maputo Mozambique Source: Demographic and Health Surveys (Ghana 2008, Uganda 2006, Mozambique 2003). 88 From Farm to Firm Welfare Differences: Linked to Location Poverty and poor access to services are not evenly distributed throughout these three cities. The differences in consumption and access to services just described reflect spatially based inequalities. Although no small area poverty maps have been created for the three cities in this report, what evidence we do have suggests a strong correlation between poverty and location, with most but certainly not all of the urban poor living in areas defined as slums (see box 4.1). Maxwell et al. (2000) find high welfare disparities in a survey of 16 areas across Greater Accra. Table 4.4 reports the welfare measures for the one middle-income neighborhood in the survey—Dansoman—and for two slum areas—Tema Newtown and Nima. The results demonstrate a strong correlation between location and measures of well-being. The quality of housing and the immediate environment are much better in Dansoman than in Tema Newtown and Nima. Per capita household expenditures in Dansoman are more than three times higher than in Tema Newtown and 2.5 times higher than in Nima, and a much lower proportion of expendi- ture is devoted to food. Finally, a child born in Dansoman is eight times more likely to survive than a child born in Tema Newtown and 10 times more likely to survive than one born in Nima. A 2005 Consultative Citizens’ Report Card Survey for Maputo high- lights how the inequalities in access to services are closely linked to dis- parities between the seven municipal districts surveyed (Maputo City Council 2005). In district one—a high-income district—98 percent of people have modern toilet facilities, whereas in district seven—a low- income neighborhood—40 percent of people have no toilet facility whatsoever, and 56 percent use traditional facilities. In district one, 96 percent of people have piped potable water, whereas in district seven, 95 percent of people collect water from wells. Waste collection coverage is 73 percent in district one and nonexistent in district seven. Similar geographically fine-tuned data do not exist for Kampala. But the combined evidence suggests that most of the city’s poor live on former agricultural and vacant lands in the urban periphery, made up of unplanned, inadequately serviced, low-density, poor-quality housing (Giddings 2009). Nuwagaba (2004) describes most of Kampala’s poor as concentrated in marginal areas of the city and finds that “slums are an indicator of poverty.” Nyakaana, Sengendo, and Lwasa (2005) describe a walk through a typical Kampala slum: houses are con- structed in road reserves, children play in front of leaking latrines, and Local Level: Intraurban Welfare Disparities and Implications for Development 89 Box 4.1 Poverty and Slums Slums are defined as settlements in which more than half of the households lack at least one of the following: improved water, improved sanitation, durable hous- ing, adequate living space, and security of tenure (UN-Habitat 2002). Urban poor households are found not only in slums, and not all those living in slums are poor. However, the strong correlation between being poor and living in a slum leads to the spatial concentration of poverty in many developing country cities. Relative to other regions, the urban poor in Africa are subjected to particularly severe forms of deprivation. According to UN-Habitat estimations, nearly two- thirds (62 percent) of city dwellers in Sub-Saharan Africa live in a slum—the high- est proportion of any region in the world. Moreover, an examination of conditions across regions suggests that slum dwellers in Sub-Saharan Africa experience a higher level of multiple shelter deprivations than slum dwellers in other regions (UN-Habitat 2008). Slum settlements are closely associated with insecurity because most tenants lack written contracts or any legal mechanism to protect their tenure rights. Unplanned and unsuitable land use heightens vulnerability. Inhabitants of slums are also particularly vulnerable to poverty traps. They tend to experience greater geographic distance from or inadequate and expensive transportation to more productive jobs. Residents in a low-income settlement outside Dar es Salaam, for example, were found to be spending between 10 and 30 percent of their income on transport. Spatial concentrations of crime and violence add to the stigmatiza- tion of slums, creating additional hurdles in access to jobs or credit (Kilroy 2007). Slums are often tacitly accepted by local governments, which find themselves unable to provide basic services for large segments of their city’s de facto resi- dents. Not only do slum dwellers receive fewer public services, but they also have to pay more for those they purchase in the private market. Those in the high- income neighborhoods of Accra, for example, have access to a constant water supply at the official rate, whereas those in middle-income neighborhoods must supplement publicly supplied water with water from private vendors at about four times the official rate, and those in slums must rely entirely on these private suppliers. The cost of water from vendors can amount to 10 percent of the monthly income of a low-income household, which in turn limits the amount of water the household uses for drinking, cooking, and hygiene. The result is higher rates of disease, greater loss in workdays, and even higher outlays of expenditure for medical attention (World Bank 2002). 90 From Farm to Firm Table 4.4 Welfare Measures in Subareas of Greater Accra Housing, Total per capita Proportion assets, and household Proportion of of live-born environment expenditure expenditure children who Area indexa (1,000 cedis) spent on food died Dansoman 1.75 1,757 39.0 1.6 Nima –0.46 699 56.8 16.3 Tema Newtown –1.42 547 61.9 13.0 Source: Maxwell et al. 2000. a. An index of quality of housing, assets, and environmental quality was created based on four assets: refrigerator, electric stove, tape deck, and television set. Factor analysis was then used to derive one factor that would summarize variables related to the quality of housing (construction material for roof, walls, and floor), the “possession” index, the source of drinking water, and the availability of sanitary facilities and garbage disposal. The principal components factor extraction method was used. Loadings smaller than 0.5 were excluded from the initial model. The final model had five variables (floor, walls, water source, possession index, and sanitary facilities), all with factor loadings greater than 0.53, and it explained 46 percent of the variance. the overflow contaminates drinking water sources, making residents particularly vulnerable to cholera. Understanding Spatial Inequalities This section examines the causes of the spatially based welfare disparities just documented here. The high inequalities in many cities in both the developed and the developing world suggest that inequality is in part a function of urban growth. But it is by no means inevitable: some Asian cities such as Beijing and Jakarta have managed to grow in size and eco- nomic strength while maintaining relatively low levels of inequality. In Malaysia, urban income inequalities have even decreased steadily since the late 1970s as urban populations have grown (UN-Habitat 2008). In this section, we therefore look beyond the sheer fact of rising pop- ulations to uncover the institutional reasons for urban inequality. As described, inequalities are rooted in the colonial past, but also are rein- forced by more recent institutional constraints and policy failures. In par- ticular, we consider the failure to adequately address the complex and confused systems of land tenure that have enabled an elite minority to monopolize access to land and housing and forced a poor majority to rely on insecure and inadequate arrangements. We also address the shortcom- ings of urban governance, which have resulted in ever-larger populations congregating in areas of cities that lack even the most basic access to energy, sanitation, or transportation infrastructures. Local Level: Intraurban Welfare Disparities and Implications for Development 91 The story of urban inequality is not uniformly gloomy, however. Chapter 11, which presents our summary and conclusions, notes recent initiatives to address unjust institutions, such as by simplifying systems of land tenure, and to improve service delivery to the poor, such as through programs to upgrade slums or reform local government. The Colonial Legacy Colonial legacies have contributed to spatially based welfare disparities in contemporary African cities. The populations of colonial Accra, Kampala, and Maputo were not large by present-day standards, but the colonial era created systemic inequalities. For example, the land use and settlement patterns of the colonial city reflected socioeconomic and racial divisions. In most African cities, the “standard of infrastructure in the neighbor- hoods inhabited by Europeans was many times higher than in African areas. . . . Europeans tended to live in bungalows or villas in large plots along clean, shady streets. . . . Africans tended to live . . . along unplanned paths, enjoying few, if any, sanitary services” (Stren and Halfani 2001). In colonial Accra, most commercial activity and planned development were located near the port; African residents lived and worked in overcrowded areas outside of this central business area (Grant and Yankson 2003). The divisions in Kampala were written into law and starker still. Colonial Kampala was divided into African, Asian, and European residen- tial areas. Most of the city’s African population lived in the Kibuga, an area subject to the authority of the king of Buganda, whereas Europeans and Asians lived in Kampala proper and were served by the municipal government. The separate systems of administration created de facto seg- regation between the Kibuga and Kampala in service delivery (Nilsson 2006). Roads, mosquito control, drainage, water supply, and sanitation were far inferior in the Kibuga compared with that in municipal Kampala (Nilsson 2006; Bryceson 2008). Because the municipality financed its services through taxes and fees, the divide had socioeconomic and racial ramifications. According to Elkan and van Zwanenberg (1975), “Life in the municipality (Kampala) was healthy but expensive, whereas in the Kibuga . . . it was insanitary and cheap. . . . Given the racial distribution of income, this led naturally to a racial distribution of residence.” Similarly, Maputo (then known as Lourenço Marques) was ruled under a trifurcated legal regime. The city was organized into separate areas for the African indígenas, assimilados (a small group of “Europeanized” Africans), and Europeans, with differentiated civil rights granted on the 92 From Farm to Firm basis of both race and location. Confined to the periphery of the city, indígenas were barred from building houses from permanent materials. As Jenkins (2006) describes, this practice gave rise to a dualistic settlement pattern: “a gracious Southern European-style planned inner ‘cement city’ surrounded by the ‘caniço’” (in reference to the reeds from which many of the buildings were constructed). The monopolis- tic pattern of landownership indicated the extent of socioeconomic inequality: in 1965, 75 percent of private land was in the hands of just 11 large-scale landowners (Jenkins 2000). By placing some areas under the jurisdiction of the traditional author- ities while imposing European legal norms in others, colonial rulers com- pounded complex systems of tenure. In Accra, as elsewhere in Ghana, customary land tenure systems—whereby ownership was corporate and vested in a “stool,” represented by the chiefs, or in a family—operated in parallel with the British conveyancing system (Gough and Yankson 2000). In Kampala, a 1900 agreement granted mailo land—a system of freehold land tenure—in perpetuity to the Bagandan majority. The origi- nal inhabitants of mailo land became tenants overnight, paying annual dues to the landowners. Although the colonial regime found these sys- tems convenient, enabling indirect control of large territories, they cre- ated a complex legacy, contributing to inelastic land supply and cementing divisions between a landed elite and a landless city. The chal- lenge of overcoming colonial legacies was made more difficult by the ambivalent attitudes of some newly independent governments toward the process of urbanization (see box 4.2). Failure to Strengthen Systems for Land Delivery The colonial period, then, resulted in the codification of a dual system of land tenure, leaving a complex legacy that has become only more confusing as cities have grown in physical size and population. Post- independence governments have generally been slow to respond to the challenging task of simplifying land tenure arrangements. In Accra, for example, the dual system of customary and formal land ownership has resulted in “persistent anarchy in the land market” (Gyimah-Boadi 2004). The city is rife with “grey areas” where two or more stools (chief- taincies) contest ownership (World Bank 2007b). With dispute resolu- tion taking up to a decade, land acquisition through formal means is perceived to be a daunting task (Durand-Lasserve et al. 2007). Large public land holdings and “huge deficits in the policing, regulation, Local Level: Intraurban Welfare Disparities and Implications for Development 93 Box 4.2 Postindependence Attitudes toward Urban Development Since independence, national governments have displayed various attitudes toward the process of urbanization. In Ghana, starting with the Nkrumah govern- ment (1957–66) the country’s leadership promoted Accra as the central growth pole of the national economy, improving the city’s transportation network and making major infrastructure investments such as in the deepwater harbor at Tema (Grant and Yankson 2003). By contrast, successive governments of independent Uganda tended to con- ceive of urban areas as “places of opportunity” that required little policy interven- tion, focusing instead on development in rural areas (Nuwagaba 2004). Under the rule of Idi Amin (1971–78), Kampala’s economic health was subverted in pursuit of brutal ideological goals: tens of thousands of Ugandans were executed in the cap- ital, and an estimated 80,000 ethnic Asian residents nationwide, owning as much as 90 percent of all businesses, fled the country. Only more recently (since 1986) has the government aimed its economic development policies at supporting commercial and industrial development in the Kampala region (Giddings 2009). In Mozambique, the official stance of the ruling FRELIMO (Liberation Front of Mozambique) party toward urban development in the years following the coun- try’s independence was even more hostile. After independence, 150,000–200,000 of Maputo’s Portuguese and assimilados residents, fearing persecution, fled. The government aimed its economic development policies at rural areas. In 1985, the state imposed Operation Production, whose expressed goal was to remove sur- plus urban workers (described as “parasitic”) to the countryside. Under this policy, as many as 50,000 people were relocated from Maputo, and it is believed that many perished. It is only since adoption of the 1990 constitution that Mozam- bique had been a free-market economy, and Maputo is increasingly seen as an essential “engine of development” (Jenkins 2006). Today, a “rural bias” in policy making persists in some countries. According to a 2008 United Nations Department of Economic and Social Affairs survey, 83 percent of African governments implement policies to reduce rural-to- urban migration, and 78 percent intervene to reduce migration to large urban agglomerations (UNDESA 2008). Although little panel data-based research is found in Africa, research from other areas suggests that restrictions on migra- tion, which may consist of denial of services or slum evictions, result in main- taining surplus labor in rural areas and low agglomeration economies in urban areas, ultimately slowing the development of both rural and urban areas. 94 From Farm to Firm acquisition and use” of such land further stifle the market (Gyimah- Boadi 2004). The institutional impediments to efficient planning and the functioning of a land and housing market in Accra include over- lapping arrangements between different agencies, poor land manage- ment, cumbersome processes of titling, conveyancing and deeds registration, and inadequate mapping (World Bank 2002). In Kampala, the development of an efficient land market has been hampered by the colonial legacy of mailo land. About 60–70 percent of the land in Kampala is owned under the mailo freehold system (World Bank 2008a, annex 1). This kind of tenure permits separating the owner- ship of land from the ownership of developments on land (World Bank 2009). Because most of the title-holding landowners do not have enough capital to develop their land, they rent or allow squatters to occupy it, usually without providing the most basic of services. Observers claim that the mailo system has resulted in a “land impasse” in Kampala (Augustus Nuwagaba, quoted by Rulekere 2006). Efforts to reform the system have only compounded the problems. For example, the 1995 constitution stated that tenants were to be compensated for the sale or development of mailo land, but this stipulation has had the effect of depressing demand, because any purchaser could face multiple compensation claims (Rulekere 2006). Leasehold land is in short supply; Giddings (2009) esti- mates that the acquisition of a public plot through leasehold arrangement can take up to five years. In postcolonial Mozambique, FRELIMO (Liberation Front of Mozambique, which was the ruling party until 2009 and is currently the majority political party in the government) nationalized land and rented or abandoned housing. From 1975 to 1987, the state dominated all for- mal activity in urban land and housing supply. After 1987, market forces increasingly prevailed, although the 1997 Land Law and 2004 constitu- tion both assert that all land belongs to the state (Jenkins 2001; Malauene et al. 2005). Although the state remains in de jure control of land alloca- tion, it has not been able to keep pace with the demand for land and housing, particularly among Maputo’s low-income groups. Cumbersome bureaucratic processes act as a disincentive for formal registration. Indeed, Malauene et al. (2005) identify more than 27 steps associated with registering land, 9 steps for requesting plans, and a further 28 steps for registering a building. And although formal access to land is nominally free, in practice, multiple fees must be paid to the Maputo Municipal Council. Among the most pertinent problems, observers find that the current legal framework, enshrined in the 1997 Land Law, is ill-suited Local Level: Intraurban Welfare Disparities and Implications for Development 95 for urban circumstances; the institutional and technical capacity is weak; land management instruments are lacking; and low-income groups tend to be unaware of their rights4 (Jenkins 2001; Malauene et al. 2005; World Bank 2008b).Together, these factors add to the scope for arbitrary deci- sion making and bribe taking among officials, which only act as further barriers to entry for the poor. In each of the three cities, inadequate legal frameworks, weak dispute settlements, low institutional and technical capacities, lengthy registration processes, poor land management, and corruption have combined to cre- ate tight formal housing markets. In Accra, because of legal ambiguities such as lack of documentation and poor boundary definition, private par- ties are reluctant to invest in the existing housing stock. As a result, Accra qualifies as a “superstar city,” in which a highly inelastic housing supply is pricing lower- and even middle-income groups out of the market (Buckley and Mathema 2007). One report calculates that the backlog of demand for formal housing units in Accra has reached 500,000, with an additional 14,000–16,000 units required each year (referenced by World Bank 2007b). In Kampala, the unwillingness of owners to sell land has led to “skyrocketing” prices (World Bank 2008a). Current land prices are from two to four times what they were in 2002, and the city has a deficit of housing units estimated at 100,000 (Giddings 2009). To take advan- tage of rising prices, the private sector has mainly engaged in high-value construction, with the hope of attracting rent from foreigners; mean- while, housing for middle- and low-income groups remains in short sup- ply (Nuwagaba 2004). Likewise, surveying the data for Maputo, Jenkins (2000) concludes that formal land and housing “of even a most basic stan- dard” is unobtainable for the vast majority of low-income city dwellers. National governments have not been blind to the urgent need to address land problems in their capital cities. In Accra, a Land Administration Pilot Project, initiated under the umbrella of a slum upgrading program in 1996, aimed to improve mapping capabilities and land titling administration. Other programs were directed at strengthen- ing records management and establishing a small unit for the digital pro- duction of registry maps. However, a broader attempt to resolve the underlying issues of tenure has not been forthcoming (World Bank 2002). Uganda has experienced similar difficulties. As of 2009, the government of Uganda’s comprehensive land policy review was in its fourth draft. The policy aims to “rationalize and simplify the complex tenure regimes.” To this end, it recommends (1) that the mailo system of ownership be elim- inated; (2) that freehold land, especially in urban areas, be converted 96 From Farm to Firm into leasehold land; and (3) that customary tenure be formalized.5 However, the current draft has generated opposition from landlords and the Buganda Kingdom (the largest mailo landlord); they view it as an attempt by the government to take away their land. According to Giddings (2009), “How fast the policy will move towards adoption, or how much compromise will be necessary given the resistance of the mailo owners, is therefore uncertain.” Constraints to Effective Urban Governance The approach to governance in many African cities has had distinct phases. The minimalist state of the colonial period was usually replaced by a more bureaucratic approach in the post-independence era, and the structural adjustment programs of the 1980s sought to make unwieldy public services more nimble (Gyimah-Boadi 2004). By the early 1990s, many African cities were facing a “crisis of governance,” manifested in the shortcomings of service delivery and severe spatially based inequalities (Stren and Halfani 2001). Decentralization programs aimed at devolving governmental functions and finances from the national to the district or municipal level were an attempt to alleviate this crisis by making govern- ments more responsive to local needs. Decentralization has been an official goal of all three of the countries featured here.6 But administrative weaknesses have constrained the actual extent of the decentralization. A recent urban profile of Accra points to the dual allegiance of decentralized departments, nonconnectivity of departments, lack of transparency, overcentralization of administration, and functional duplication of public and parastatal agencies as just some of the obstacles to effective local governance at the city level (UN-Habitat 2009). The Maputo Municipal Council is constrained by a skills shortage (in 2005 only 1.5 percent of staff had a higher education and 7 percent had technical training), weak operational supervision, inadequate infor- mation systems, and poor budget management tools (World Bank 2006). Many of the obstacles to effective local governance in Kampala are tied up in the complex land tenure system. For example, the arduous bureau- cratic procedures required to purchase a public plot of land only add to the incentives for corruption (see box 4.3). The ability of local urban authorities to implement decisions has been further limited by the willingness of national governments to transfer the financial resources needed to match devolved responsibil- ities (United Cities and Local Governments 2007). Contrary to notions of an “urban bias,” central governments tend to favor rural areas in the Local Level: Intraurban Welfare Disparities and Implications for Development 97 Box 4.3 Corruption in Kampala Councilors have demanded immediate investigations into Kampala City Council senior staff who have constantly failed to execute their duties, leading to con- struction of several illegal structures in the city. Makerere University Councilor, Mr Bernard Luyiga said the engineering and planning departments should be investigated over failure to realize the city plan. “We have several properties con- structed in the road reserves while others have altered the approved plans. The two departments are primarily responsible and if there is any political influence, it should be exposed,” he said. Lubaga Division representative Mr Zachel Mberazze, said several roads and access lanes in the central business district have been blocked by illegal struc- tures which has tainted council’s image and worsened traffic congestion in the city. . . . The developments come after a Ministry of Works, Transport and Commu- nication report recommended that the Inspector General of Government investi- gates the continuous sale of Kisenyi lane, a public utility to individuals. . . . The report signed by the MoW (Ministry of Works) Permanent Secretary Charles Muganzi says council senior staffs were responsible for the increasing city mess. “The ministry wants in-depth investigations conducted to establish the individuals who took part in the evil act of selling off three plots of land that were set aside as road reserve in Kisenyi parish, Kampala Central Division,” the report, a copy of which Daily Monitor has obtained, reads in part. The report indicates that plots 215,239 and 203 on Block 12 Kafumbe Mukasa Road oppo- site Nakivubo Stadium, which were reserved as roads accessing the commercial area, were encroached on in November 2006. “It has been established that these plots had been planned as accesses to the planned commercial area. And all constructions on the said plot are illegal and being carried out in the wrong place,” Mr Mungazi said. . . . The MoW report also recommends disciplinary action against the KCC officials. Source: Reprinted from Robert Mwanje, “Uganda: Councilors Want KCC Officials Punished Over Illegal Structure,” Daily Monitor, October 21, 2009, http://allafrica.com/stories/200910210139.html. distribution of funds. Ugandan policy makers’ characterization of urban areas as “places of opportunity” has historically affected the design of policy and program interventions in favor of rural areas. As for Ghana, the World Bank (2007a) has noted that its urban District Assemblies are disfavored in the transfer of central funds. Also according 98 From Farm to Firm to the World Bank (2006), municipalities in Mozambique receive only 1 percent of the national budget. Meanwhile, cities’ own revenue-generating capacities have been stymied by a high degree of informality—whether of land, property, or enterprise—as well as by poor administrative capacity (Durand-Lasserve et al. 2007). In 2001, the Accra Metropolitan Assembly estimated that it was able to benefit from less than one-third of its fiscal base. Only 5 per- cent of properties in Maputo are taxed, and the Maputo Municipal Council lacks an updated cadastre and other tools to increase property tax revenues. In addition, local fees and user charges are outdated and dif- ficult to collect, and the laws governing municipal fiscal autonomy are very restrictive (World Bank 2006). Likewise, in Kampala outdated valu- ations, frequent property sales and exchanges, and inadequate title regis- tration have limited the extent to which local authorities can levy property taxes (Fjeldstad 2006). National and local authorities have recognized the urgent need to strengthen urban governance in order to deliver better and more equi- table services to city dwellers. Accra has been implementing slum upgrad- ing programs over the last two decades, and most recently the city turned its attention to revamping its collapsed mass transport system (UN- Habitat 2009). Slum upgrading has resulted in improved road paving, water, sanitation, solid waste removal, and street lighting in selected parts of the city. The first-generation projects were top-down initiatives designed and fully funded by the central government. A more recent environmental sanitation project adopted a much more participatory approach and engaged local government in both funding and imple- mentation (World Bank 2002). These programs have led to markedly improved welfare indicators in the areas in which they have been implemented, and they also have stimulated local economic develop- ment (box 4.4). However, they have not yet been replicated on a large enough scale to lead to improvements in citywide welfare indicators (World Bank 2008a). In Maputo, decentralization has become more meaningful by involving submunicipal officials and neighborhood structures in participatory plan- ning and service assessments. For example, the Consultative Citizens’ Report Card Survey described earlier has fed into a participatory process to formulate a 10-year development program for the city called ProMaputo. The first phase of ProMaputo recognizes that the municipal government has an extremely limited capacity to raise its own revenues and to provide its citizens with quality services. Therefore, the initial Local Level: Intraurban Welfare Disparities and Implications for Development 99 Box 4.4 Slum Upgrading in Accra In 1985, with World Bank support, the government of Ghana implemented a pilot infrastructure upgrading project in the East Maamobi community of Accra. The government’s reaction upon completion of the scheme was favorable, and more schemes were initiated in 1988 under the Priority Works Project, in 1996 under the Urban 2 Project, and in 1997 under the Urban Environmental Sanita- tion Project (UESP). The UESP was implemented in seven urban communities in Ghana, including three in Accra—Sukura, Old Teshie, and West Maamobi. The slum upgrading com- ponent of the UESP set cost-per-hectare limits and promoted community partic- ipation throughout the identification, planning, design, and implementation phases. Efforts were also made to engage local governments: they were required to contribute 11 percent of the capital costs and were responsible for managing implementation of the UESP. Although they have not addressed social infrastructure such as health care clinics and schools, the upgrading schemes have improved environmental, sani- tation, and economic conditions in the selected communities: • Improved solid waste collection has reduced the accumulation of garbage in areas where children play and parents cook and clean their dishes and utensils. • Improved drainage has led to reduced flooding in most areas, thereby slowing the spread of waterborne diseases. • Paved streets and improved access to communities have resulted in a prolifera- tion of small businesses and kiosks in areas that once were not economically viable. • Upgrading has motivated individual households to invest more in improving their shelter structures. • The introduction of street lighting in some areas has reduced the incidence of crime. The benefits of upgrading extend beyond the specific areas upgraded. For example, people from the areas around Maamobi use the communal toilets, waste containers, and running water facilities in the upgraded areas. Because of the success of these programs, the government of Ghana and the World Bank are considering preparation of an urban upgrading project for the Greater Accra Met- ropolitan Area, which could set the stage for a national upgrading program (World Bank 2002). 100 From Farm to Firm focus is on improving governance, financial stability, and institutional capacity in order to provide a platform for subsequent sustainable improvements to service delivery (World Bank 2006). In Kampala, after accusations of corruption and ineffectiveness were made against the Kampala City Council, Parliament passed a bill in November 2010 that provided for the administration of Kampala by the central government, although proposals to expand the official city limits and abolish the council were dropped. The new Kampala Capital City Authority to “oversee the delivery of quality services to the population” and the Metropolitan Physical Planning Authority to promote a new structure plan and to “ensure that land use in the City and metropolitan area follow designated plans, irrespective of the tenure of land.” The Buganda Kingdom and some opposition groups had opposed the new law (Mpagi and Ssenkabirwa 2009, Namutebi and Bekunda 2010). Conclusion Cities in Sub-Saharan Africa are growing at an extraordinarily fast pace, and they continue to attract migrants to the economic opportunities and greater amenities they offer. In the three capitals discussed in this part of the report, poverty levels are lower and the overall living standards, as measured by access to social and infrastructure services, are better than those in rural areas, although the degree of disparity varies considerably. But urbanization also has its “dark side” (Spence, Annez, and Buckley 2009), manifested in the prevalence of slums and the absence of decent shelter or basic services for those who dwell in them. Colonial histories, a complex mix of traditional, colonial, and postcolonial land tenure sys- tems, the resultant dysfunctional land markets, and weak governance all have a role to play in the stark spatial inequalities found in Accra, Kampala, Maputo, and many other African cities. As noted in WDR 2009, historical evidence indicates that such inequalities will persist longer than the rural-urban welfare gaps, dimin- ishing only at the more advanced stages of development. But this find- ing does not mean that national and urban governments should not take action to alleviate urban inequalities. Failure to invest sufficiently in infrastructure in rapidly growing cities results in congestion, crime, pollution, and social tension, which, in turn, undermine the potential benefits of agglomeration. The policies needed to address these spatial inequalities are discussed briefly in chapter 11, this report’s summary and conclusions. Here, it is Local Level: Intraurban Welfare Disparities and Implications for Development 101 worth noting that not all policies to address urban inequalities will target the city. WDR 2009 emphasizes the importance of framing the policy debate on cities in the context of the overall rural-urban transformation as well as regional economic growth. For example, introducing better services in rural areas will reduce “push factors” for people whose primary reason for moving to cities is to access better amenities. And policies to promote regional integration would enable larger agglomeration economies and opportunities for job creation (box 4.5). Indeed, for anyone designing policies it is better to think about costs and benefits along the rural-urban continuum rather than in distinctly rural or urban spaces. For large cities in particular, the success of all these policies rests on improving coordination between the national, municipal, and local levels of government. Responsibilities are often transferred down to local governments that lack the resources or the capacity to address the problems of a metropolitan region. It is hardly surprising, Box 4.5 Promoting Agglomeration through Regional Integration Recent work by Collier and Page (UNIDO 2009) suggests that because of Africa’s political fragmentation and many landlocked countries, urban welfare and eco- nomic growth would benefit greatly from policies that promote regional integra- tion. They argue, consistent with WDR 2009, that bigger cities have the potential to generate greater economies of scale. Simulating a merger of 10 African countries, Collier and Venables (2008), cited in UNIDO (2009), conclude that if each country had one city of 10 million, after the merger 6 of the 10 cities would increase in size and thus in productivity. Four cities would shrink, although, depending on the extent of integration, their losses could be compensated through income redistribution. Collier and Page (UNIDO 2009) conclude that “small low-income countries are at a massive disadvantage with regard to industrialization,” but they also acknowl- edge that the current trend away from vertical integration of industries may make city size less critical. However, other factors, such as an adequate concen- tration of good-quality infrastructure in a limited physical space, thereby creat- ing what Collier refers to as “islands of protected governance,” are critical for encouraging agglomeration economies even in a “sea of corruption.” Sources: UNIDO (2009) and Collier’s presentation of the UNIDO report at the World Bank, February 23, 2010. 102 From Farm to Firm then, that the one city that concurrently implemented spatially neutral institutions, connective infrastructure, and targeted interventions with success was Singapore, a city-state. And even if lines of authority and responsibility between various levels of government are clarified, national governments must take on more responsibility for financing the urban transformation (Spence, Annez, and Buckley 2009). The task of ensuring that the bright lights of the city are not just an illusion is not purely a technical question but one of political will and public pressure. Notes 1. Accra Metropolitan Assembly, http://ama.ghanadistricts.gov.gh. 2. The population estimate for Maputo in 2009 was 1.1 million, but because Maputo and Matola are adjacent municipalities, their combined population estimate of over 1.8 million provides a better measure of the size of the met- ropolitan area. 3. UN-Habitat, http://www.unhabitat.org/content.asp?cid=3294&catid=460& typeid=13&subMenuId=0. 4. The World Bank (2008b) finds that only 15 percent of urban households have any knowledge of how to obtain a land title under the current law. 5. Republic of Uganda, Ministry of Lands, Housing and Urban Development, “Drafting the National Land Policy, Emerging Issues for Discussion,” May 2007; cited by Giddings (2009). 6. Ghana’s Local Government Act of 1993 charged District Assemblies with ensuring the development of basic infrastructure, the provision of sanitary conditions and education services, and protection of the environment. Since passage of Uganda’s 1997 Local Government Act, the Kampala City Council has had responsibility for ensuring economic development and service deliv- ery in the capital, although the recently passed Kampala Capital City Bill has changed the allocation of responsibilities. Also in 1997, Mozambique passed a municipal legislation package that established 33 urban local governments with limited but autonomous powers and democratically elected mayors and assemblies. In accordance with the legislation, Maputo’s Municipal Assembly is responsible for land use, basic water and sanitation services, urban roads, solid waste management, and the municipal environment. References Adjasi, C. K. D., and K. A. Osei. 2007. “Poverty Profile and Correlates of Poverty in Ghana.” International Journal of Social Economics 34 (7): 449–71. Local Level: Intraurban Welfare Disparities and Implications for Development 103 Arndt, C., R. James, and K. Simler. 2006. “Has Economic Growth in Mozambique Been Pro-Poor?” Journal of African Economies 15 (4): 571–602. Aryeetey, E., and A. McKay. 2007. “Ghana: The Challenge of Sustained Poverty Reduction.” In Delivering on the Promise of Pro-Poor Growth, ed. T. Besley and L. Cord. Washington, DC: World Bank and Palgrave Macmillan. Bryceson, D. 2008. “Creole and Tribal Designs: Dar es Salaam and Kampala as Ethnic Cities in Coalescing Nation States.” Crisis States Working Papers Series 2, London School of Economics, U.K. Buckley, R., and A. Mathema. 2007. “Is Accra a Superstar City?” Policy Research Working Paper 4453, World Bank, Washington, DC. Collier, P., and A. J. Venables. 2008. “Trade and Economic Performance: Does Africa’s Fragmentation Matter?” Paper prepared for the Annual World Bank Conference on Development Economics, Cape Town, South Africa, June 9–11, 2008. Durand-Lasserve, A., E. Fernandes, G. Payne, and C. Rakodi. 2007. “Social and Economic Impacts of Land Titling Programs in Urban and Peri-Urban Areas: A Review of the Literature.” Study funded by the government of Norway fol- lowing the World Urban Forum in Canada, June 2006. Elkan, W., and R. van Zwanenberg. 1975. “How People Came to Live in Cities in Colonialism in Africa.” In The Economics of Colonialism, Vol. 4, ed. P. Duignan and L. H. Gann, 655–72. Cambridge, U.K.: Cambridge University Press. Fjeldstad, O-H. 2006. “Local Revenue Mobilization in Urban Settings in Africa.” CMI Working Papers, Chr. Michelsen Institute, Bergen, Norway. http://www.cmi.no/publications/file/2338-local-revenue-mobilization-in- urban-settings-in.pdf. Ghana Statistical Service. 2007. “Pattern and Trends of Poverty in Ghana, 1991–2006.” Accra. Giddings, S. 2009. “The Land Market in Kampala and Its Effect on Settlement Patterns.” International Housing Coalition, Washington, DC. Gough, K., and P. Yankson. 2000. “Land Markets in African Cities: The Case of Peri-Urban Accra, Ghana.” Urban Studies 37 (13): 2485–500. Grant, R., and P. Yankson. 2003. “City Profile: Accra.” Cities 20: 65–74. Grest, J. 2006. “Urban Governance, State Capacity and the Informalization of Urban Management in Maputo: Some Thoughts on Poverty, Informalization and Survival in the City.” Notes for “GDRI: Governing Cities in Africa; Law, Local Institutions and Urban Identities since 1945. Regional Workshop: South African Cities,” Stellenbosch, South Africa, November 16–17. Gyimah-Boadi, E. 2004. “The Search for a Developmental Public Service in Ghana: Challenges and Prospects.” Paper presented at the Public Services Commission 7th Annual Lecture, Accra, Ghana, May 19. 104 From Farm to Firm Jenkins, P. 2000. “Urban Management, Urban Poverty and Urban Governance: Planning and Land Management in Maputo.” Environment and Urbanization 12 (1): 137–52. ———. 2001. “Strengthening Access to Land for Housing for the Poor in Maputo, Mozambique.” International Journal of Urban and Regional Research 25 (3): 629–48. ———. 2006. “Image of the City in Mozambique: Civilization, Parasite, Engine of Growth or Place of Opportunity?” In African Urban Economies: Viability, Vitality or Vitiation? ed. D. Bryceson and D. Pots. New York: Palgrave Macmillan. Kessides, C. 2006. The Urban Transition in Sub-Saharan Africa: Implications for Economic Growth and Poverty Reduction. Washington, DC: Cities Alliance. Kilroy, A. 2007. “Intra-Urban Spatial Inequality: Cities as ‘Urban Regions.’” Background paper, World Development Report 2009, World Bank, Washington, DC. Malauene, D., A. Chilundo, B. Cau, and M. Mubai. 2005. “Land Registration in Maputo and Matola Cities, Mozambique.” International Institute for Environment and Development, London. Maputo City Council. 2005. “Municipal Score Card Survey Urban Services Provision.” Maputo, Uganda. Maxwell, D., C. Levin, M. Armar-Klemesu, M. Ruel, S. Morris, and C. Ahiadeke. 2000. “Urban Livelihoods and Food and Nutrition Security in Greater Accra, Ghana.” Research Report 112, International Food Policy Research Institute, Washington, DC. Mpagi, C. M., and A. Ssenkabirwa. 2009. “Cabinet Stuck Over Kampala Expansion.” Daily Monitor (Kampala), July 13. http://www.monitor.co.ug/. Namutebi, J., and Bekunda, C. 2010. “House Passes Kampala Takeover Bill.” New Vision Online, 3 Nov. 2010. http://www.newvision.co.ug/D/8/13/736999 Nilsson, D. 2006. “A Heritage of Unsustainability: Reviewing the Origin of the Large Scale Water and Sanitation System in Kampala, Uganda.” Environment and Urbanization 18 (2): 369–85. Nuwagaba, A. 2004. “Understanding the Causes of Increased Urban Poverty in Uganda: The Case of Kampala and Mbale.” Background paper, World Bank Uganda Poverty Assessment, World Bank, Washington, DC. Nyakaana, J. B., H. Sengendo, and S. Lwasa. 2005. “Population, Urban Development and the Environment in Uganda, the Case of Kampala City and Its Environs.” Makerere University, Kampala, Uganda. Rulekere, G. 2006. “Land Tenure Policies Complicating Development in Kampala.” April 24. http://www.ugpulse.com. Local Level: Intraurban Welfare Disparities and Implications for Development 105 Spence, M., P. Annez, and R. Buckley, eds. 2009. Urbanization and Growth. Commission on Growth and Development. Washington, DC: World Bank. Stren, R., and M. Halfani. 2001. “The Cities of Sub-Saharan Africa: From Dependency to Marginality.” In Handbook of Urban Studies, ed. R. Paddison, 466–85. London: Sage Publications. Tuffour, J. A., and J. Apallo. 2005. “Inequality and Social Welfare in Ghana in the 1990s.” Working paper 2005-01, Department of Economics, St. Francis Xavier University, Nova Scotia, Canada. UNDESA (United Nations Department of Economic and Social Affairs). 2008. World Population Policies 2007. New York: UNDESA. http://www.un.org/esa/ population/publications/wpp2007/Publicationindex.htm. UN-Habitat. 2002. Expert Group Meeting on Urban Indicators, Secure Tenure, Slums and Global Sample of Cities. Nairobi: United Nations Human Settlements Programme. Urban Secretariat and Shelter Branch in close collaboration with United Nations Statistic Division and the Cities Alliances. http://www.citiesalliance.org/ca/sites/citiesalliance.org/files/ expert-group-meeting-urban-indicators%5B1%5D.pdf. ———. 2008. State of the World’s Cities 2008/9: Harmonious Cities. Nairobi: United Nations Human Settlements Programme. ———. 2009. Ghana: Accra Urban Profile. Nairobi: United Nations Human Settlements Programme. UNIDO (United Nations Industrial Development Organization). 2009. Industrial Development Report 2009. Breaking In and Moving Up: New Industrial Challenges for the Bottom Billion and the Middle-Income Countries. Vienna: UNIDO. United Cities and Local Governments. 2007. Decentralization and Local Democracy in the World. First Global Report. Barcelona: United Cities and Local Governments. United Nations. 2010. World Urbanization Prospects: The 2009 Revision. Department of Economic and Social Affairs, Population Division. New York: United Nations. World Bank. 2002. Upgrading Low Income Settlements, Country Assessment Report Ghana. Washington, DC: World Bank. ———. 2006. “Program Appraisal Document on a Proposed Credit in the Amount of SDR 20 Million (USD 30 Million Equivalent) to the Republic of Mozambique for ProMaputo: The Maputo Municipal Development Program.” Report 37530-MZ, World Bank, Washington, DC. ———. 2007a. “Ghana Meeting the Challenge of Accelerated and Shared Growth: Country Economic Memorandum.” Report 40934-GH, World Bank, Washington, DC. 106 From Farm to Firm ———. 2007b. “Republic of Ghana Urban Development and Economic Growth.” World Bank, Washington, DC. ———. 2008a. “Africa’s Urbanization for Development: Understanding Africa’s Urban Challenges and Opportunities.” Working Paper 45298, World Bank, Washington, DC. ———. 2008b. “Beating the Odds: Sustaining Inclusion in a Growing Economy. A Mozambique Poverty, Gender and Social Assessment.” Report 40048- MZ, World Bank, Washington, DC. ———. 2008c. World Development Report 2009: Reshaping Economic Geography. Washington, DC: World Bank. ———. 2009. “Uganda: Legal and Judicial Sector Study Report. Legal Vice Presidency.” Report 49701, World Bank, Washington, DC. PA R T I I Managing the Rural-Urban Transformation: Examples from South Asia CHAPTER 5 Pathways Out of Poverty: Managing the Rural-Urban Transformation in South Asia Rural areas are home to three-quarters of the poor in developing countries, and most of the rural poor are directly or indirectly depend- ent on agriculture for their livelihood.1 Lifting this mass of people out of poverty will entail two interrelated processes: (1) shifting employ- ment from agriculture, where labor productivity tends to be low, to the higher-productivity nonagriculture sectors; and (2) increasing the concentration of people and economic activities in urban areas. Not surprisingly, as described in the overview and chapter 1 to this report, policy challenges for managing these two transformation processes were at the center of the 2008 and 2009 editions of the World Development Report (WDR). World Development Report 2008: Agriculture for Development places agricultural growth and transformation of the rural economy at the heart of a strategy for inclusive economic development (World Bank 2007). World Development Report 2009: Reshaping Economic Geography considers the challenges of transformation at three different geographical levels: urbanization at the local level, rural-urban disparities at the national level, and regional integration at the international level (World Bank 2008). Both World Development Reports note that the process of rural-urban transformation is geographically uneven. 109 110 From Farm to Firm Part II of this report sheds further light on the texture of transforma- tion in five countries in South Asia: Bangladesh, India, Nepal, Pakistan, and Sri Lanka. Each of these countries is at a different stage in the transformation process. Of the five, only Nepal is still a predominantly agricultural country where urbanization is at an incipient stage. The other four countries are “transforming” or “urbanizing,” although, as indi- cated by the WDR 2009 agglomeration index, urbanization levels are much higher in India and Pakistan than in Sri Lanka and Bangladesh. Intracountry differences are even more striking. For example, the Kathmandu Valley in Nepal is nearly fully urbanized, with a poverty inci- dence of 3.4 percent. By contrast, the eastern hill region of Nepal is almost entirely dependent on agriculture and has a poverty incidence of 43 percent. The interdependence of rural and urban transformation is acknowledged in both WDRs. But the reports tend to put greater emphasis on the impact of urbanization on rural development (“neigh- borhood effects” in the parlance of WDR 2009) than on the impact of rural policies and institutions on urban development. In practice, urban- ization depends on the flow of goods, capital, and people from rural areas. Interventions in rural, lagging, or agricultural spaces can create significant distortions and spillovers in urban, leading, or nonagricul- tural spaces, and vice versa. The chapters that follow thus aim to explore how the rural-urban transformation process is shaped by the policies, institutions, and initial conditions across space. Among other things, we consider the extent of distortions and spillovers between rural and urban areas. For example, how do agrarian institutions such as land laws shape the rural-urban transformation process? How does proximity to urban centers affect nonfarm rural activities? And what are the push and pull factors shaping migration decisions? Much of part II of this report is structured around country-level case studies for three reasons. First, because of the high degree of intracoun- try differences just noted and documented in much greater detail in chapter 6, the country is the most appropriate unit of analysis. Second, as shown in the WDRs, a country’s stage of rural-urban transformation is determined by a complex interaction of history, policies, and institu- tions. A case study approach enables us to investigate the specific policies and institutions that matter in a particular country context—land laws, for example, may be a binding constraint in one country, whereas in another, uneven infrastructure and services may be the most pertinent issue. Third, data availability is inevitably an important consideration in selecting case study topics. Pathways Out of Poverty: Managing the Rural-Urban Transformation in South Asia 111 The case studies demonstrate how the various policies and institutions in countries in South Asia have led to differences in the level and pace of rural-urban transformation. For example, relative to that of other coun- tries in the region, the provision of infrastructure in Nepal remains weak, in part because of the terrain. As a result, for most of the Nepalese the nearest urban center with a population of 100,000 or more is on average more than eight hours away, limiting the mobility of workers from rural to urban areas. By contrast, Sri Lanka has historically pursued a policy of equitable provision of infrastructure that should have aided the mobility of labor. And yet, because of restrictions on agricultural land transactions in Sri Lanka, relative to that of some East and South Asian countries (such as India and the Republic of Korea), the pace of transformation from an agriculture-based economy to a nonagriculture-based one has been rather slow. This part of the report is organized as follows. Chapter 6 provides country-level diagnostics that document sectoral employment and the income pattern across the five countries. The distribution of economic activities is then paired with information on population density and poverty to identify the stages of urbanization and geographic transforma- tion. Chapter 7 identifies some of the principal institutional and policy constraints affecting the rural-urban transformation in South Asia, includ- ing economic policies, factor markets, access to social services, and infra- structure development. Chapters 8–10 focus on specific country cases. Chapter 8 considers the effects of unequal land distribution in Pakistan and the land market restrictions on rural-urban transitions in Sri Lanka. Chapter 9 looks at two aspects of rural-urban transformation drawn from case studies of Bangladesh: (1) the impact of infrastructure and urban access on shifts from agricultural to nonagricultural employment in rural areas; and (2) the role of agglomeration economies, amenities, and infrastructure in the location choice of manufacturing firms. Chapter 10 concludes this part by shedding some light on the role of infrastructure, amenities, and income in determining the flow of migrants across regions of Nepal. Chapter 11 summarizes both parts of this report, including the key findings from case studies, and draws some policy lessons. Note 1. Part II of this report was produced by a team led by Forhad Shilpi and com- posed of Brian Blankespoor, Toru Nishiuchi, Daan Struyven, Dana 112 From Farm to Firm Thomson, Emily Schmidt, Wei Xiao, and Haomiao Yu. It was prepared under the supervision of Marisela Montoliu Munoz. Katy Hull edited an earlier version of this part of the report. References World Bank. 2007. World Development Report 2008: Agriculture for Development. Washington, DC: World Bank. ———. 2008. World Development Report 2009: Reshaping Economic Geography. Washington, DC: World Bank. CHAPTER 6 Patterns of Rural-Urban Transformation in South Asia Despite considerable achievement in poverty reduction during the last decade, the differences in living standards across regions, and across rural and urban areas, remain substantial in South Asian countries. The widen- ing gaps in living standards have increased social and political tensions. Managing the rural-urban transformation to ensure more inclusive eco- nomic development has thus become an important consideration among policy makers in South Asian countries. As discussed at the outset of this report, the 2008 and 2009 editions of the World Development Report categorize countries (or areas) according to their stage of sectoral or geographical transformation (World Bank 2007, 2008). The typologies in WDR 2008—“agriculture-based,” “trans- forming,” and “urbanized”—are predicated on the contribution of agricul- ture to economic growth and the percentage of the poor residing in rural areas. The typologies in WDR 2009—“1-D,” “2-D,” and “3-D”—are based on measures of economic “density,” the “distance” of the poor from areas of economic density, and the presence of “divisions.” In this chapter, we employ similar typologies to assess patterns of rural- urban transformation in South Asia. We begin by examining the extent to which the poor are located in economically lagging or rural areas and the relative contribution of agriculture to economic growth and employment. 113 114 From Farm to Firm We then consider various measures of urbanization in each country before addressing the contribution of nonfarm activities (whether urban or rural) to growth and poverty reduction. Poverty Incidence and Geographical Distribution The South Asian region is home to 1.4 billion people, 72 percent of whom reside in rural areas. Of the half-billion poor, about 389 million (75 percent of the total) live in rural areas. But living standards vary widely across countries. The per capita gross domestic product of Sri Lanka ($1,074) is almost four and a half times higher than that of Nepal ($243)—see table 6.1.1 In Bangladesh, the consumption of 40 percent of the population falls below the official poverty line; in Sri Lanka, almost 85 percent of the population lives above the poverty line.2 Most of the South Asian countries saw substantial growth in their GDP from 1998 to 2007 (table 6.1). The annual average GDP growth between 1998 and 2007 was the highest in India (7.2 percent) and low- est in Nepal (3.6 percent). The slower population growth during this period translated into healthy growth in per capita income in India and Sri Lanka (higher than 4 percent). The income growth was associated with a substantial reduction in the incidence of poverty, but in many cases an increase in inequality as well. Thus almost all of the South Asian coun- tries experienced a nearly 10 percentage point or more decline in the poverty headcount ratio during the last decade, but the Gini coefficient of distribution of per capita consumption expenditure increased consid- erably in India (from 0.29 in 1993 to 0.35 in 2005), Sri Lanka (from 0.32 in 1991 to 0.40 in 2006), and Nepal (from 0.34 in 1995 to 0.41 in 2003). Table 6.1 Poverty, Inequality, and Economic Growth, South Asia Annual average growth, Gini coefficient GDP per capita 1998–2007 (%) Poverty of per capita (constant 2000 Per capita headcount consumption US$, 2006) GDP income ratio (%) expenditure Bangladesh 419 5.7 3.7 40.0 (2005) 0.31 (2005) India 637 7.2 5.6 27.5 (2005) 0.35 (2005) Nepal 243 3.6 1.5 30.9 (2003) 0.41 (2003) Pakistan 630 5.1 2.7 29.2 (2004) 0.30 (2004) Sri Lanka 1,074 5.0 4.3 15.2 (2006) 0.40 (2006) Sources: World Bank Development Data Platform (DDP), http://go.worldbank.org/DNBRRS9TB0, and various poverty assessments. Patterns of Rural-Urban Transformation in South Asia 115 Intracountry differences in living standards are even more striking. Figure 6.1 plots the poverty headcount ratios for the leading and lagging regions within Bangladesh, Nepal, Pakistan, and Sri Lanka. The leading regions are Western Province in Sri Lanka, Sindh in Pakistan, Kathmandu in Nepal, and the eastern region in Bangladesh. The lagging regions are Uva Province in Sri Lanka, Balochistan in Pakistan, the rural eastern hills in Nepal, and the western region in Bangladesh. According to the esti- mates for 2007, the incidence of poverty in Western Province of Sri Lanka was only 8 percent compared with 27 percent in Uva Province. In Nepal, the estimates of headcount ratios for 2004 show even starker differences: only 3.3 percent in the capital city, Kathmandu, but 42.9 percent in the rural eastern hills. Box 6.1 describes the location of the poor relative to areas of economic density in each of the countries. Contribution of Agriculture to Growth and Employment In the South Asian region as a whole, agriculture remains the most important employer of the labor force. Even in Sri Lanka, the least poor country in the region, agriculture accounts for 34 percent of total employment (table 6.2). The importance of agriculture is much greater in Nepal, where it employs more than 90 percent of the labor force. Consistent with its contribution to employment, agricultural growth Figure 6.1 Poverty Headcount Ratio in Lagging and Leading Regions, South Asia 60 50 40 percent 30 20 10 0 Sri Lanka Pakistan Nepal Bangladesh leading lagging Source: World Bank staff estimates based on latest poverty assessments. Note: For years from which data were taken, see table 6.1. 116 From Farm to Firm Box 6.1 Economic Density versus Density of the Poor Comparisons of the locations of areas of high economic density with areas of high density of the poor reveal three patterns. First, and not surprising, economic activities are much denser in urban areas than in rural areas. Although urban poverty rates are lower, higher population density implies that the number of poor can be large in the more urbanized areas. These areas include Western Province in Sri Lanka, Punjab in Pakistan, the eastern region of Bangladesh, the Kathmandu Valley in Nepal, and the western region of Uttar Pradesh, India. Second, countries have remote regions with a high density of the poor. These regions include the western part of Bangladesh, Central Province of Sri Lanka, the eastern part of Nepal, and a large part of India such as Bihar, Orissa, and part of Uttar Pradesh. Some of these regions are also home to ethnic minorities and a large share of the lower castes, who face considerable barriers to moving. Third, some relatively isolated areas have a higher incidence of poverty than less isolated areas, but are sparsely populated. These areas include Balochistan in Pakistan, the far western region in Nepal, Uva Province in Sri Lanka, and tribal areas in India. Table 6.2 Selected Indicators for Contribution of Agriculture to Economy and Rural Poverty, South Asia Contribution of Share of agriculture in agriculture to Rural poverty Share of GDP Employment GDP growth headcount poor in rural (%, 2006) (%, 2006) (%, 1992–2001) ratio (%) areas (%) Bangladesh 20 51 13.3 43.8 (2005) 82.2 India 18 60 9.6 28.1 (2005) 73.7 Nepal 35 93 32.1 34.6 (2003) 95.3 Pakistan 20 45 12.0 20.6 (2008) 80.0 Sri Lanka 12 34 7.1 15.7 (2006) 79.2 Sources: World Bank Development Data Platform (DDP), http://go.worldbank.org/DNBRRS9TB0, and various poverty assessments. had the strongest poverty-reducing effect in India from 1960 to 1990 (Ravallion and Datt 1996). But the role of agriculture in GDP is diminishing (table 6.2). Its contribution to economic growth is greatest in Nepal, accounting for 35 percent of GDP in 2006. During the same period, its contribution Patterns of Rural-Urban Transformation in South Asia 117 was (close to) 20 percent in Bangladesh, India, and Pakistan, but only 12 percent in Sri Lanka. The average annual growth in agriculture in these countries was just above 2.5 percent compared with more than 5 percent growth in overall GDP. In India, Bangladesh, and Sri Lanka, the contribution of agriculture to overall GDP growth has declined (figure 6.2). During the 1980s, agriculture’s contribution to growth was about 20 percent in two of the five countries—India and Nepal. Therefore, in the 1980s these two countries would have fit into the WDR 2008 category “agriculture-based.” Since 2001, only Nepal, where agriculture has contributed more than 30 percent to GDP growth, still qualifies as an agriculture-based economy. The other countries in the region are best described as “transforming.” In all of the South Asian countries, the incidence of poverty is much higher in rural areas than in urban areas. For example, in Nepal in 2003 the incidence of poverty in rural areas was three and a half times as high as that in urban areas—34.6 percent and 9.6 percent, respectively (tables 6.2 and 6.3). The rural-urban gap is small only in India, where there is a 2.3 percentage point difference between rural and urban poverty. As is typical in both agriculture-based and transforming economies, the majority of the poor still lives in rural areas in all five countries. The share of rural poor in total poor ranges from 74 percent in India to 95 percent in Nepal. Figure 6.2 Contribution of Agriculture to GDP Growth, South Asia 60 50 40 percent 30 20 10 0 Bangladesh India Nepal Pakistan Sri Lanka 1981–90 1991–2000 2001–08 Source: World Bank Development Data Platform (DDP), http://go.worldbank.org/DNBRRS9TB0. 118 From Farm to Firm Table 6.3 Selected Urbanization Indicators, South Asia % of Population population Agglomeration Annual growth Poverty density per in urban index in urban incidence in km2 (2006) areas (2006) (%, 2000) population (%) urban areas Bangladesh 1,198 25 48.0 3.5 (1995–2005) 28.4 (2005) India 378 28 52.4 2.6 (1993–2000) 25.8 (2005) Nepal 198 15 26.0 6.5 (1995–2003) 9.6 (2003) Pakistan 206 34 53.6 3.1 (1998–2004) 10.1 (2008) Sri Lanka 310 38 38.2 1.1 (1995–2006) 6.7 (2006) Sources: World Bank Development Data Platform (DDP), http://go.worldbank.org/DNBRRS9TB0; World Bank 2008; and various poverty assessments. Intracountry differences in employment structure and poverty inci- dence reveal that even within a transforming country, some regions have agriculture-based economies. The intracountry differences in employ- ment and income structure, like those for living standards, are starker than cross-country differences. For example, in Western Province of Sri Lanka, less than 10 percent of the economically active population is employed in agriculture (figure 6.3). Outside Western Province, agriculture accounts for at least one-third of total employment, with its share rising as high as 66 percent in Uva Province, the poorest, where agriculture accounts for 44 percent of the province’s GDP. These patterns hold true in other countries as well. For example, in Nepal 81 percent of the labor force is employed in agriculture in the rural eastern hills compared with 21 percent in the urban Kathmandu Valley. Urbanization, Nonagricultural Activities, and the Rural-Urban Transformation Level of urbanization is a key factor in determining the urbanization strategy for development. Historically, the Gangetic floodplains have sup- ported a high-density population. A great number of urban centers—both large and small—are located in these floodplains and in the coastal areas. Among South Asian countries, India alone has 27 cities with a popula- tion of more than 1 million and 3,700 smaller cities and towns.3 Bangladesh, another country located in the Gangetic plain, has three cities with a population of more than 1 million and 154 smaller cities and towns. The urban populations of all South Asian countries have experienced considerable growth during the last decade (table 6.3). The highest Patterns of Rural-Urban Transformation in South Asia 119 Figure 6.3 Agriculture’s Contribution to GDP and Employment, Sri Lanka 70 60 50 percent 40 30 20 10 0 n n rn a l a ra uw Uv r r te te he nt es es m Ce ut ga W W So rth ra rth No ba No Sa province share of agriculture in employment share of agriculture in GDP Source: World Bank staff estimate based on Household Income and Expenditure Survey, 2002. growth took place in Nepal (6.5 percent), followed by Bangladesh (3.5 percent), and Pakistan (3.1 percent). These urban growth rates exceeded overall population growth rates (2.0 percent in Bangladesh, 2.3 percent in Nepal, 2.2 percent in Pakistan, and 0.8 percent in Sri Lanka). The urbanization rate computed from national statistics—which, of course, vary from country to country—ranges from 35 percent in Pakistan and Sri Lanka to 15 percent in Nepal. Because the national definition of urban area varies from country to country, the comparative measure provided by the agglomeration index in WDR 2009 is useful. The index measures level of urbanization based on the density of population and distance to centers of human settlement of 50,000 or more.4 According to this index, the urban share of popula- tion is 48 percent in Bangladesh, 52 percent in India, and 54 percent in Pakistan. In Nepal, only about 26 percent of population can be regarded as living in urban-like settlements. Both national statistics on urbaniza- tion and the agglomeration index suggest that Bangladesh, India, Pakistan, and Sri Lanka are rapidly urbanizing. In Nepal, urbanization is incipient. That said, the level of urbanization in any of the five coun- tries is subject to considerable variations. Although overall urbanization is incipient in Nepal, some of its regions are urbanizing rapidly. In the central hill region, which includes the Kathmandu Valley, the urban share of population is about 40 percent, and Kathmandu Valley itself is 120 From Farm to Firm almost fully urbanized. In Pakistan, one of the most urbanized countries in South Asia, the urban share of population is only about 13 percent in the northwestern frontier province. The contribution of urbanization to economic growth cannot be dis- cerned precisely because GDP is reported by sector, not location, and a large portion of nonfarm activities are located outside urban areas. But recent World Bank poverty assessments indicate substantial growth in consumption in urban areas. As noted, the incidence of poverty is lower in urban areas than in rural areas in all South Asian countries. The poverty trends during the last decades indicate a larger decline in urban poverty than in rural poverty in all countries except India, where the rural-urban gap was small initially. In Nepal, urban poverty declined by 52 percent (from 21.6 to 9.6 percent) between 1995–96 and 2003; rural poverty declined by 21 percent over the same time period. Similar patterns are found in Bangladesh, Pakistan, and Sri Lanka. In recent decades, urbanization has been found to reduce poverty in rural areas as well. Cali and Menon (2009) report that urbanization had a strong, robust effect on rural poverty reduction in post-1991 India. This poverty-reducing effect of urbanization remains significant even after controlling for rural-to-urban migration and reclassification of rural areas. The World Bank’s 2009 poverty assessment of India finds the link- age between urban growth and rural poverty reduction to be strongest for smaller cities and towns. For Bangladesh, Shilpi (2008) finds a stronger spillover of urban growth to poverty reduction in surrounding rural areas. Rural Nonfarm Activities and the Rural-Urban Transformation Because GDP growth is reported by sector, not location, it is not possible to provide a precise breakdown of the relative contribution of nonagricul- tural activities to growth in urban and rural areas. Growth in nonfarm activities in urban areas appears to have made a significant contribution to overall growth in those areas, and rural areas have also undergone a sec- toral shift from farm to nonfarm activities. The contributions of rural non- farm activities to rural employment and household income are significant (figure 6.4), ranging from about a third in Nepal and Pakistan to more than half in Sri Lanka and Bangladesh.5 The larger contribution of non- farm activities to income relative to employment in most countries sug- gests that, on average, these activities provide a higher labor return than farming. For example, in Bangladesh the incidence of poverty is lower Patterns of Rural-Urban Transformation in South Asia 121 Figure 6.4 Contribution of Rural Nonfarm Sectors to Rural Income and Employment, South Asia 70 60 50 percent 40 30 20 10 0 Bangladesh Nepal Pakistan Sri Lanka India (2005) (2003) (2001) (2002) (2005) share of nonfarm activities in rural employment share of nonfarm activities in rural household income Source: Various poverty assessments. among rural households with access to nonfarm employment than those depending purely on agriculture (World Bank 2004). For India, Foster and Rosenzweig (2003) report that growth in nonfarm activities in rural areas reduced inequality significantly, and the impact of nonfarm employment growth on poverty reduction was larger than that of agricultural produc- tivity growth. According to employment data, a large fraction of nonfarm activities are located in rural areas. For example, in Pakistan the share of rural areas in total employment is 50 percent in the manufacturing sector and 47 percent in the service sector. The share of rural areas in total nonagri- cultural employment is much higher in Bangladesh, about 66 percent. Even in a predominantly agricultural country—Nepal—more than half of those employed in nonfarm activities live in rural areas. Conclusion The evidence from South Asian countries demonstrates clearly that coun- tries are at different stages of the rural-urban transformation. There are substantial differences in the level and pace of urbanization, in agriculture’s 122 From Farm to Firm contribution to growth and employment, and in the importance of the rural nonfarm sector in income and employment generation. Despite these differences, Bangladesh, India, Pakistan, and Sri Lanka can be grouped broadly as transforming and rapidly urbanizing countries. Nepal remains the only country in which agriculture is still the mainstay of the economy and urbanization is at the incipient stage. Intracountry differences in the pattern of rural-urban transformation are, however, much starker than cross-country differences. Within each country, regions range from agriculture based to highly urbanized. In terms of distribution of the poor and of economic density, and the pres- ence of barriers to mobility, regions in these countries fit all combina- tions of density, distance, and division suggested by WDR 2009. The presence of such differences implies a need for examining the challenges of rural-urban transformation on a much finer geographical scale. The large intracountry variations in the pattern of rural-urban trans- formation are somewhat surprising. As WDR 2009 observes, divisions and distances within a country are likely to be smaller than those at the international level because of the presence of national borders, curren- cies, and macroeconomic policies. But this is not true for most South Asian countries. However, both WDRs also emphasize that observed patterns of economic transformation are outcomes of underlying poli- cies, institutions, and initial conditions. In the next chapter, we explore both the intra- and intercountry differences in policies, institutions, and initial conditions. Notes 1. At purchasing power parity, the per capita GDP is $1,340 in Bangladesh, $2,740 in India, $1,040 in Nepal, $2,570 in Pakistan, and $4,210 in Sri Lanka. 2. At a $1.25 per day poverty line, the incidence of poverty is 50 percent in Bangladesh (2005), 44 percent in rural India and 36 percent in urban India (2004), 55 percent in Nepal (2003), 22 percent in Pakistan (2004), and 14 percent in Sri Lanka (2002). 3. Census of India, 2001. http://censusindia.gov.in. 4. The population density has to be at least 150 persons per square kilometer, and distance has to be less than or equal to one hour of travel time from the center of a city or town of at least 50,000 people. 5. Because the definitions of rural area and urban area are different across coun- tries, the shares of nonfarm sectors in employment and income are not strictly comparable across countries. Patterns of Rural-Urban Transformation in South Asia 123 References Cali, M., and Carlo M. 2009. “Does Urbanisation Affect Rural Poverty? Evidence from Indian Districts.” Spatial Economics Research Centre Discussion Paper 14, London School of Economics, U.K. Foster, A., and M. Rosenzweig. 2003. “Agricultural Development, Industrialization and Rural Inequality,” Brown University, Providence, RI. Ravallion, M., and G. Datt. 1996. “How Important to India’s Poor Is the Sectoral Composition of Economic Growth?” World Bank Economic Review 10 (1): 1–25. Ravallion, M., and D. van de Walle. 2008. “Does Rising Landlessness Signal Success or Failure for Vietnam’s Agrarian Transition?” Journal of Development Economics 87 (2): 191–209. Shilpi, F. 2008. “Migration, Sorting and Regional Inequality: Evidence from Bangladesh.” Policy Research Working Paper 4616, World Bank, Washington, DC. World Bank. 2004. Bangladesh: Promoting the Non-Farm Sector in Bangladesh. Report 29719-BD, World Bank, Washington, DC. ———. 2007. World Development Report 2008: Agriculture for Development. Washington, DC: World Bank. ———. 2008. World Development Report 2009: Reshaping Economic Geography. Washington, DC: World Bank. CHAPTER 7 Policies, Institutions, and Initial Conditions As shown in chapter 6, South Asian countries are at different stages of the rural-urban transformation. The key indicators of transformation— such as the incidence and geographical distribution of poverty, the con- tribution of agriculture to employment and income, and levels of urbanization—vary considerably within and across countries. As argued in the 2008 and 2009 editions of the World Development Report, rural- urban transformations are influenced by policies and institutions (World Bank 2007, 2008b). Although any policy could conceivably affect trans- formation, in this chapter we focus on the major levers identified by the WDRs: economic policies, land and labor market institutions, invest- ments in human capital in the form of health and education policies, and investments in physical capital in the form of connective infrastructure. Economic Policies During the decades after their independence, South Asian countries implemented distortive trade and exchange rate policies. Although eco- nomic reforms in subsequent decades reduced distortions, liberalization progressed at an uneven pace across countries. Sri Lanka led the way, implementing major trade and exchange rate reforms as early as 1977. 125 126 From Farm to Firm India, by contrast, lagged behind until a balance-of-payments crisis kick-started a wave of reform in 1991. Bangladesh, Nepal, and Pakistan also liberalized their economies during the early 1990s. The liberaliza- tion packages typically included a reduction in, or even the elimination of, quantitative restrictions on trade, a reduction in tariff rates and sim- plification of tariff structures, and rationalization of exchange rate regimes. As shown in figure 7.1, the average tariff rates in Bangladesh, India, and Pakistan were between 47 and 56 percent during the early 1990s; by 2006 the rate had fallen to 15 percent or below in all South Asian countries. In almost every country in South Asia, reforms in the agriculture sec- tor lagged far behind monetary and trade policy reforms. Agricultural policies in South Asian countries have been shaped by the desire to ensure food self-sufficiency, resulting in trade restrictions, price support, generous input subsidies, and public investments in irrigation and research and extension systems. Figure 7.2 presents the average tariff on agricultural products in each country. The average tariff rate protecting the agriculture sector in India in 2004–05 was about 40 percent—almost four times the level of India’s average industrial tariff and among the highest in the world. A comparison of figures 7.1 and 7.2 reveals a simi- lar skew toward agricultural tariffs for Bangladesh and Sri Lanka. Figure 7.1 Average Tariff Rate (Unweighted), South Asia 60 50 40 percent 30 20 10 0 Bangladesh India Nepal Pakistan Sri Lanka 1992–94 2005–07 Source: United Nations Conference on Trade and Development (UNCTAD) TRAINS database. Policies, Institutions, and Initial Conditions 127 Figure 7.2 Average Tariff Rate on Agricultural Products, South Asia 45 40 35 30 percent 25 20 15 10 5 0 Bangladesh India Nepal Pakistan Sri Lanka (2004–05) (2004–05) (2002–03) (2004–05) (2004) Source: Anderson and Martin 2009. In addition to tariff protection, India and Sri Lanka have provided agri- cultural subsidies for power and fertilizers, among other investment and subsidy schemes. In India, the total subsidies were estimated to be about $7.8 billion in 2004, with much of it in the form of electricity subsidies for rice and wheat crops (Pursell, Gulati, and Gupta 2007). In 2006, fertilizer subsidies in Sri Lanka were doubled from about $40 million to $80 million, and they rose even further after the world price spike in 2007. Recent estimates of the nominal rate of assistance (NRA) to farmers demonstrate the extent of total agricultural support (Anderson and Martin 2009). During the early 1990s, the NRA was negative in Bangladesh, Pakistan, and Sri Lanka, signifying implicit taxation of farm- ers’ incomes (figure 7.3). By the early 2000s, the NRA had become pos- itive in all three countries and substantial in India (15.8 percent) and Sri Lanka (9.5 percent). These levels of NRA would translate into a subsidy per farmer of $4 in Pakistan, $5 in Bangladesh, $40 in Sri Lanka, and $57 in India. The implicit subsidization of agriculture in recent years has protected farmers’ incomes in India and Sri Lanka. But subsidization has also come with a cost. Most of the subsidies went to two subsistence crops—rice and wheat—which deterred farmers from diversifying into high-value crops such as fruits, vegetables, and livestock products. High spending on 128 From Farm to Firm Figure 7.3 Nominal Rate of Assistance, South Asia 20 15 10 percent 5 0 –5 –10 Bangladesh India Pakistan Sri Lanka 1990–94 2000–04 Source: Anderson and Martin 2009. subsidies has also been at the expense of investments in other rural infra- structure such as roads (World Bank 2006). Factor Markets: Land and Labor Land market institutions can have far-reaching effects on the rural-urban transformation by altering incentives to invest or to move from rural to urban areas. Even though all countries in South Asia have instigated some form of land reform, the pace of implementation has been uneven. India has been a leader, particularly in the area of tenancy reform. By contrast, Pakistan has not yet dealt with its feudal landownership system in some provinces such as Sindh. Government interventions have focused on three main areas of the land markets. First, ceiling legislation has been a common feature of land reform in almost all South Asian countries. This legislation allows govern- ments (and states in India) to expropriate land held by private owners in excess of ceilings and to transfer it to poor farmers and landless agricul- tural workers. Ceilings vary from country to country and across states in India—from 15 acres in the Indian state of Tamil Nadu and about 20 acres in Bangladesh to 500 acres of irrigated (and 1,000 acres of unir- rigated) land in Pakistan. But in spite of de jure ceilings, implementation Policies, Institutions, and Initial Conditions 129 has been slow in all countries, with the practice of subdivision enabling landowners to avoid expropriation. As a result, substantial inequality in the distribution of landownership remains, particularly in Pakistan. WDR 2008 estimates the Gini coefficient of landownership distribution to be 0.61 in Pakistan, 0.48 in Bangladesh, and 0.45 in India (World Bank 2007). Second, almost all countries have tenancy legislation aimed at improving tenure security and limiting rents. Legal provisions differ across countries (and states in India) in terms of conditions for permit- ting tenancy, the maximum rent ceilings, and the types of property rights awarded to tenants. But the actual implementation of tenancy laws has been slow and weak in most counties, with the exception of some states in India such as West Bengal. Many landlords were able to evict tenants in anticipation of the laws coming in effect, and even long-term tenants remain vulnerable to eviction despite the legislation. When effectively enforced, however, such as in West Bengal, tenancy laws have significant positive effects on farm productivity (Banerjee, Gertler, and Ghatak 2002). Third, governments in South Asian countries have also intervened in the land sales markets, restricting the sale and subdivision of land. For example, in Sri Lanka large tracts of land came under government ownership after enactment of the Crown Lands Encroachment Ordinance, and these were subsequently distributed to landless farm- ers under the Land Development Ordinance (LDO). Recipients of LDO land have the right to occupy and cultivate the land in perpetuity, subject to restrictions imposed on sale, leasing, and mortgaging, and to conditions related to abandoning or failing to cultivate the land. In India, the government maintains laws that restrict the sale of land from tribals to nontribals, and some states impose subdivision restrictions to prevent fragmentation of holdings below a minimum size. Sri Lanka and some Indian states have also imposed restrictions on the conversion of agricul- tural land to nonagricultural use, even at the urban periphery. Labor market institutions can have major effects on rural-urban trans- formations through their impacts on job creation, firm size, and sectoral transformation, among other factors. In most South Asian countries, labor market regulations apply primarily to the larger formal sector firms (for example, those in India having more than 20 workers and those in Bangladesh having 10). The regulations typically stipulate the minimum hours of work, minimum wages, the right of association, and safety and health standards. All countries also have rules for termination of workers 130 From Farm to Firm and dispute resolution. In India, no less than 45 central laws and 170 state statutes deal with labor market issues ranging from industrial relations to social security and insurance. In Sri Lanka, the Termination of Employment Act makes it virtually impossible to fire workers who have been employed for more than six months for reasons other than serious and well-documented disciplinary problems. In India, terminations have to be approved by the state authority. The cost of firing is highest in Sri Lanka, worth about 178 weeks of salary (figure 7.4). Compared with the difficulty of firing, the hiring of workers presents few or no problems in Bangladesh, India, and Sri Lanka. By contrast, the difficulties of hiring and firing are almost equally large in Nepal and Pakistan. Because labor regulations tend to “bite” only in the larger formal sec- tor firms, their impact is uneven across sectors. Of all countries in the region, India has been best documented. There, regulations have caused the loss of 30–40 percent of formal manufacturing jobs (Ahsan and Pages 2007). According to Besley and Burgess (2004), amendments of the Industrial Disputes Act in a pro-worker direction lowered output, employment, investment, and productivity in formal manufacturing firms and increased urban poverty. Aghion, Burgess, and Redding (2008) find there was higher output growth in pro-employer states than in pro- worker states after deregulation of License Raj in India in 1991. Ahsan and Pages (2007) report that an increase in redundancy cost increases Figure 7.4 Difficulty of Hiring and Redundancy Costs, South Asia 200 180 160 140 120 100 80 60 40 20 0 Bangladesh India Nepal Pakistan Sri Lanka difficulty of hiring index (0–100) redundancy costs (weeks of salary) Source: Doing Business database, http://www.doingbusiness.org/; World Bank (2007). Policies, Institutions, and Initial Conditions 131 salaries for employed workers, but lowers employment, output, and investment, and increases the share of informal work. The higher cost of labor disputes reduces salaries for employed workers, and also leads to lower employment, lower output, lower investment, and a higher share of informal work. These regulations, by dissuading firms from taking advantage of economies of scale, may have slowed down the transition from farm to nonfarm jobs and from informal to formal jobs. Human Capital Endowments Human capital endowments are important determinants of urbanization and economic growth. Examples are the health and education status of South Asian countries (table 7.1). The cross-country differences in health indicators are quite substantial. Sri Lanka ranks highest in per capita health care expenditure, access to improved sanitation, and infant survival rates, among other measures of health. The per capita health care expen- diture is lowest in Bangladesh. The number of physicians to the popula- tion varies between 1 to 5,000 persons in Nepal to 1 to 1,400 in Pakistan. In Bangladesh, India, and Nepal, less than 30 percent of the population has access to improved sanitation facilities. As a result of these disparities in health care inputs, life expectancy varies between 65.7 years in Bangladesh and 74.0 years in Sri Lanka. Cross-country differences in infant mortality rates are even more striking. The probability of an infant dying at birth is more than six times higher in Pakistan (7.8 percent) than in Sri Lanka (1.2 percent). In education, gender differences in gross enrollment rates have been eliminated in all countries except Pakistan, where women remain at a sig- nificant disadvantage. Bangladesh lags behind India and Sri Lanka in terms of average years of schooling (8 years in Bangladesh versus 10 in India), primary school completion rates (56.3 percent in Bangladesh ver- sus 96.2 percent in Sri Lanka), and adult literacy (53.5 percent in Bangladesh versus 90.8 percent in Sri Lanka). Indeed, adult literacy remains a challenge in all countries except Sri Lanka, with almost half of all adults in Nepal and Pakistan and one-third in India classified as illiter- ate. But even in Sri Lanka, challenges remain. Two-thirds of primary school graduates in Sri Lanka lack basic language and math skills (World Bank 2006). Intracountry differences in access to schools and health facilities are somewhat less striking. In Nepal, a country with difficult terrain, a typi- cal rural household lives within 15 minutes of a school and half an hour 132 From Farm to Firm Table 7.1 Selected Human Development Indicators, South Asia Year Bangladesh India Nepal Pakistan Sri Lanka Health indicator Physicians (per 1,000 persons) 2004 0.26 0.60 0.21 0.74 0.55 Health expenditure, total (% of GDP) 2006 3.2 3.6 5.1 2.0 4.2 Health expenditure per capita (current US$) 2006 12 29 17 16 62 Access to improved sanitation (% population) 2006 36 28 27 58 86 Life expectancy (years) 2007 65.7 63.4 66.3 66.2 74 Infant mortality (per 1,000 live births) 2006 52 57 46 78 12 Education indicator Expected years of schooling 2006 8.0 10.0 8.9 6.5 — Gross enrollment (% of school-age population) 2006a 103 112 126 84 108 Gross enrollment: male (% of school-age population) 2006a 101 114 129 94 108 Gross enrollment: female (% of school-age population) 2006a 105 109 123 73 108 Primary school completion rate, total (% of relevant age group) 2006 56.3 85.7 76.0 61.8 96.2 Adult literacy (% of population aged 15+) 2007 53.5 66.0 56.5 54.2 90.8 Sources: World Bank Development Data Platform (DDP), http://go.worldbank.org/DNBRRS9TB0; Food and Agri- culture Organization (FAO) nutrition country profiles; United Nations Children’s Fund (UNICEF), http://unicef.org; and Human Development Index Database. Note: — = not available. a. Estimates for Bangladesh are for 2004. of a hospital (table 7.2). In the Mid-Western Region, however, both schools and hospitals are marginally less accessible. At the other end of the spectrum, Sri Lanka is renowned for its equitable provision of health and education services across different regions, as depicted in figure 7.5. Most Sri Lankans live within 30 minutes of a health facility, 1.4 kilome- ters of a basic health clinic, and 4.8 kilometers of a government-sponsored free health care facility (World Bank 2010). Access to schools is similarly equitable across regions in Sri Lanka. Intracountry differences in access to health care and education are also small in Bangladesh and Pakistan. Policies, Institutions, and Initial Conditions 133 Table 7.2 Access to Education and Health Care Facilities in Rural Areas: Nepal, 2003 Median Mean Standard hours hours error hours Travel time to nearest school Eastern 0.17 0.31 0.02 Central 0.17 0.34 0.03 Western 0.17 0.25 0.02 Mid-Western 0.25 0.37 0.05 Travel time to nearest hospital Eastern 0.50 0.78 0.08 Central 0.50 0.76 0.08 Western 0.50 0.94 0.15 Mid-Western 0.75 0.97 0.11 Source: World Bank staff estimates, Nepal Living Standards Survey (NLSS), 2003. The only exception to this general pattern is India, where health and education indicators are much worse in the lagging regions. In 2000, for example, 52.77 percent of children were malnourished in the low-income states versus 29.33 percent in the northeastern special category states.1 In the low-income states, only 6.41 percent of villages have access to high schools compared with 25.00 percent in the higher-income states. Similarly, female adult literacy varies between 39 percent in the low- income states and 67 percent in the northeastern special category states. But relatively equitable access to health care facilities and schools across regions within a country (with the exception of India) does not translate into equal quality of service. In Sri Lanka, school performance is worst in the northern and eastern provinces. In Bangladesh, at 74 percent, the absentee rate for doctors in rural primary health clin- ics is very high (Kremer et al. 2005). And in poor neighborhoods of Delhi, qualified public doctors render worse care than unqualified pri- vate sector doctors (Das and Hammer 2005). The inequalities in the quality of health care and education have both spatial and socioeco- nomic dimensions. Over 33 percent of public spending on health care in India is enjoyed by the richest 20 percent of the population, while less than 10 percent goes to the poorest consumption quintile (World Bank 2006). The quality of education received by the lowest consumption decile of Indian students is much worse than that of the topmost decile (Das and Zajonc 2008). 134 From Farm to Firm Figure 7.5 Access to Health Care Facilities, Sri Lanka Source: World Bank 2010. Connective Infrastructures As indicated in WDR 2009, by reducing the distance between human set- tlements and centers of economic density, connective infrastructure is a crucial means of easing rural-urban transformations (World Bank 2008b). In this section, we address the intra- and intercountry differences in roads, power, and communications. Road densities in most of the South Asian countries fare well by international standards (figure 7.6). Bangladesh, India, Pakistan, and Sri Lanka have more than 1 kilometer of road for each square kilometer of area. Road density per capita is highest in Sri Lanka, which has 5 kilometers of road per 1,000 inhabitants; India has 3 kilometers of road Policies, Institutions, and Initial Conditions 135 Figure 7.6 Road Density, South Asia 6 5 square kilometers 4 3 2 1 0 Bangladesh India Nepal Pakistan Sri Lanka road density per km2 road density per capita Source: World Bank Development Data Platform (DDP), http://go.worldbank.org/DNBRRS9TB0. per 1,000 inhabitants. By contrast, the transport network in Nepal remains one of the weakest because of the country’s difficult terrain, and road density is only about half a kilometer per 1,000 inhabitants. The share of paved roads differs significantly across countries, ranging from 9.5 percent in Bangladesh (2002–06) to 81.0 percent in Sri Lanka (2003)—see table 7.3. Road density alone does not measure accessibility, because it reveals nothing about road quality or the extent to which roads are connect- ing villages to markets and urban centers (rather than other villages). Even in Sri Lanka, which has the highest share of paved roads, road quality is an issue. As measured by the International Roughness Index (IRI), roads in Sri Lanka are poor, with only 2 percent of national roads of sufficient quality to support traffic volumes. Estimates of travel times by road to cities of 100,000 or more are available for all coun- tries except India. The average travel time is about 1.7 hours in Bangladesh, 2.4 hours in Pakistan and Sri Lanka, and 8.2 hours in Nepal (table 7.3). Intracountry differences in remoteness are even more striking. Road quality tends to be much poorer and accessibility much lower in the less wealthy and more rural regions. Some regions of Nepal (figure 7.7) and Pakistan are 38 hours or more away from a city of 100,000 or more. 136 From Farm to Firm Table 7.3 Selected Infrastructure Indicators, South Asia Year Bangladesh India Nepal Pakistan Sri Lanka Travel time to cities of 100,000 persons or more Average (hours) 2009 1.69 — 8.24 2.38 2.37 Minimum (hours) 2009 0.01 — 0.37 0.02 0.02 Maximum (hours) 2009 10.3 — 38.2 44.5 15.2 Telephone lines (per 100 persons) 2007 0.7 3.5 2.5 3.0 13.7 Cell phones (per 100 persons) 2007 21.7 20.8 11.6 38.7 39.9 Rail lines (km) 2005 2,855 63,465 — 7,791 1,200 Roads (10,000 km) 2003–06 24 (2003) 332 (2006) 2 (2004) 26 (2006) 10 (2003) Paved roads (% of all) 2002–06 9.5 (2003) 47.4 (2002) 56.9 (2004) 65.4 (2006) 81.0 (2003) Access to electricity (%) 2000 20.4 43.0 15.4 52.9 62.0 Sources: World Bank staff estimates (travel time); all others, World Bank Development Data Platform (DDP), http://go.worldbank.org/DNBRRS9TB0. Note: — = not available. Because of Nepal’s difficult terrain, the average travel time to a paved road is 4.87 hours and to a commercial bank 3.49 hours, in contrast to 0.18 hours and 0.35 hours, respectively, in urban areas. Even in Sri Lanka, some remote locations are more than 15 hours away from the nearest large town (figure 7.8), and roads are narrower and bumpier off main routes and in rural areas than on major national routes and in Colombo. Although road densities are high and travel times are shorter in Bangladesh, regions within the country remain cut off from each other because of the lack of bridges on larger rivers. There are also substantial differences in access to electricity (table 7.3). In 2000 the share of households with access to electricity was 62.0 percent in Sri Lanka, and then dropping to 52.9 percent in Pakistan, Policies, Institutions, and Initial Conditions 137 Figure 7.7 Travel Times, Nepal Sources: District boundaries: MapLibrary; travel times: estimates by World Bank based on data from Nepal Living Standard Survey, 2003; cities: Global Rural-Urban Mapping Project (GRUMP); 2000 population estimates, Center for International Earth Science Information Network (CIESIN), Columbia University. 43.0 percent in India, 20.4 percent in Bangladesh, and 15.4 percent in Nepal. Power shortages are a critical concern for firms; more than 75 per- cent of the firms in Bangladesh (and almost 40 percent of those in India, Pakistan, and Sri Lanka) perceive power shortages to be a major obstacle. In Bangladesh, resolution of power shortages would add two percentage points to economic growth (World Bank 2008a). Intracountry differences in access to power are also very significant. In India, 99 percent of the villages in the higher-income states (but only 69 percent of those in low- income states) have access to power.2 In Kathmandu, Nepal, nearly all households in urban areas have access to electricity in contrast to one- quarter of rural residents. Access to a landline phone is very limited in all countries, ranging from 13.7 percent of households in Sri Lanka to 0.7 percent in Bangladesh. Cell phone penetration rates in 2007 were 11.6 percent in Nepal, 20.8 percent in India, 21.7 percent in Bangladesh, 38.7 percent in Pakistan, and 39.9 percent in Sri Lanka (table 7.3). In Nepal, of the 3,919 village development committees, more than 1,000 do not have any telephone link. 138 From Farm to Firm Figure 7.8 Travel Times, Sri Lanka Sources: District and Divisional Secretariat Divisions: MapLibrary; travel times: estimates by World Bank based on road network; cities: Global Rural-Urban Mapping Project (GRUMP); 2000 population estimates, Center for Inter- national Earth Science Information Network (CIESIN), Columbia University. Conclusion Our preliminary analysis suggests that, as indicated by WDR 2008 and WDR 2009, policies and institutions will have a profound effect on the pace and shape of rural-urban transformations. In South Asia, macroeco- nomic policies and human capital investments vary across countries, but are relatively uniform within countries. Land and labor institutions, however, as well as connective infrastructure, tend to vary both across countries and within countries. In the chapters that follow, we present in-depth results from case studies focusing on the role of institutions and connectivity in determining some key processes of the rural-urban transformation. Policies, Institutions, and Initial Conditions 139 Notes 1. The low-income states are Bihar, Chattisgarh, Jharkhand, Orissa, Rajasthan, Uttar Pradesh, and Madhya Pradesh. The northeastern special category states are Arunchal, Assam, Manipur, Megalaya, Mizoram, Nagaland, and Tripura. 2. The higher-income states are Goa, Haryana, Maharashtra, Punjab, and Tamil Nadu. The group of low-income states includes Bihar, Chattisgarh, Jharkhand, Orissa, Rajasthan, Uttar Pradesh, and Madhya Pradesh. References Aghion, P., R. Burgess, and S. Redding. 2008. “The Unequal Effects of Liberalization: Evidence from Dismantling the Licence Raj in India.” American Economic Review 98 (4): 1397–1412. Ahsan, A., and C. Pages. 2007. “Are All Labor Regulations Equal? Assessing the Effects of Job Security, Labor Dispute, and Contract Labor Laws in India.” Policy Research Working Paper 4259, World Bank, Washington, DC. Anderson K., and W. Martin. 2009. Distortions to Agricultural Incentives in Asia. Washington, DC: World Bank. Banerjee A., P. Gertler, and M. Ghatak. 2002. “Empowerment and Efficiency: Tenancy Reform in West Bengal.” Journal of Political Economy 110 (2): 239–80. Besley, T., and R. Burgess. 2004. “Can Labor Regulation Hinder Economic Performance? Evidence from India.” Quarterly Journal of Economics 119 (1): 91–134. Das, J., and J. Hammer. 2005. “Which Doctor? Combining Vignettes and Item Response to Measure Clinical Competence.” Journal of Development Economics 78 (2): 348–83. Das J., and T. Zajonc. 2008. “India Shining and Bharat Drowning: Comparing Two Indian States to the Worldwide Distribution in Mathematics Achievement.” Policy Research Working Paper 4644, World Bank, Washington, DC. Kremer, M., N. Chaudhury, F. H. Rogers, K. Muralidharan, and J. Hammer. 2005. “Teacher Absence in India: A Snapshot.” Journal of the European Economic Association 3 (2–3): 658–67. Pursell, G., A. Gulati, and K. Gupta. 2007. “Distortions to Agricultural Incentives in India.” Agricultural Distortions Working Paper 34, World Bank, Washington, DC. World Bank. 2006. Can South Asia End Poverty in a Generation? South Asia Region Unit. Washington, DC: World Bank. ———. 2007. World Development Report 2008: Agriculture for Development. Washington, DC: World Bank. 140 From Farm to Firm ———. 2008a. Harnessing Competitiveness for Stronger Inclusive Growth. Bangladesh Second Investment Climate Assessment. Washington, DC: World Bank. ———. 2008b. World Development Report 2009: Reshaping Economic Geography. Washington, DC: World Bank. ———. 2010. “Sri Lanka: Connecting People to Prosperity.” Draft report of South Asian Poverty Reduction and Economic Management Unit, World Bank, Washington, DC. CHAPTER 8 Land Market Institutions and the Rural-Urban Transformation Access to land is a key determinant of a household’s economic well-being and social status. The institutions that govern land rights can influence people’s incentives to invest in productivity improvements and determine their access to credit and public services. The impact of land institutions can extend beyond household welfare to the long-term path and pace of economic transformation. In this chapter, we use case studies to explore the impact of two fundamental land market institutions on the rural- urban transformation. First, we look at the effects of inequality of landownership on rural and urban employment patterns and wages in Pakistan. Second, we consider the impact of restrictions on land market transactions on the rural-urban transformation in Sri Lanka. Impact of Land Inequality: The Existing Evidence The unequal distribution of land is a major source of income inequality and higher incidence of poverty in rural areas in most developing coun- tries (World Bank 2007c). Landownership inequality affects productivity and growth through two principal mechanisms. First, it can induce a higher incidence of share tenancy, which tends to depress investment in land improvement and newer technology. For example, Jacoby and 141 142 From Farm to Firm Mansuri (2008) find that in Pakistan, investments are lower in leased land than owned land. Second, landownership inequality tends to cement the position of rural elites, who may be incentivized to hamper investments in public institutions. For example, Banerjee and Iyer (2002) report that the colo- nial Zamindari system in India compounded inequalities in land distribu- tion and created an entrenched landlord elite, even after the abolition of the Zamindari system itself. This development resulted in lower agricul- tural investment and productivity and lower public investments in health, education, and agricultural technology in districts dominated by land- lords. Indeed, according to Galor, Moav, and Vollrath (2009), land inequality has a significant negative effect on public investment in educa- tion at the state level in the United States. They argue that, because edu- cation facilitates the movement of workers from the agriculture to nonagriculture sectors, landowners have an incentive to limit public investment in education. The negative influence of land inequality on public education suggests that transitions from farm to nonfarm occupations are likely to be slower in countries and areas with higher land inequality. The ultimate effect of landownership on the transformation from agricultural to nonagricultural activities is complex, because it affects not only income distribution and access to credit but also savings and investment. The evidence summa- rized in box 8.1 suggests that the impact of land inequality on the pattern and pace of occupational transformation is likely to be nonlinear, depend- ing on the extent of inequality itself. Land Inequality and Occupational Transitions in Pakistan Among the South Asian countries, Pakistan has the most unequal distri- bution of agricultural land. According to the 2000 agricultural census, the Gini coefficient of the landownership distribution is 0.66. The Pakistan Social and Living Standards Measurement Survey (PSLM), 2004–05, cites the coefficient as higher still, 0.77. There are, however, differences in the distribution of landownership across provinces, and the Gini coefficient is particularly high in Punjab (0.76) and Balochistan (0.78). Unequal distribution of land is cited as a major source of income inequality in rural Pakistan. Adams and Alderman (1992) find that between a third and three-quarters of income inequal- ity in rural areas can be attributed to high inequality in landownership. Regression-based decompositions of income inequality also confirm the key role of land inequality in determining rural income inequality Land Market Institutions and the Rural-Urban Transformation 143 Box 8.1 Inequality and the Evolution of Occupational Structure: Theoretical Insights Banerjee and Newman (1993) contend that the distribution of wealth affects the employment composition of an economy and vice versa. In their model, because of capital market imperfections, people can borrow only limited amounts. As a result, occupations that require higher levels of investment are beyond the reach of poor people, who choose instead to work for wealthier employers. Depending on the labor market conditions and on their wealth, other agents become self- employed in low-scale production or remain idle. The structure of occupation in an economy determines how much people save and how much risk they bear. These factors give rise to a new distribution of wealth in a dynamic setting. In a static equilibrium, high inequality is associated with more wage work and lower wages, and thus greater poverty. In a dynamic setting, the degree of inequality determines the economy’s growth path. When the distribution of initial wealth is highly unequal (with a large number of poor, a smaller middle class, and a minor- ity of highly affluent people), the economy may converge to stagnation because of low wages and thus low savings and demand. When inequality is not so extreme and there is a sizable middle class, the economy may converge to a pros- perous equilibrium. The inequality in this case forces poorer agents to become workers and over time allows large-scale industrial production. When inequality in wealth distribution is only moderate and very few poor people are available to participate in wage work, the economy converges to an equilibrium where self- employment in smaller business enterprises predominates. (Naschold 2009). But empirical evidence on the impact of land inequal- ity on occupational transitions in Pakistan is sparse. To investigate the impact of land inequality on occupational structures and wages in Pakistan, we run regressions on several rounds of the Labor Force Survey data from Pakistan. The empirical analysis presents several challenges. First, the unobserved location characteristics that may have affected wage distribution can be correlated with the distribution of land. Second, there may be reverse causality, because occupation structure may affect savings and investment and thus land distribution. The empirical methodology used to deal with these problems is elaborated in Shilpi (2009) and omitted here for the sake of brevity. In this chapter, we pres- ent the main results of the empirical analysis. 144 From Farm to Firm For employment patterns, our regressions indicate that the relationship between landownership inequality and employment structure is nonlin- ear (table 8.1). Except for manufacturing employment and nonagricul- tural self-employment, the index of land inequality and its squared term are jointly statistically significant. The estimated coefficients suggest a concave relationship between land inequality and employment in non- farm self-employment, wage work in agriculture, and services employ- ment. Thus when inequality is low, self-employment in farming predominates. As inequality increases, more and more people engage in off-farm work. At a very high level of inequality, people revert to self- employment in farming.1 The pattern is reversed only for employment in public services. Like land inequality, education is an important determinant of participa- tion in nonfarm activities. An education level higher than or equal to middle school increases significantly the probability of participation in nonfarm self-employment and services employment (both private and public). Education between the primary and middle school levels also improves the rate of participation in nonfarm self-employment relative to self-employment in agriculture. As expected, those who are engaged in agri- cultural wage work lack primary-level education. The manufacturing sector also seems to attract mainly those who lack primary-level education. Regression results also suggest a concentration of activities in and around cities. The farther the distance to urban centers with a population of 100,000 or more, the greater is the negative effect on participation in manufacturing and services activities. This finding indicates a geographical concentration of wage employment in nonfarm activities in and around large urban centers. The metropolitan and urban dummies are also signif- icant, pointing to a concentration of nonagricultural activities in those areas. Manufacturing activities are concentrated in and around metropoli- tan cities (but not in smaller towns and cities) and are also more prevalent in areas with better access to electricity. By contrast, services are concen- trated in smaller cities and towns, in addition to metropolitan areas. The employment patterns in urban areas are also found to be respon- sive to land inequality in surrounding rural areas.2 When the analysis is repeated for the rural subsample (61 percent of the sample), urban sub- sample (13 percent), and metropolitan subsample (26 percent), the results are similar to those for the full sample for rural areas. In urban areas, land inequality has significant nonlinear effects on self-employment in nonfarm activities only. Self-employment in the nonagriculture sector is by far the largest employment category in smaller urban cities and towns, Table 8.1 Employment Choice of Individuals, Pakistan: Marginal Effects from Multinomial Logit Regressions Self-employment Wage employment Nonagriculture Agriculture Manufacturing Service Public service Index of land 0.027 0.015 0.01 0.02 –0.013 inequality (1.71) (3.27)** (0.67) (2.73)** (2.65)** Land inequality –0.003 –0.002 –0.002 –0.002 0.002 squared (1.39) (3.28)** (1.08) (2.40)* (2.85)** Log (distance) 0.001 0.000 –0.003 –0.002 0.001 (0.84) (1.53) (2.82)** (2.76)** (1.85) Middle school and up 0.025 –0.018 –0.105 0.036 0.098 (3.15)** (6.41)** (13.53)** (8.02)** (22.39)** Primary to middle 0.077 –0.019 –0.035 0.007 0.025 school (9.25)** (7.82)** (4.62)** (1.51) (5.71)** Phone in workplace 0.254 –0.028 –0.211 0.081 –0.004 (yes = 1) (5.82)** (1.37) (5.08)** (5.19)** (0.31) Household has 0.141 –0.034 0.169 0.001 –0.048 electricity (3.36)** (2.60)** (4.27)** (0.05) (4.07)** Urban (yes = 1) 0.174 –0.022 –0.014 0.047 0.036 (8.75)** (3.58)** (0.71) (5.47)** (5.81)** Metro (yes = 1) 0.216 –0.024 0.155 0.104 0.057 (9.96)** (4.46)** (7.51)** (10.50)** (7.15)** Number of observations 45,023 45,023 45,023 45,023 45,023 Source: World Bank staff estimates based on Pakistan Labor Force Survey, 2005–06. Note: Robust t-statistics are in parentheses. All regressions included province dummies and other household- and individual-level controls. 145 *p < .05. **p < .01. 146 From Farm to Firm accounting for 42 percent of all employment. The nonlinear effect implies that self-employment in nonagricultural activities in smaller towns and cities is low when land inequality in the surrounding rural areas is small, but rises with an increase in inequality, and eventually falls as inequality becomes extreme. In metropolitan areas, a similar concave relationship between manufacturing employment and land inequality is observed. By contrast, private employment in services in metropolitan areas is not very responsive to land inequality in the surrounding rural areas. We also analyze the average growth in employment in different activ- ities from 1999 to 2005.3 The results suggest faster growth of total employment in areas near large urban centers and in areas with higher land inequality. The impact of land inequality is significantly negative only for growth in manufacturing employment. The negative impact of land inequality on this growth may seem puzzling when considered in the light of overall employment growth in the same areas. However, it is pos- sible that areas with higher land inequality may experience lower growth in demand for manufacturing goods. This result seems to suggest that large landowners are spending a large share of their income on goods pro- duced outside the local area. And what is the impact of land inequality on real wages? Our regres- sion results indicate that inequality of ownership has a significant nega- tive effect on wages in agriculture and manufacturing (table 8.2). The negative effect of land inequality on wages is expected because areas with a higher concentration of ownership of agricultural lands are likely to face an increasing supply of labor. For wage growth, the regression results sug- gest a significant negative effect of land inequality on the growth of man- ufacturing wages only. The negative effect of inequality on both manufacturing employment and wage growth may seem puzzling at first. The regression results in table 8.1 show clearly that manufacturing employs mainly uneducated and unskilled workers in semiurban and urban areas. But distance to large urban centers does not matter much for wages (except in services) once provincial dummies are included in the regressions (table 8.2). Combined, these results suggest that workers are mobile across locations, at least within the province. Thus manufacturing employment may decline locally in areas with high land inequality, but mobility of workers also ensures an overall downward pressure on wages for unskilled work- ers, particularly in manufacturing activities. Indeed, land inequality, education, and infrastructure are all important drivers of wages. Education commands a sizable premium in all types of Land Market Institutions and the Rural-Urban Transformation 147 Table 8.2 Real Wages, Pakistan: Individual-Level Instrumental Variable Regressions, Labor Force Survey, 2005–06 Log of real wages in Agriculture Manufacturing Services Index of land inequality –0.225 –0.101 –0.043 (2.30)* (5.34)** (1.31) Log (distance) –0.006 –0.003 0.011 (0.72) (1.03) (2.31)* Middle school and up 0.166 0.107 0.327 (2.13)* (4.95)** (8.64)** Phone in workplace (yes = 1) 0.198 –0.081 0.615 (0.41) (0.70) (4.63)** Urban (yes = 1) –0.025 0.118 (0.77) (2.22)* Metro (yes = 1) 0.032 0.317 (0.98) (5.90)** Number of observations 1,325 9,608 3,875 Source: World Bank staff estimates based on Pakistan Labor Force Survey, 2005–06. Note: Robust t-statistics are in parentheses. All regressions included province dummies and other household- and individual-level controls. *p < .05. **p < .01. wage work (table 8.2). The premium for an above-middle school educa- tion is greatest in service activities. Workers with education above primary school but below the middle school level command a substantial pre- mium in manufacturing and services activities. Among workers with dif- ferent levels of education, only those with a higher than secondary education experienced positive growth in wages. Wages in agriculture and services grew at a faster rate in areas with more urbanized services, such as better access to phones. But manufacturing wages declined in more urbanized settings because of the increased supply of labor from rural migrants. To the extent that investment in education and infrastructure are adversely affected by land inequality, as argued by Banerjee and Iyer (2002) and Galor, Moav, and Vollrath (2009), the regression results in tables 8.1 and 8.2 underestimate the adverse effect of land inequality. Impact of Land Market Restrictions: The Existing Evidence The restrictions on land market transactions common in developing coun- tries (World Bank 2003) have been shown to affect sectoral and spatial 148 From Farm to Firm transformation. Hayashi and Prescott (2008) find that during 1855–1940, agricultural employment remained nearly unchanged in Japan, despite very large rural-urban income disparity. They argue that the prewar patri- archy forced the designated heirs to stay in agriculture. This informal barrier to labor mobility caused misallocation of labor across activities, depressing per capita output by about a third. The introduction of parti- ble inheritance after the war led to a mass exodus from rural areas, a sharp change in employment patterns, and an increase in per capita income. Yang (1997) argues that the inalienability of land rights under the house- hold responsibility system in China increased migration costs, slowing sectoral transformation. Removal of the control in 1988 was followed by a surge in migration, although most rural-to-urban migrants were “float- ing populations” whose families remained in rural areas to retain the households’ land earnings. Overall, the existing evidence suggests that the removal of land mar- ket restrictions increases long-term investment in agriculture, improves participation in nonfarm activities, and generally allows land-poor house- holds to gain better access to land (box 8.2). Land Market Restrictions and Rural-Urban Transitions in Sri Lanka Among South Asian countries, land market restrictions are perhaps the most stringent in Sri Lanka. The Crown Lands Encroachment Ordinance of 1840 transferred all lands without private title to the state. Under the Land Development Ordinance (LDO) of 1935, the government intro- duced a system of protected tenure: private recipients of LDO land had the right to occupy and cultivate the land in perpetuity subject to restrictions imposed on sale, leasing, and mortgaging, and to conditions related to abandoning or failing to cultivate the land. Although subsequent amendments have weakened some conditions on mortgages and transfers, the basic provisions of unitary succession and a ban on the subdivision of plots and land rental remain largely intact. Restrictions on LDO leases are strict by South Asia standards, but they are similar to those found in pre-reform China, in Vietnam, and in many African countries where commons are still an important source of agricultural land. Figure 8.1 shows the incidence of LDO leases in Sri Lanka. The per- centage of land under the LDO is much lower in the urbanized Western Province than elsewhere in the country. LDO land is particularly preva- lent in North Central Province. Our empirical analysis seeks to uncover the impact of LDO restric- tions on the pace of sectoral transformation in Sri Lanka. As discussed in chapter 7, Sri Lanka implemented a broad-based economic liberalization Land Market Institutions and the Rural-Urban Transformation 149 Box 8.2 Evidence on the Effects of Land Reforms Tenancy reforms. Tenancy laws attempt to increase tenure security by registering sitting tenants and by establishing limits on the amount of rent to be paid. According to Banerjee, Gertler, and Ghatak (2002), successful implementation of such reforms in West Bengal, India, resulted in a significant increase in agricultural productivity. Looking at Pakistan, Jacoby and Mansuri (2008) report an increase in investment in land improvement measures by share tenants with an increase in tenure security. By contrast, Deininger, Jin, and Nagarajan (2008) find that in India, tenancy restrictions tend to reduce the demand for and supply of rental con- tracts, limiting the ability of the poor and landless to acquire land through such contracts. However, Deininger, Jin, and Nagarajan (2009) find that tenancy and other land reforms (including redistribution of land) in India have a positive effect on a household’s investment in human capital and the growth of income, con- sumption, and nonland assets. Liberalization of land sales and rental markets. In 1993 Vietnam distributed land use certificates known as red books. Red books came with the rights to sell, rent, mortgage, and bequeath land. They led to a vibrant land lease and sales market, allowing transfers of land to more productive users (Markussen, Tarp, and Van den Broeck 2009). Do and Iyer (2008) find that this reform had a positive and signifi- cant impact on long-term agricultural investments and on the time devoted to nonfarm activities. Meanwhile, Ravallion and van de Walle (2008) observe that land reforms increased the incidence of landlessness. Rising rural landlessness has been a benign, even positive, factor in the process of aggregate poverty reduc- tion, because farm households have taken up new opportunities in the labor market, and the landless have experienced a higher rate of poverty reduction rel- ative to the rest of the population. In China, the liberalization of land tenure arrangements increased the share of households participating in land rental arrangements from 2.3 percent to 9.4 percent in 2000. Benjamin, Brandt, and Rozelle (2000) report that the rental market shifted land to more productive and land-poor producers, with positive implications for both efficiency and equity. Land titling. Clear and secure titles to land are important for the conversion of land from one use (or user) to another. Evidence from rural land titling programs in Andra Pradesh, India, suggests that a proper title increases land value by 15 percent (World Bank 2007a). Although evidence from recent land titling pro- grams in urban slums shows that titling has little impact on households’ access to land, it demonstrates a significant positive effect on women’s labor market partic- ipation and hours of work (Galiani and Schargrodsky 2005; Field 2007). 150 From Farm to Firm Figure 8.1 Area under Land Development Ordinance (LDO) Restrictions, Sri Lanka Source: World Bank staff estimates based on agriculture census of Sri Lanka, 1998. and industrial deregulation program in 1977, almost a decade and a half earlier than India (although agriculture still receives considerable govern- ment support in the form of input subsidies and trade protection). The country is also renowned for its equitable provision of education, health care, and other social services to its citizens regardless of their location. Yet in spite of this largely favorable policy environment, the pace of sec- toral transformation has been relatively slow. In 1960 agriculture’s share of total employment was about 56 percent in Sri Lanka, 60 percent in the Republic of Korea, and 78 percent in Indonesia. Today, those shares stand at 44 percent, 9 percent, and 44 percent, respectively (figure 8.2). Even Land Market Institutions and the Rural-Urban Transformation 151 Figure 8.2 Agriculture’s Share in Total Employment, Selected Countries in Asia 80 70 60 50 percent 40 30 20 10 0 1961 1966 1971 1976 1981 1986 1991 1996 2001 2006 India Indonesia Korea, Rep. Sri Lanka Source: World Bank Development Data Platform (DDP), http://go.worldbank.org/DNBRRS9TB0. India experienced a faster decline in agriculture’s share of employment and income during this period. Comparatively, the regional pattern in Sri Lanka indicates a higher level of sectoral transformation in Western Province, the wealthiest of the provinces, and a much higher dependence on agriculture in other provinces (table 8.3). In the wealthy western region, 44 percent of the population is employed in services and 26 percent in manufacturing. Only 12 percent is employed in agriculture (including both farmers and wage laborers). At the other end of the spectrum, in Uva Province, the poorest of the provinces, 34 percent are farmers, and another 30 percent are agricultural wage laborers. Manufacturing employment accounts for only 6 percent of the provincial labor force. Using a regression-based model, we examined the impact of LDO restrictions on sectoral employment and wages in Sri Lanka. The econo- metric details are reported in a background paper by Shilpi (2010). This section summarizes the main results. The empirical analysis indicates that LDO restrictions have slowed the pace of transition out of agriculture. Estimates from the regression model show that the percentage of area under LDO leases has had a significant negative impact on the probability of participation in all types of non- agricultural employment (table 8.4). The direct effect of distance to the 152 From Farm to Firm Table 8.3 Sectoral Employment Shares in Sri Lanka, 2002 Percentage of total employment Self-employment Wage employment Province Agriculture Nonfarm Agriculture Manufacturing Services Western 8 18 4 26 44 Central 15 12 29 12 32 Southern 21 15 20 18 26 North Western 21 16 11 21 31 North Central 33 13 18 8 27 Uva 34 8 30 6 22 Sabaragamuwa 18 13 20 17 32 Sri Lanka 18 15 16 18 34 Source: World Bank staff estimates based on Household Income and Expenditure Survey, 2002. Table 8.4 Effects of the Land Development Ordinance (LDO) on Employment Choice, Sri Lanka Employment in Self-employment, Wage Wage Wage nonagriculture agriculture manufacturing services Percent area under LDO –0.183 0.238 –0.162 –0.224 (3.23)*** (3.91)*** (2.40)** (2.29)** Travel time to large city –0.004 0.014 –0.018 –0.007 (1.03) (4.22)*** (4.00)*** (1.42) Area LDO*travel time 0.027 –0.029 0.047 –0.037 (1.75)* (1.62) (2.17)** (1.42) Education level (year) 0.002 0.013 0.039 –0.018 (0.31) (1.70) (4.56)*** (1.59) Number of observations 24,123 24,123 2,4123 24,123 Source: World Bank staff estimates based on Household Income and Expenditure Survey, 2002. Note: Robust z-statistics are in parentheses. All regressions include a large number of individual, household, and location controls and a district-level fixed effect. *p < .10. **p < .05. ***p < .01. nearest large city is negative and significant for nonfarm self-employment and wage employment in manufacturing. The interaction of LDO and distance is also statistically significant for all activities except services (table 8.4). For wage employment in manu- facturing and self-employment in nonfarm enterprises, the probability of participation in these activities declines with the increase in land under Land Market Institutions and the Rural-Urban Transformation 153 LDO restrictions. The interaction effect is positive. This finding implies that the impact of the LDO is more negative in locations that have bet- ter market access; in more remote places, LDO restrictions have a much smaller adverse effect. In contrast with nonfarm employment, the effects of LDO restrictions and distance are positive for agricultural wage labor. The result implies that relative to self-employment in farming, participation in agricultural wage labor is higher in areas with higher LDO restrictions. The higher incidence of agricultural wage labor probably stems from the lease restric- tions on LDO land, which impede those residents with larger farms from leasing land out and force those with inadequate land to turn to agricul- tural labor for supplementary income. In addition to keeping a larger proportion of people dependent on agri- culture, land market restrictions tend to keep them poorer because they earn less per unit of labor (table 8.5). The coefficient of land under the LDO is negative and statistically highly significant for agricultural and services wages. For manufacturing wages, it is negative but statistically sig- nificant at the 10 percent level. The magnitude of the estimated coeffi- cients implies the largest negative effect of the LDO on services wages, followed by agricultural and then manufacturing wages. The results sug- gest that the interaction of the proportion of area under the LDO and Table 8.5 Market Access, Land Tenure Arrangement, and Income, Sri Lanka Log (real annual wage) Agriculture Manufacturing Services Percent area under LDO –3.213 –1.304 –9.000 (4.79)*** (1.90)* (2.50)** Travel time to large city –0.292 –0.067 –0.057 (6.26)*** (1.60) (1.00) Area LDO*travel time 0.867 0.501 3.034 (4.42)*** (2.12)** (2.36)** Education level (year) 0.009 0.047 0.076 (2.54)** (14.53)*** (21.49)*** Number of observations 3,350 3,804 7,444 Source: World Bank staff estimates based on Household Income and Expenditure Survey, 2002. Note: Robust z-statistics are in parentheses. All regressions include a large number of individual, household, and location controls and a district-level fixed effect. *p < .10. **p < .05. ***p < .01. 154 From Farm to Firm travel time to a large city has a positive and statistically significant effect on all wages (table 8.5). The effect of the share of land under the LDO thus depends on access to large urban centers and vice versa. This inter- action effect implies a larger negative effect of the LDO in locations with better access to urban centers. Wages are lower in a location with a higher proportion of land under the LDO compared with a location that is equi- distant from an urban center but has a lower proportion of land under the LDO. Our empirical analysis thus suggests that employment in locations with severe land regulations is less diversified toward nonfarm activities in manufacturing and services, and that those employed in nonfarm labor earn much less on average. These results are quite remarkable because we already control for much of the agglomeration externalities related to urbanization and increased density through district-level fixed effects. The regressions also control for differences in service provision across districts. As in Pakistan, our empirical results indicate that level of education is an important determinant of employment structure and wages in Sri Lanka. According to table 8.5, education yields a significant premium in services and manufacturing. The employment choice estimation reported in table 8.4 indicates that higher education significantly improves the probability of securing a job in the manufacturing sector. Surprisingly, higher education has little effect on securing a job in services, although its return is highest in service activities. As in most developing countries, service activities in Sri Lanka comprise both low- and high-skill jobs. Low-skill service jobs seem to be more prevalent than high-skill ones, causing access to service jobs to be nonresponsive to education level on average. Summary of Case Studies and Policy Implications The empirical analysis for Pakistan demonstrates that inequality in the distribution of land has forced a large portion of the population to seek off-farm employment, thereby depressing wages, particularly in agricul- ture and manufacturing. The effect is not confined to rural areas. Self- employment in smaller cities and towns and manufacturing in metropolitan areas are also affected. The adverse effect on wages stifles savings, investment, and demand, forcing more and more workers into subsistence (self-employment) work. The impact of land inequality on the growth of employment and wages suggests an exodus of workers from areas high in land inequality to urban areas, where they are absorbed into Land Market Institutions and the Rural-Urban Transformation 155 self-employment (in smaller cities and towns) and unskilled jobs in man- ufacturing (in metropolitan areas). As a result, land inequality has a sig- nificant negative effect on real wage growth in manufacturing. Our results suggest that by improving the access of the poor to land, Pakistan can ensure smoother and faster sectoral transformation and a significant reduction in income inequality. Three major attempts at redis- tributive land reform in Pakistan have failed, the most recent in 1977. Land reform has neither political support nor the backing of Islamic religious authorities. As noted by a World Bank report, any attempt to enact redistributive land reform would have to create a win-win situation (World Bank 2007b). Policy measures to increase access to land could include: greater access to credit to enable the poor to purchase land, land taxation to minimize the holding of land for speculative purposes, and measures to improve the efficiency of land sales and rental markets. Land purchase schemes that include grant components for the poorest landless households is another option, although the fiscal costs could limit the scale of such an operation. The empirical analysis for Sri Lanka indicates that an increase in the incidence of LDO leases is associated with lower nonfarm employment diversification. Areas with a higher percentage of land under the LDO have disproportionately more people dependent on agricultural wage labor. The finding that nonfarm enterprise income also declines with an increase in land under the LDO points to the negative effect of mortgage restrictions on the development of credit markets. The limited expansion of nonfarm enterprises in these areas means that wages for all types of workers have been depressed in areas with a higher incidence of LDO leases. Thus, although LDO leases have created a middle-class peasantry in Sri Lanka, they seem to have lowered the income prospects for agricul- tural workers, who are among the poorest in rural areas. Because LDO leases are quite secure, they should not depress long-term investment in land. The rental restrictions can nevertheless create produc- tivity inefficiency if they restrict the transfer of land from inefficient to efficient producers, although, in practice, informal leases are quite preva- lent in LDO areas (PSIA 2008). Moreover, any productive inefficiency created by the LDO restrictions in agriculture has been mitigated by mas- sive public investment in irrigation schemes and other infrastructure development projects. As a result of these generous investments, today land productivity in many settlement areas is much higher than that in the rest of the country. However, sales restrictions are found to reduce the value of LDO lands by 15–25 percent (PSIA 2008). Relaxing leasing and 156 From Farm to Firm sales restrictions on land under the LDO is likely to have positive effects on rural incomes, poverty reduction, and longer-term structural transfor- mation in Sri Lanka. Finally, in both countries land reforms would have to be comple- mented with investment in human capital, because higher levels of edu- cation would help the poor to secure better nonfarm employment. In the medium to long run, the modernization of Pakistan and Sri Lanka’s man- ufacturing and services sectors will depend on the availability of a highly skilled labor force. Notes 1. This pattern of private employment is consistent with the prediction of Banerjee and Newman (1993). 2. These results are provided in the background paper by Shilpi (2009). 3. These results are omitted for the sake of brevity. References Adams, R., Jr., and H. Alderman. 1992. “Sources of Income Inequality in Rural Pakistan: A Decomposition Analysis.” Oxford Bulletin of Economics and Statistics 54 (4): 591–608. Banerjee A., P. Gertler, and M. Ghatak. 2002. “Empowerment and Efficiency: Tenancy Reform in West Bengal.” Journal of Political Economy 110 (2): 239–80. Banerjee, A., and L. Iyer. 2002. “History, Institutions, and Economic Performance: The Legacy of Colonial Land Tenure Systems in India.” American Economic Review 95 (4): 1190–213. Banerjee, A. V., and A. F. Newman. 1993. “Occupational Choice and the Process of Development.” Journal of Political Economy 101 (2): 274–98. Benjamin, D., L. Brandt, and S. Rozelle. 2000. “Aging, Well-being, and Social Security in Rural North China.” Population and Development Review 26: 89–116. Deininger, K., S. Jin, and H. K. Nagarajan. 2008. “Efficiency and Equity Impacts of Rural Land Rental Restrictions: Evidence from India.” European Economic Review 52 (5): 892–918. ———. 2009. “Land Reforms, Poverty Reduction, and Economic Growth: Evidence from India.” Journal of Development Studies 45 (4): 496–521. Do, Q.-T., and L. Iyer. 2008. “Land Titling and Rural Transition in Vietnam.” Economic Development and Cultural Change 56 (April). Land Market Institutions and the Rural-Urban Transformation 157 Field, E. 2007. “Entitled to Work: Urban Property Rights and the Labor Supply in Peru.” Quarterly Journal of Economics 122 (4): 1561–602. Galiani, S., and E. Schargrodsky. 2005. “Property Rights for the Poor: Effects of Land Titling.” Business School Working Papers, Universidad Torcuato Di Tella, Buenos Aires. Galor, O., O. Moav, and D. Vollrath. 2009. “Inequality in Landownership, the Emergence of Human-Capital Promoting Institutions, and the Great Divergence.” Review of Economic Studies 76 (1): 143–79. Hayashi, F., and E. C. Prescott. 2008. “The Depressing Effect of Agricultural Institutions on the Prewar Japanese Economy.” Journal of Political Economy 116 (4): 573–632. Jacoby, G., and G. Mansuri. 2008. “Land Tenancy and Non-Contractible Investment in Rural Pakistan.” Review of Economic Studies 75 (3): 763–88. Markussen, T., F. Tarp, and K. Van den Broeck. 2009. “The Forgotten Property Rights: Restrictions on Land Use in Vietnam.” Discussion Paper 09-21, Department of Economics, University of Copenhagen. Naschold, F. 2009. “Microeconomic Determinants of Income Inequality in Rural Pakistan.” Journal of Development Studies 45 (5): 746–68. PSIA (Poverty and Social Impact Analysis). 2008. Land Reforms in Sri Lanka: A Poverty and Social Impact Analysis. Washington, DC: World Bank. Ravallion, M., and D. van de Walle. 2008. “Does Rising Landlessness Signal Success or Failure for Vietnam’s Agrarian Transition?” Journal of Development Economics 87 (2): 191–209. Shilpi, F. 2009. “Land Inequality and Occupation Choice: Evidence from Pakistan.” World Bank, Washington, DC. ———. 2010. “The Effect of Land Market Restrictions on Employment Pattern and Wages: Evidence from Sri Lanka.” World Bank, Washington, DC. World Bank. 2003. Land Policies for Growth and Poverty Reduction. New York: Oxford University Press. ———. 2007a. India: Land Policies for Growth and Poverty Reduction. New Delhi: Oxford University Press. ———. 2007b. “Pakistan: Promoting Rural Growth and Poverty Reduction.” South Asia Region, Sustainable and Development Unit, Report 39303-Pk, World Bank, Washington, DC. ———. 2007c. World Development Report 2008: Agriculture for Development. Washington, DC: World Bank. Yang, D. 1997. “China’s Land Arrangements and Rural Labor Mobility.” China Economic Review 8 (2): 101–15. CHAPTER 9 Geographical Linkages and the Rural-Urban Transformation As described in chapter 8, land market policies intended for rural areas have implications for the urban labor market. In this chapter, we address the issue of geographical spillovers along the entire rural-urban contin- uum by considering at three principal questions. First, how important are small towns and cities in the national economic landscape? Second, look- ing especially at Bangladesh, does urbanization affect employment pat- terns in rural areas? And, third, what areas—metropolitan, periurban, smaller cities, or rural—are more attractive to emerging firms? Poverty, Urbanization, and Employment Structure along the Rural-Urban Continuum In most developing countries, the rural-urban differences in economic structure and poverty incidence are well documented, but evidence on the economic structure of cities of different sizes is still very sparse. Here we provide evidence on poverty, population, and employment patterns along the rural-urban continuum in selected South Asian countries. Evidence from these countries reveals that the incidence of poverty is highest in rural areas, followed by smaller towns and cities, and lowest in metropolitan areas. For example, the poverty incidence in Bangladesh is 159 160 From Farm to Firm about 26 percent in metropolitan areas, about 38 percent in periurban areas and smaller cities, and 43 percent in rural areas (figure 9.1).1 A similar pattern is observed in Pakistan (figure 9.2). In India, 43 per- cent of people residing in smaller towns are poor compared with 32 per- cent in medium-size cities and 20 percent in metropolitan cities. Although poverty rates are lower in the larger metropolitan cities, they Figure 9.1 Poverty Headcount Ratio: Bangladesh, 2004–05 50 45 40 35 30 percent 25 20 15 10 5 0 metro area periurban area towns rural area Source: World Bank staff estimates based on Household Income and Expenditure Survey, 2004–05. Figure 9.2 Poverty Headcount Ratio: Pakistan, 2004–05 40 35 30 25 percent 20 15 10 5 0 Punjab Sindh NWFP Balochistan province rural urban metro cities Source: World Bank staff estimates based on Pakistan Social and Living Standard Measurement Survey, 2004–05. Note: NWFP = North West Frontier Province. Geographical Linkages and the Rural-Urban Transformation 161 are typically home to high inequalities. For example, child and infant mortality rates are highest among the poor in large cities in India (World Bank 2011). Differences in access to services mirror poverty incidence. The per- centage of households with access to electricity for lighting is highest in metropolitan areas in both Bangladesh and Pakistan (figures 9.3 and 9.4). In Bangladesh, 76 percent of households in metropolitan areas have access to electricity compared with only one-third of households in peri- urban areas and smaller towns.2 Although the disparity in electricity access between metropolitan cities and smaller towns in Pakistan (“urban” in figure 9.4) is much smaller, substantial disparities remain in access to piped water and telephones. More than 200 cities and towns dot Bangladesh and Pakistan. And yet urbanization in both of these countries is dominated by a few large met- ropolitan cities. Two metropolitan cities with a population of more than 1 million—Dhaka and Chittagong—account for 43 percent of Bangladesh’s urban population. In Pakistan, eight cities with a population of more than 1 million account for 58 percent of the urban population. Another 24 percent of the urban population in Pakistan resides in 48 cities with populations of 100,000 to 1 million.3 The pattern of urbanization in South Asian countries indicates two stylized facts. First, rural areas continue to account for a large share of Figure 9.3 Access to Services: Bangladesh, 2000 80 70 60 50 percent 40 30 20 10 0 metro periurban towns rural electricity piped water Source: Staff estimate based on population census of Bangladesh, 2000. 162 From Farm to Firm Figure 9.4 Access to Services: Pakistan, 2004–05 100 90 80 70 60 percent 50 40 30 20 10 0 metro area urban area rural area electricity piped water phone Source: World Bank staff estimates based on Pakistan Social and Living Standard Measurement Survey, 2004–05. Figure 9.5 Percentage of Urban Population in Small, Medium, and Large Cities, Bangladesh and Pakistan 70 60 50 40 percent 30 20 10 0 small cities medium cities large cities Bangladesh Pakistan Source: World Bank staff estimates based on population censuses. population and therefore employment, even in the nonfarm sector. For example, nearly half of all manufacturing activities in each country are located in the rural areas of Bangladesh, Nepal, and Pakistan. Thus employment transition in rural areas will remain critical for the overall rural-urban transformation. Second, consistent with the weaker provision Geographical Linkages and the Rural-Urban Transformation 163 of services, smaller towns and cities are not always able to attract manu- facturing activities. Rural or periurban areas in close proximity to larger urban markets may be better placed strategically to attract manufactur- ing activities. These two issues are analyzed in detail in the following case studies using data from Bangladesh, where the growth of the nonfarm sector is particularly critical to national development. At 1,148 persons per square kilometer, the population density in Bangladesh is among the highest in the world. With the margin of cultivation almost exhausted and cropping intensity approaching its physical limit, agriculture is unable to offer addi- tional employment opportunities. The bulk of the 1 million or so annual new entrants into the labor force will have to be absorbed into the non- farm activities in both rural and urban areas. The first case study presents evidence on the relative importance of farm and urban linkages to the growth of rural nonfarm employment. The second case study analyzes the relative importance of agglomeration economies, infrastructure provision, and spatial spillovers in determining the location decisions of manufacturing firms across the rural–urban continuum. Case Study 1: The Rural Employment Pattern and Urban Linkages The existing evidence highlights two observations about the pattern of nonfarm development in rural areas of developing countries. First, the growth of nonfarm activities is often driven by growth in agricultural pro- ductivity, at least at the initial stage (Haggblade, Hazell, and Dorosh 2006), because of production, consumption, and labor market linkages between the farm and nonfarm sectors. Second, a large share of nonfarm activities is located in and around towns and cities (Renkow 2006) because of the demand and supply side interlinkages between rural and urban areas (von Thünen 1966; Fuijta, Krugman, and Venables 1999). A background study undertaken for this chapter considers the relative strength of farm and urban linkages in determining nonfarm employment in Bangladesh (Deichmann, Shilpi, and Vakis 2009). Here we summarize the main findings. Proximity to large cities is an important determinant of the nature of nonfarm activities in a region. The probability of employment in high- return wage work is greater the closer a person is to a large metropolitan city such as Dhaka or Chittagong (table 9.1). The probability of self- employment is lower for those people living farther from urban centers, 164 From Farm to Firm Table 9.1 Urban Access, Cash Crop Potential, and Nonfarm Employment in Bangladesh: Multinomial Logit Results Regression Wage employment Dependent variable = type of nonfarm occupation Low return High return Self-employment Distance to major cities (log) 0.01 –0.04 –0.015 (1.44) (2.71)** (1.47) Distance*crop suitability –0.006 –0.006 –0.008 (4.64)** (1.93) (3.58)** Distance to towns (km) 0.001 –0.001 0.001 (0.73) (1.41) (1.92) Source: Deichmann, Shilpi, and Vakis 2009. Note: All regressions include individual characteristics (gender, education, age), household characteristics (size, landownership, asset ownership, electricity connection), village characteristics (whether village has electricity, nongovernmental organization programs, credit programs, markets), distance to town, and an intercept term as regressors. Robust z-statistics are in parentheses. *p < .05. **p < .01. although the estimated effect is not statistically significant. In contrast to the important influence of metropolitan areas, access to smaller rural towns (with populations of about 5,000) exerts little influence on non- farm activities. The effect of agricultural potential also depends on how far a village is from urban centers. The results suggest that better-paying wage employ- ment and self-employment in the nonfarm sector are more likely for those living closer to urban centers. Those who are living farther away from such centers are less likely than those living closer to be in well-paid nonfarm jobs if they are living in areas with greater agricultural potential. Low-return wage work, which pays less than or equal to the median agricultural wage of a region, appears to have no significant relationship to access to urban centers. These jobs seem to cater to local demand and are distributed almost evenly over geographical space. Lack of connectivity is not only likely to depress growth in agricultural productivity and diversification in areas with higher agricultural potential but also to discourage the growth of higher-paying nonfarm activities in those areas. The empirical results highlight the need for improved con- nectivity in Bangladesh between regions with higher agricultural poten- tial (such as the northwestern region) and urban centers to stimulate growth in high-return wage employment and self-employment in non- farm activities. In the second case study, we further examine the pattern of location of nonfarm activities over the entire rural–urban continuum. Geographical Linkages and the Rural-Urban Transformation 165 Case Study 2: Agglomeration, Spillover, and Location of Nonfarm Activities Nonfarm activities in developing countries are often clustered in a few locations because of either the natural endowments or the agglomeration economies available at those locations (Krugman 1991; Henderson 1997; Fujita, Krugman, and Venables 1999). Agglomeration economies arise through knowledge and information sharing, economies of scale, and better matching of workers and jobs. Concentration also increases competition, incentivizing firms to make productivity-enhancing invest- ments. However, a greater concentration of activities in an area can lead to congestion, pollution, and transport and services bottlenecks. Without ade- quate investment in infrastructure and service provision, the diseconomies from congestion can more than offset agglomeration economies, choking off the further growth of nonfarm activities in that area. The regression results presented in the preceding section suggested a concentration of rural nonfarm activities in proximity to large urban centers—the Dhaka and Chittagong metropolitan areas—in Bangladesh. In this section, in considering the location pattern of nonfarm activities across the rural–urban continuum, we classify areas according to four cat- egories: (1) metropolitan areas (Dhaka, Chittagong, Rajshahi, Khulna, Barisal, and Sylhet); (2) periurban areas consisting of the upazillas (sub- districts) bordering the six metropolitan cities; (3) small towns—the dis- trict headquarters (sadar upazillas); and (4) rural areas—the rest of the country. As noted, these four areas differ markedly in terms of their con- nectivity to major markets and access to services.4 The empirical analysis focuses on firms that employ 10 or more peo- ple. These firms account for more than a third of total nonfarm employ- ment in Bangladesh. The existing evidence indicates that nonfarm activities are highly local- ized in metropolitan and periurban areas (figure 9.6). According to the 2006 economic census, metropolitan areas account for 43 percent of total employment in firms of 10 or more, followed by rural areas (24 percent) and periurban areas (23 percent). Even within metropolitan areas, activi- ties are clustered in Dhaka and Chittagong. Dhaka alone accounts for 72 percent of total employment and 81 percent of total firms located in metropolitan areas, with Chittagong a distant second (23 percent of employment and 12 percent of firms). However, diseconomies of congestion have led to a decline in number of nonfarm firms in metropolitan areas as firms have moved to periurban and rural areas. Between 2000 and 2006, employment in metropolitan 166 From Farm to Firm Figure 9.6 Shares of Nonfarm Employment and Enterprises by Area: Bangladesh, 2006 50 45 40 35 30 percent 25 20 15 10 5 0 metro periurban urban rural employment enterprise Source: World Bank staff estimates based on economic census of Bangladesh, 2006. areas remained flat, and the number of firms declined by an annual rate of 2.5 percent (figure 9.7). This decline in the number of firms in metro- politan areas is not surprising in view of the high land prices, congestion, and pollution in those cities, particularly Dhaka. In contrast with metro- politan areas, rural and periurban areas experienced substantial increases in both employment and number of firms (5.2 percent and 1.8 percent, respectively, in rural areas and 4.4 percent and 2.5 percent in periurban areas). Small towns, by contrast, experienced only slight growth in employment and number of firms. The estimates of Ellison and Glaeser’s (1997) concentration index show that only two industries—textile and wearing apparel—display a high degree of concentration in Bangladesh.5 These industries are, how- ever, the two most important in terms of their contribution to nonfarm employment. Manufacturing of food and beverages, tobacco, and non- metallic mineral products display intermediate levels of concentration. The rest of the two-digit industries show very low levels of concentration, and hotels and restaurants are the least-concentrated activity in Bangladesh. To ascertain the relative importance of infrastructure provision and agglomeration forces in determining the dynamics of the clustering of nonfarm activities, we analyze the location decisions of manufacturing Geographical Linkages and the Rural-Urban Transformation 167 Figure 9.7 Annual Percentage Growth in Nonfarm Employment and Enterprises: Bangladesh, 2000–06 6 5 4 percentage per year 3 2 1 0 –1 –2 –3 metro periurban urban rural employment enterprise Source: World Bank staff estimates based on economic censuses of Bangladesh, 2000 and 2006. start-ups using unit record data from the 2000 and 2006 economic cen- suses. The analysis has been carried out at the two-digit industry and upazilla levels.6 The empirical methodology is described in the back- ground paper by Shilpi (2009). Here we summarize the main results. Urbanization, as measured by population density and the percentage of urban households in a location, has a significant and positive role in attract- ing start-ups to a location (table 9.2). Because the regression included dummies to capture any systematic differences in investment climate con- straints across regions, the percentage of urban households in a location can be taken as a summary measure of the availability of urban services such as sanitation, garbage disposal, and infrastructure. These results there- fore indicate that firms prefer locations with better urban services. The regression results in table 9.2 reveal that the degree of competition measured by the number of firms per capita in an activity at a given loca- tion has a consistently positive and statistically significant effect on firm start-ups in all four regions. These results indicate that competition attracts more productive entrants to a location that already has a concentration of that activity, which is consistent with the observations of Porter (1990). But the interaction of the competition index with firm size has a negative 168 From Farm to Firm Table 9.2 Agglomeration Economies and Start-up Firms’ Location, Bangladesh Metropolitan Periurban Rural areas areas Towns areas Urbanization economies Log (population) 0.301 0.301 0.301 0.301 (3.81)** (3.81)** (3.81)** (3.81)** % of household urban 0.008 0.008 0.008 0.008 (5.67)** (5.67)** (5.67)** (5.67)** Agglomeration economies Specialization index (SI) 1.64 1.29 0.86 2.829 (1.32) (2.79)** (1.44) (9.17)** SI*firm size (FS) 0.03 0.03 0.006 0.004 (1.98)* (8.61)** (1.66) (3.10)** Competition index (CI) 0.005 0.003 0.003 0.001 (11.36)** (31.55)** (10.71)** (52.66)** CI*firm size (FS) 2.0E–05 –1.0E–05 –5.0E–05 –4.5E–06 (3.85)** (8.29)** (3.72)** (10.47)** Diversity index (DI) 0.24 0.28 –0.02 0.025 (4.03)** (8.29)** (1.10) (2.03)* DI*firm size (FS) 0.003 –8.0E–04 2.0E–04 7.8E–5 (0.75) (1.84) (1.04) (0.88) Source: World Bank staff estimates based on economic census of Bangladesh, 2006. *p < .05. **p < .01. sign for all areas except metropolitan. Thus larger firms prefer locations with a relatively lower density of competitors in the case of periurban areas, rural areas, and smaller towns. Specialization economies are measured by a location’s share of total employment for an activity. The results indicate that start-ups in peri- urban and rural areas seek out locations with an existing clustering of sim- ilar activities. In these areas, the positive externality from locating near similar firms outweighs congestion costs. But in metropolitan areas, con- gestion costs seem to neutralize the positive externalities from specializa- tion; the regression results suggest no statistically significant effect of specialization on the entry of firms. Not surprisingly, specialization has only a very weak effect in attracting new entrants in smaller towns, which are typically home to a range of activities. The geographical concentration of various types of activities in a loca- tion allows knowledge transfers across firms and makes complementary services available. Our regression results show that the diversity index (measured by the inverse of log of the Herfindahl index of employment Geographical Linkages and the Rural-Urban Transformation 169 concentration) has a positive effect on an entrant’s location choice only in metropolitan and periurban areas. The regression results suggest positive spillovers from the major met- ropolitan growth poles of Dhaka and Chittagong (table 9.3). The market size in growth poles is measured by the total population of the closest growth pole divided by the distance from an upazilla to that growth pole. The influence of market size of the growth poles is positive and statisti- cally significant in all areas except for small towns. Specialization in growth poles has a positive and significant effect on firm entry in metro- politan and periurban areas, and on the entry of larger firms in rural areas. Firms are more likely to choose a location that has better access to mar- kets and firms in growth poles. By contrast, the regression results indicate that, in medium-size cities, access to markets and specialization have only a weak effect on firm entry. Larger firms are more likely to start up in small towns if the nearest medium-size town has a larger market size and a concentration of firms Table 9.3 Spillover and Start-up Firms’ Location, Bangladesh Metropolitan Periurban areas areas Towns Rural areas Spillover from growth poles Market size (MS) 2.90E–06 1.50E–06 1.20E–07 1.62E–06 (13.34)** (9.87)** (0.55) (10.67)** Specialization (SP) 9.95 6.17 –3.88 3.29 (7.27)** (5.08)** (1.41) (1.71) MS*firm size (FS) 1.3–E–11 3.60E–09 9.60E–10 3.23E–09 (0.01) (3.86)** (0.62) (3.39)** SP*FS –0.035 –0.012 –0.0026 –0.01 (2.60)** (0.85) (0.07) (0.46) Spillover from city of size >100,000 Market size (MS) 2.28E–06 1.81E–06 –1.56E–06 –1.50E–06 (3.80)** (5.56)** (1.83) (0.67) Specialization (SP) 7.55 2.27 2.56 34.48 (2.51)* (1.42) (0.90) (2.51)* MS*firm size (FS) –2.73E–09 –2.33E–09 6.85E–08 –5.00E–08 (0.37) (0.91) (2.80)** (1.52) SP*FS –0.09 –0.035 0.3 0.02 (1.11) (4.39)** (3.62)** (0.27) Source: World Bank staff estimates based on economic census of Bangladesh, 2006. *p < .05. **p < .01. 170 From Farm to Firm in the same activity. And market size in the nearest medium-size town appears to have a positive effect on start-ups in metropolitan and periur- ban areas. Specialization in the nearest medium-size cities encourages entry into rural and metropolitan areas, but it discourages entry by larger firms in periurban areas. These results suggest that specialization in medium-size cities positively influences firm start-ups in small towns and rural areas. But medium-size towns also pull some large firms away from periurban areas. Agglomeration economies have a significant positive effect on the pro- ductivity of firms located outside of metropolitan areas. The regression based on the Non-Metro Investment Climate Survey (NMICA 2007) shows that the index of specialization has a statistically significant and positive effect on productivity. Similar to our findings for the enterprise start-ups, competition has a statistically significant and positive effect on productivity (table 9.4). The index representing diversity has no signifi- cant impact on productivity. The proximity to markets in growth poles influences the productivity of enterprises significantly and positively. By contrast, specialization in growth poles and in the closest medium-size towns seems to affect productivity negatively. When the effects of all different types of specialization variables are evaluated at their mean val- ues, the regression coefficients imply a positive influence of specialization on firm productivity. By survey design, the selected locations in the Table 9.4 Productivity and Spatial Externality in Bangladesh: Regression Results from Non-Metro Investment Climate Survey Sample Coefficient t-statistic Specialization index 4.125 (2.62)** Competition index 0.001 (3.23)** Diversity index 0.055 (0.28) Market size of nearest growth pole 6.86E–07 (4.12)** Specialization of nearest growth pole –1.502 (2.15)* Market size of nearest city of population >100,000 1.43E–07 (0.91) Specialization of nearest city of population >100,000 –4.353 (2.60)** Log (area) 0.121 (1.37) Log (population) 0.128 (1.31) % urban 0.002 (1.04) Source: World Bank staff estimates based on Non-Metro Investment Climate Survey, 2008. *p < .05. **p < .01. Geographical Linkages and the Rural-Urban Transformation 171 NMICA 2007 survey are close to urban markets, which may explain why the regression indicates the absence of a substantial effect of own popu- lation on firm-level productivity. Overall, our results suggest that the growth of nonfarm activities, par- ticularly manufacturing activities, outside the two main metropolitan areas of Dhaka and Chittagong will be important for broad-based growth and structural transformation in Bangladesh. Although these two cities manifest the highest level of clustering in the country, according to the regression results the congestion costs already outweigh the benefits of agglomeration. These results are consistent with the evidence of extremely high real estate prices, severe traffic congestion, pollution, and lack of basic urban services in these cities. Although investment in urban services and reforms in urban management may improve the overall investment climate in metropolitan areas, they are unlikely to relax the land constraint imposed by rivers, particularly in the case of Dhaka.7 Compared with Dhaka and Chittagong, other areas—particularly peri- urban and rural areas close to metropolitan cities—offer many advan- tages: cheaper land and labor and little or less congestion. The provision of urban services and maintenance of basic infrastructure in periurban areas would enable further clustering of nonfarm activities. Smaller towns and cities fare poorly in terms of accessibility to larger urban markets and the provision of services. Whereas periurban areas are within 45 kilometers of four major metropolitan areas (Dhaka, Chittagong, Khulna, and Rajshahi), the smaller towns are on average 160 kilometers away and rural areas 190 kilometers away. The provision of services in smaller towns is comparable to that in periurban areas, and yet smaller towns suffer from lack of connectivity. As a result, these towns and cities have not been able to attract enough specialized activities. Connecting these locations to markets can help the clustering of firms, allowing agglomeration economies to take hold. Notes 1. The definitions of metropolitan versus smaller cities and towns differ across countries. In Bangladesh, the six administratively defined metropolitan cities are included in the metro sample, whereas in Pakistan, the 14 largest cities are considered metropolitan areas (populations of more than 300,000). 2. Similarly, there are disparities in access to electricity across cities of different sizes in India, although the difference is not as stark as in Bangladesh (World Bank 2008). 172 From Farm to Firm 3. In contrast to Bangladesh and Pakistan, more than 70 percent of India’s urban population lives in medium-size and smaller cities and towns (population of less than 1 million). 4. Bangladesh is administratively organized into 64 districts and more than 480 upazillas (subdistricts). 5. The Ellison and Glaeser (1997) concentration index was estimated for all International system of Industry Classification (ISIC) two-digit industries using economic census data. 6. Bangladesh has 484 upazillas. Each of the areas specified here—metropolitan cities, periurban areas, smaller towns and cities, and rural areas—consists of many upazillas. 7. Expansion of Dhaka city is constrained by the presence of several rivers to the east and south of the city. References Deichmann, U., F. Shilpi, and R. Vakis. 2009. “Urban Proximity, Agricultural Potential and Rural Non-farm Employment: Evidence from Bangladesh.” World Development 37 (3): 645–60. Ellison, G., and E. L. Glaeser. 1997. “Geographic Concentration in U.S. Manufacturing Industries: A Dartboard Approach.” Journal of Political Economy 105: 889–927. Fujita, M., P. Krugman, and A. J. Venables. 1999. The Spatial Economy: Cities, Regions and International Trade. Cambridge, MA: MIT Press. Haggblade, S., P. Hazell, and P. Dorosh. 2006. “Sectoral Growth Linkages between Agriculture and the Rural Nonfarm Economy.” In Transforming the Rural Nonfarm Economy, ed. S. Haggblade, P. Hazell, and T. Reardon, chap. 7. Baltimore: Johns Hopkins University Press. Henderson, J. V. 1997. “Externalities and Industrial Development.” Journal of Urban Economics 42: 449–70. Krugman, P. 1991. “Increasing Returns and Economic Geography.” Journal of Political Economy 99: 483–99. Porter, M. 1990. The Competitive Advantage of Nations. New York: Free Press. Renkow, M. 2006. “Cities, Towns and the Rural Nonfarm Economy.” In Transforming the Rural Nonfarm Economy, ed. S. Haggblade, P. Hazell, and T. Reardon, 183–98. Baltimore: Johns Hopkins University Press. Shilpi, F. 2009. “Agglomeration and Location Choice of Rural Enterprises: Evidence from Bangladesh.” World Bank, Washington, DC. von Thünen, J. H. 1966. Isolated State: An English Edition of Der Isolierte Staat. Oxford, U.K.: Pergamon Press. Geographical Linkages and the Rural-Urban Transformation 173 World Bank. 2008. “Harnessing Competitiveness for Stronger Inclusive Growth: Bangladesh Second Investment Climate Assessment.” Report No. 56635, Washington, DC. ———. 2011. Persepectives on Poverty in India: Stylized Facts from Survey Data. Washington, DC: World Bank. CHAPTER 10 Labor Mobility and the Rural-Urban Transformation Chapter 8 examined the impacts of agrarian institutions on the transition from agricultural to nonagricultural activities. Chapter 9 then considered two more dimensions of the rural-urban transformation: sectoral transfor- mation in rural space and the determinants of firms’ decisions to locate in smaller towns, or rural, periurban, or metropolitan areas. In this chapter, we consider an additional process associated with the rural-urban trans- formation: migration of workers from rural to urban areas. Rural-to-urban migration has contributed significantly to the rapid urbanization of South Asia. Migrants to urban areas are made up of those in search of better living standards—“pull” migrants—and those who have been forced out of rural areas by a lack of job opportunities, poor public services, or natural or human-produced disasters—“push” migrants. Pull migrants frequently are well educated and skilled, raising concerns about a “brain drain” from the area of origin. However, recent evidence suggests that such concerns are misplaced. Migration benefits not only the migrants themselves but also those left behind (Clemens and McKenzie 2009). Moreover, migration helps to moderate the differences in living standards between destination and origin areas (box 10.1). In addition to facilitating regional convergence of living standards, migration plays a critical role in the reorganization of people and activities 175 176 From Farm to Firm Box 10.1 The Recent Evidence on Migration: A Selected Review Recent studies argue that simply comparing the earnings of migrants and non- migrants overestimates returns to migration (Gabriel and Schmitz 1995; Dahl 2002; Akee 2006). For example, McKenzie, Gibson, and Stillman (2006) show that ignoring the self-selection of migrants would lead to an overestimation of the gains from migration by 9–82 percent. Despite this selection issue, migration tends to benefit the migrant workers. Beegle, de Weerdt, and Dercon (2008) report that in Tanzania, the poverty rate among those who moved out of Kagera region dropped by 23 percentage points compared with a 12 percentage point drop among those who moved within the region and a 4 percentage point drop among those who did not move. Recent literature also highlights the role of networks in the migration process. Carrington, Detragiache, and Vishwanath (1996) argue that the presence of a large migrant population in the place of destination reduces migration costs and generates path dependence. Likewise, Munshi (2003) finds that interpersonal networks play a significant role in helping Mexican migrant workers in the United States. Using data on refugees resettled in various parts of the United States, Beaman (2006) reveals a more complex story: an influx of refugees initially over- whelms the network as it struggles to provide job-relevant information, but in the longer term, the network has a positive effect as new migrants find their way into employment. Meanwhile, Munshi and Rosenzweig (2005) find that in India, strong mutual assistance networks in the place of origin discourage migration. Work migrants often send remittances home, either on a regular basis or to deal with external shocks. These remittances often improve the living standards of nonmigrants and help them to invest (Adams 1997; Woodruff 2001; Lokshin, Bontch-Osmolovski, and Glinskaya 2007; Beegle, de Weerdt, and Dercon 2008). For some countries, they also constitute a significant source of foreign exchange and play an important macroeconomic role. Returning migrants may also invest some of the proceeds from migration into self-employment at home (e.g., McCormick and Wahba 2001; Mesnard and Ravallion 2001). over geographical space. For one to understand how the spatial pattern of population distribution, and thus urbanization, will evolve, it is important to identify factors that influence migrants’ choices of destination. The prospect of earning better wages is perhaps the most frequently cited explanation for the selection of destination. But other factors may Labor Mobility and the Rural-Urban Transformation 177 influence migration decisions as well, such as the provision of services or ease of assimilation at the destination (Lall, Timmins, and Yu 2009). Although recent literature has focused on four main issues—returns to migration, the role of networks, remittances, and return migration (box 10.1)—in this chapter we use the case of Nepal to investigate the determinants of migration. We begin by considering the extent of migration across South Asian countries before focusing on Nepal’s expe- rience. In Nepal, the bulk of migration is internal—from rural to urban areas—rather than to other countries. Our empirical analysis uncovers the motivation for migration within Nepal, considering the relative importance of infrastructure, provision of services, incomes, and social networks in migrants’ decisions. Internal Migration in South Asia Historically, many regions of the Indian subcontinent experienced sub- stantial seasonal migration. Reliable estimates of migration flows across regions during recent years are, however, difficult to obtain because internal migration (both temporary and permanent) remains a sparsely researched topic. Estimates based on recent population censuses suggest the presence of substantial differences in migration rates among countries. When migrants are defined as those who are living in areas different from their place of birth, about 30 percent of India’s population, 20 percent of Sri Lanka’s, 15 percent of Nepal’s, and 9 percent of Pakistan’s can be classified as migrants.1 Evidence from censuses also suggests an acceleration of migration rates during recent years. In India and Sri Lanka, about 20 percent of migrants moved during the four years before the census. In Nepal, about one-third of migrants moved during those years. In Nepal, Pakistan, and Sri Lanka, more than 60 percent of migration is from rural to urban areas. In India, rural to rural migration accounts for the bulk (53 percent) of all migration. Moreover, in India, most migration takes place within the province or state; only 14 percent of migrants moved to a state different from their birth state. Women and men migrate at different paces, and for different rea- sons, in these countries. In India, women are more mobile than men— constituting more than 70 percent of all migrants. In Nepal, women make up just over half of all migrants. Nearly three-quarters of South Asian women migrate because of marriage. Work migration among women is 178 From Farm to Firm minuscule in India (about 2 percent) and small in Nepal (22 percent). By contrast, men migrate principally because of work. Among all male migrants, more than a third migrated for work in India. The share is much higher in Nepal (54 percent) and Sri Lanka (45 percent). Among adult males, nearly 70 percent moved for work in Nepal. The pace of work migration has picked up in recent years in most South Asian countries. The evidence suggests that these migrants are often better educated than those left behind. For example, in Sri Lanka, of households moving in the country, the proportion of people with an O-level education or above is much higher among the heads of migrant households than among those who remained in the district of origin (World Bank 2010). A similar pattern is also true for Nepal. Internal Work Migration in Nepal Home to the Himalayas, Nepal remains one of South Asia’s remotest countries. The mountain and rural hills regions in Nepal offer few earning opportunities because of their difficult geography and lack of physical infrastructure. For households in these regions, temporary and permanent migration has been an important livelihood strategy. For centuries, people migrated to the Terai plains and adjacent areas in India during off-seasons to supplement their income. This migration continues to this day. According to a World Bank poverty assessment (World Bank 2006), remittances received by households from both internal and external migrants have been important in reducing the incidence of poverty in Nepal. Between 1995–96 and 2003–04, the poverty headcount ratio declined from 42 percent to 31 percent. A simulation exercise conducted by Lokshin, Bontch-Osmolovski, and Glinskaya (2007) indicates that if remittances had remained at their 1995–96 level, then the incidence of poverty in 2003–04 would have been about 3.9 percentage points higher. In other words, more than a third of the 11 percentage point decline in the poverty headcount ratio between 1995–96 and 2003–04 in Nepal was attributable to higher remittances. Although both internal and external migration are important in Nepal, we focus on the movement and motivation of internal migrants, who constitute the majority of Nepali migrants. According to the population census sample survey data (2001), about a fifth of adult workers in Nepal are migrant workers. The same census data also indi- cate that almost 70 percent of adult working-age male migrants move in search of new job opportunities. Labor Mobility and the Rural-Urban Transformation 179 Figures 10.1 and 10.2 display the out- and in-migration of male adult workers in Nepal estimated from census data. The districts of origin are distributed widely across the country (figure 10.1). By contrast, a small number of destination districts receive a disproportionate number of work migrants (figure 10.2). A comparison of figures 10.1 and 10.2 indi- cates a movement of migrants from the hill and mountain districts toward the Terai plains and urban hill areas. Compared with native workers, migrant workers are on average younger and better educated. About 40 percent of migrants are in the age 20–30 group compared with 19 percent of the native population (figure 10.3). Three-quarters of migrant workers compared with 40 percent of nonmigrants are aged 40 or younger. And nearly half of migrants have education above the secondary level compared with 12 percent of nonmigrants (figure 10.4). This higher level of education of migrants improves their access to salaried jobs. About 64 percent of migrants are employed in salaried jobs compared with 25 percent of nonmigrants (figure 10.5). Agriculture— the main occupation for nonmigrants—employs just 13 percent of migrants. Figure 10.1 Out-migration of Workers by District: Nepal, 2001 Source: World Bank staff estimates based on population census of Nepal, 2001. 180 From Farm to Firm Figure 10.2 In-migration of Workers: Nepal, 2001 Source: World Bank staff estimates based on population census of Nepal, 2001. Figure 10.3 Migrants and Nonmigrants by Age Category, Nepal 45 40 35 30 percent 25 20 15 10 5 0 20–30 years 30–40 years 40–50 years 50+ years age group nonmigrant migrant Source: World Bank staff estimates based on population census of Nepal, 2001. Labor Mobility and the Rural-Urban Transformation 181 Figure 10.4 Education Level of Migrants and Nonmigrants, Nepal 80 70 60 50 percent 40 30 20 10 0 up to primary primary-secondary above secondary nonmigrant migrant Source: World Bank staff estimates based on population census of Nepal, 2001. Figure 10.5 Employment of Migrants and Nonmigrants, Nepal 70 60 50 40 percent 30 20 10 0 agriculture salary/wage own business other nonmigrant migrant Source: World Bank staff estimates based on population census of Nepal, 2001. 182 From Farm to Firm Migration rates for adult males vary across ethno-linguistic groups. For example, Brahmins (high-caste Hindus) constitute 11.7 percent of adult male nonmigrant workers but 34.5 percent of adult male migrants (table 10.1). And migration propensity is much higher among people speaking the Nepali language (74 percent among migrants versus 45.3 percent among nonmigrants). The characteristics of migrants’ destinations differ substantially not only from those of their origin but also from those of other potential des- tinations. Table 10.2 reports the location characteristics of the actual des- tination compared with those of alternative destinations.2 All estimates in table 10.2 are relative to the district of origin. The consumption and housing price premiums in the actual destination district is positive rela- tive to those in the district of origin. The average housing price premium is higher in alternative destinations compared with that in the district of origin. More important, consumption and housing price premiums in the actual destination were significantly higher than those in the alterna- tive destinations in 1995–96. The actual destinations also offered lower rice prices and better services (faster travel times to banks and more paved roads) compared with those in both the origin and alternative des- tinations. The distance to the actual destination from the district of origin Table 10.1 Ethnic Profile of Migrants and Nonmigrants (Adult Males), Nepal percent Migrants Nonmigrants (adult males) Ethnicity Brahmin 34.5 11.7 Chhetri 21.5 15.6 Newar 7.4 7.9 Tharu 3.1 6.7 Magar 6.1 6.0 Tamang 4.2 5.9 Other 23.2 46.2 Language Nepali 73.9 45.3 Maithili 6.2 13.2 Bhojpuri 1.3 7.3 Newar 4.4 6.1 Tharu 2.0 5.8 Tamang 3.7 5.5 Other 8.5 16.8 Source: World Bank staff estimates based on population census of Nepal, 2001. Labor Mobility and the Rural-Urban Transformation 183 Table 10.2 Comparing Actual Destination and Alternative Destinations of Migrants, Nepal Mean in Mean in actual alternative Difference in mean destination destinations t-statistic Significance Consumption Average consumption expenditures (log) 0.075 –0.047 –33.587 *** Differential in log consumption due to education and language 0.009 0.012 0.907 Prices and services Log of rice price –0.088 0.021 47.577 *** Housing price premium (log) 0.377 0.210 –12.188 *** Travel time to nearest paved road –0.745 0.104 79.624 *** Travel time to nearest bank –0.373 0.092 71.245 *** Population and distance Population density 0.281 –0.033 –86.063 *** Log (population) 0.329 –0.208 –73.980 *** Elevation in meters –0.316 0.167 56.973 *** Ethno-caste similarity index –0.042 –0.060 –13.660 *** Language similarity index –0.123 –0.101 7.429 *** Religion similarity index 0.008 –0.016 –13.706 *** Distance (thousand km) 0.116 0.278 115.060 *** Source: Fafchamps and Shilpi 2008. Note: All figures are relative to the district of origin. ***p <.01. was also smaller than those to the alternative destinations. Migrants also chose locations with larger populations, higher rates of urbanization (measured by population density), and lower elevation. Social distance is measured by the proportion of individuals who share the same language, religion, and ethno-caste group as the migrant.3 In terms of social proximity, migrants on average face a population in their actual destination and alternative destinations that is more different from them in both language and ethno-caste than in their district of origin. A different pattern is observed for religion; if anything, migrants are more likely to encounter someone of their religion in their district of destina- tion than in their district of origin. The actual destination appears to be more diverse than alternative destinations in the language spoken but less diverse in religion and ethno-caste similarity measures. 184 From Farm to Firm Destination Choice of Migrants To understand what attracts a migrant to a particular location, we per- formed a regression analysis to ascertain the relative roles of income, serv- ices, and physical and social distances in the internal migration process in Nepal. We used an unusually detailed database from the sample survey of Nepal’s 2001 population census. The sample survey covers 11 percent of Nepal’s population, providing information on 2.5 million individuals (520,624 households). We focused our analysis on adult household heads (older than age 20) who migrated to seek work opportunities only. In the sample survey are some 16,850 such persons, representing about 70 percent of all adult male migrant household heads. The robustness of the empirical analysis is checked by estimating the regressions for the all adult male household heads sample. The empirical methodology involves several steps, including strategies to guard against reverse causality, as elaborated in the back- ground paper by Fafchamps and Shilpi (2009). Here we summarize the principal findings. Table 10.3 reports the estimation results from a sample of adult male work migrants and all adult male migrants. The distance variable meas- ures the distance between the district of origin and the district of poten- tial destinations, including the actual destination. This distance is a proxy for migration cost. In both regressions (male work migrants and all male migrants), distance has the expected negative sign and is strongly signifi- cant. This finding implies that migrants tend to move to a location that is closer to their district of origin than other potential locations. The average log consumption in table 10.3 is defined as the difference in average consumption (log) between the district of origin and the dis- tricts of potential and actual destinations.4 Similarly, the consumption dif- ferential stemming from education and language measures the education and language premium in the districts of potential destinations relative to the district of origin. The regression coefficients of both of these terms are positive. The coefficient of the education and language premium is highly significant and has a much larger magnitude. The regression results suggest that migrants do respond to income (and thus consumption) differentials and choose a destination that provides higher average consumption than alternative destinations. In particular, migrants choose destinations that offer larger returns to their skills. The coefficients of population density and log of population are posi- tive and strongly significant in both regressions, indicating that migrants Labor Mobility and the Rural-Urban Transformation 185 Table 10.3 Consumption and Choice of Migration Destination: Male Migrants, Nepal Adult male work migrants All adult male migrants Coefficient t-statistic Coefficient t-statistic Average consumption expenditures (log) 0.331* 1.753 0.433** 2.105 Log consumption differential due to education and language 0.533*** 6.330 0.614*** 7.197 Prices and services Log of rice price –4.492*** –6.809 –4.173*** –6.016 Housing price premium (log) 0.428*** 7.314 0.409*** 5.905 Travel time to nearest paved road –0.770*** –8.297 –0.831*** –8.431 Travel time to nearest bank 0.276 1.569 0.335 1.589 Elevation (1,000 meters) –0.512*** –3.149 –0.553*** –2.978 Population Population density 0.947*** 7.316 1.019*** 7.356 Log (population) 0.268** 2.099 0.357*** 2.669 Social distance Ethno-caste similarity index 0.749*** 2.607 0.794*** 2.921 Language similarity index 1.486*** 8.037 1.622*** 8.812 Religion similarity index –0.443 –1.416 –0.744** –2.265 Distance Distance (km, hundreds) –9.932*** –13.425 –10.62*** –13.481 Log-likelihood –47,880.35 –66,980.53 Number of observations 1,072,804 1,555,740 Pseudo R2 0.289 0.314 Source: Fafchamps and Shilpi 2008. Note: Fixed effect conditional legit estimation. Standard errors are corrected for clustering across district of origin. *p < .10. **p < .05. ***p < .01. move to locations that already have high concentrations of people. Because population tends to concentrate in areas with agglomeration economies and higher economic densities, this result suggests that the pull of opportunities in urban areas is quite important in attracting migrants to those locations. The regressions included indicators of services. Access to facilities such as schools, hospitals, and markets is measured by travel time to a paved road. Access to financial services is represented by the travel time to a bank. Elevation provides an indication of the difficulty of transporting 186 From Farm to Firm goods and people. The cost of living is measured by the price of rice and the housing price premium; the latter also provides an indication of pub- lic goods (such as air quality and congestion) that are not included directly as regressors in the estimation. The results in table 10.3 show that all but one of the coefficients (travel time to a bank) are highly signifi- cant. Migrants choose destinations that have much better access to schools, hospitals, and markets, as measured by travel to a paved road. The coefficient of elevation is negative, implying that, on average, migrants choose destinations where it is easier to transport goods and people. The regression coefficient implies that housing costs are higher in the destina- tion district than in alternative destinations, indicating that migrants also value public goods and services that are unobservable in regressions. Finally, conditional on the housing price premium and other services, migrants choose destinations where rice prices—an important measure of the cost of living in Nepal—are lower than in alternative destinations. Conditional on other regressors, the ethno-caste and language proxim- ity indexes are significant, with positive signs (table 10.3). This finding implies that the social proximity between the migrant and the population of the destination district is higher than that in alternative destinations. The religion proximity index is, however, not significant in the work migrants’ sample.5 Taken together, these results suggest that, conditional on the material benefits of migration, migrants prefer to move to a desti- nation where they can integrate more easily and possibly use networks to access jobs and housing. To provide a sense of the importance of different factors in deter- mining migration destination, we estimate the relative magnitude of a one standard deviation change in each regressor, keeping other regres- sors unchanged. The absolute values of these magnitudes are shown in figure 10.6. The most important regressor in terms of magnitude is dis- tance. Migrants strongly prefer a nearby destination. The second most important factor is the cost of living (measured by the price of rice). Factors related to access to services (travel time to the nearest road and housing price premium) also have relatively strong effects on destination choice. Among different measures of social proximity, the language simi- larity index has the largest relative impact. The effect of consumption variables on migration destination is smaller in magnitude. For example, a one standard deviation increase in anticipated relative consumption has an effect on destination that corresponds to a third of the effect of a one standard deviation in the housing price premium, and one-sixth of a one standard deviation in the log of the price of rice. Labor Mobility and the Rural-Urban Transformation 187 Figure 10.6 Effects of Regressors on Destination Choice (Absolute Values), Nepal 2.0 1.8 1.6 magnitude of one SD change in regressor 1.4 1.2 1.0 0.8 0.6 0.4 0.2 0.0 n m ice m ad n ity sim ) sim y y y ce nk n rit rit rit io io iu iu io an ba ns ro pr pt at ila ila ila em em at de st ed pu elev m tra ng p ice to sim ul di pr im e pr su tra pav op n e r tio im e ge n n ge (p st io ric co la to lt ua ua ca lig g ve ge lo e ng o- ng re po ra hn la la i lt e us et av d ve ho an n io at uc ed Source: Fafchamps and Shilpi 2008. Note: SD = standard deviation. Evidence from other developing countries underscores the role of serv- ices in attracting migrants to a location. Econometric estimations based on data from Sri Lanka find that migration decisions in the 1990s were influenced by district-level differences in access to well water and elec- tricity, particularly for the less educated (World Bank 2010). Analysis of migration decisions in Brazil suggests that there, too, migrants are attracted to locations based on the quality of public goods and services offered, and poorer migrants are willing to accept lower wages to access better services (box 10.2). The importance of physical and social distances in migrants’ destina- tion choices implies that migration flows are likely to be more vibrant across locations that are in close proximity. Thus smaller cities and towns can become important destinations for migrants from surrounding areas. But the growth of these cities and towns, according to the earlier empir- ical results, will depend on the availability of services. Investment in serv- ices in smaller towns and cities (or migrants’ districts of origin) can relieve congestion pressure from metropolitan cities while reducing rural-urban gaps in economic well-being. The costs and benefits of such investment programs, however, have to be weighed against the costs and benefits of providing services in larger cities. For example, investments in infrastructure in remote districts may prove to be prohibitively expensive, and investments in metropolitan 188 From Farm to Firm Box 10.2 Migrating to Economic Density in Brazil: Rational Decisions or Bright Lights? Economists have long argued that migration decisions are motivated by the pos- sibility of earning higher wages. But because many migrants do not find jobs after moving, this attraction may be irrational. Recent empirical evidence from four decades of Brazilian census data is enlightening. Working-age men migrated not only to look for better jobs but also to gain better access to basic public services such as piped water, electricity, and health care. These findings indicate that results from models of migration behav- ior that focus only on a migrant’s desire to move in search of better jobs can be biased because places with better public services also have more job opportuni- ties. Firms like to locate where workers would like to live. By ignoring the impor- tance of public services, some econometric estimates may overstate a migrant’s willingness to move in response to wage differences. For determination of how much public services matter, a rich set of data on public services at the municipality level was combined with individual records from Brazil’s census to evaluate the relative importance of wage differences and public services in a migrant’s decision to move. Predictably, wage differences are the main factor influencing migration choices. For the better-off, basic public services are not important in the decision to move. But for the poor, differences in access to basic public services matter. In fact, poor migrants are willing to accept lower wages to achieve access to better services. For example, a Brazilian minimum wage worker earning R$7 an hour was willing to pay R$420 a year to have access to better health services, R$87 for a better water supply, and R$42 for electricity. Source: World Bank 2008, based on Lall, Timmins, and Yu 2009. areas could help to reinforce preexisting agglomeration economies. Our findings suggest that, provided a rigorous cost-benefit analysis is per- formed, geographically targeted investments in public services could ease the rural-urban transition process. Notes 1. In Bangladesh, the migration data needed for a comparison with other South Asian countries are not available. The estimate for Sri Lanka includes those who migrated because of conflicts in part of the country. Labor Mobility and the Rural-Urban Transformation 189 2. Alternative destinations consist of all districts other than districts of birth and actual destination. 3. The construction of social distance variables is described in the background paper by Fafchamps and Shilpi (2009). These indexes are in practice simi- lar to the ethno-linguistic fractionalization measure common in the growth literature. 4. The income estimates from developing countries such as Nepal are notoriously weak because of the predominance of self-employment. Relative to income data, consumption data are of better quality. For this reason, we present the results using consumption measures. Regression estimates using income meas- ures can be found in the background paper by Fafchamps and Shilpi (2009). 5. For the sample of adult male migrants, the coefficient of the religion simi- larity index is negative and statistically significant. This finding indicates that work migrants tend to choose urban destinations that are more diverse in religion. References Adams, R. 1997. Remittances, Investment, and Rural Asset Accumulation in Pakistan. Washington, DC: International Food Policy Research Institute. Akee, R. K. Q. 2006. “Deciphering Immigrant Self-Selection: New Evidence from a Developing Country.” Kennedy School of Government, Harvard University, Cambridge, MA. Beaman, L. A. 2006. “Social Networks and the Dynamics of Labor Market Outcomes: Evidence from Refugees Resettled in the US.” Department of Economics, Yale University, New Haven, CT. Beegle, K., J. de Weerdt, and S. Dercon. 2008. “Migration and Economic Mobility in Tanzania: Evidence from a Tracking Survey.” Policy Research Working Paper 4798, World Bank, Washington, DC. Carrington, W. J., E. Detragiache, and T. Vishwanath. 1996. “Migration with Endogenous Moving Costs.” American Economic Review 86 (4): 909–30. Clemens, M. A., and D. McKenzie. 2009. “Think Again: Brain Drain.” Foreign Policy, October 22. Dahl, G. B. 2002. “Mobility and the Return to Education: Testing a Roy Model with Multiple Markets.” Econometrica 70 (6): 2367–420. Fafchamps, M., and F. Shilpi. 2008. “Determinants of Choice of Migration Destination.” Policy Research Working Paper 4728, World Bank, Washington, DC. ———. 2009. “Isolation and Subjective Welfare: Evidence from South Asia.” Economic Development and Cultural Change 57 (4): 641–83. 190 From Farm to Firm Gabriel, P. E., and S. Schmitz. 1995. “Favorable Self-Selection and the Internal Migration of Young White Males in the United States.” Journal of Human Resources 30 (3): 460–71. Lall, S., C. Timmins, and S. Yu. 2009. “Connecting Lagging and Leading Regions: The Role of Labor Mobility.” Brookings-Wharton Papers on Urban Affairs, 151–74, Brookings Institution, Washington, DC. Lokshin, M., M. Bontch-Osmolovski, and E. Glinskaya. 2007. “Work-Related Migration and Poverty Reduction in Nepal.” Policy Research Working Paper 4231, World Bank, Washington, DC. McCormick, B., and J. Wahba. 2001. “An Econometric Model of Temporary International Migration and Entrepreneurship.” Department of Economics, University of Southampton, Southampton, U.K. McKenzie, D., J. Gibson, and S. Stillman. 2006. “How Important Is Selection? Experimental vs. Non-Experimental Measures of the Income Gains from Migration.” IZA Discussion Paper 2087, Institute for the Study of Labor, Bonn. Mesnard, A., and M. Ravallion. 2001. “Is Inequality Bad for Business? A Nonlinear Empirical Model of Entrepreneurship.” ARQADQ, University of Toulouse, and DECRG, World Bank, Washington, DC. Munshi, K. 2003. “Networks in the Modern Economy: Mexican Migrants in the US Labor Market.” Quarterly Journal of Economics 118 (2): 549–99. Munshi, K., and M. Rosenzweig. 2005. “Why Is Mobility in India So Low? Social Insurance, Inequality and Growth.” Economic Growth Center, Yale University, New Haven, CT. Woodruff, C. 2001. “Remittances and Microenterprises in Mexico.” Graduate School of International Relations and Pacific Studies, University of California, San Diego, La Jolla. World Bank. 2006. “Nepal—Resilience amidst Conflict: An Assessment of Poverty in Nepal, 1995–96 and 2003–04.” Report 34834-NP, World Bank, Washington, DC. ———. 2008. World Development Report 2009: Reshaping Economic Geography. Washington, DC: World Bank. ———. 2010. “Sri Lanka: Connecting People to Prosperity.” Draft, South Asian Poverty Reduction and Economic Management Unit. Washington, DC. CHAPTER 11 Summary and Conclusions Today’s transition from rural to urban economies is an immense develop- ment challenge. Urbanization may spur growth and reduce poverty, but it also goes hand in hand with inequalities. Developing country policy makers and the development community are thus looking for ways in which to manage the rural-urban transition that ensure inclusive growth. Three Levels of Perspectives on Urbanization and Shared Growth: Sub-Saharan Africa As noted at the outset of this report, rural-urban transformations have been the focus of two consecutive editions of the World Development Report: WDR 2008, Agriculture for Development, and WDR 2009, Reshaping Economic Geography (World Bank 2007, 2008). Part I of this report is an empirical exploration of some of the principal propositions of WDR 2009, using examples from three Sub-Saharan countries: Ghana, Mozambique, and Uganda. Urbanization is inextricably linked to economic development. It accompanies, and in many ways drives, the structural transformation of economies that are largely agricultural to ones that are increasingly manufacturing and services-based. As part of this process, “leading” areas 191 192 From Farm to Firm emerge, characterized by average standards of living higher than those in “lagging” areas. The spatial inequalities that arise are often viewed as an inevitable part of the development process, rewarding higher productiv- ity, economies of scale, and agglomeration effects in the more advanced sectors of the economy. Although leading and lagging areas often corre- spond to rural-urban divides, the same spatial disparities may be repli- cated on larger or smaller geographical scales—for example, in the form of North-South divides or intraurban inequalities. Even if they are linked to the process of development, spatial inequal- ities can be far from benign. They may increase total inequality, which can diminish the impact of growth on poverty reduction. Spatial inequalities often correspond to ethnic, cultural, or religious divisions, leading to friction and unrest that may impede or even derail the devel- opment process. Although unequal outcomes are expected to some degree as a country develops, if left unchecked, spatial inequalities may become structural inequalities that translate into inequalities of oppor- tunity. What, then, is the scope for public policy to manage the rural- urban transformation and urbanization in a way that promotes equitable development? Our cross-country findings in chapter 2 suggest that few low- and middle-income countries fit neatly into the paradigm in WDR 2009 that rural-urban welfare disparities first diverge and then converge once coun- tries reach upper-middle-income status. Even in countries with the same levels of urbanization or GDP per capita, rural-urban differences may be enormous or relatively small. And, depending on the welfare meas- ure used, almost as many cases of convergence of living standards can be found within countries over time as cases of divergence. The exis- tence of so much diversity of experience at the country level implies that policies and institutions can play a crucial role in rural-urban transforma- tions, a proposition that is explored in much greater depth in part II of this report, which describes and analyzes the rural-urban transformation in South Asia. The decomposition analysis in chapter 3 provides some insights of the kinds of policies and institutions that could help to narrow rural-urban divides in Ghana, Mozambique, and Uganda—countries that qualify as “agriculture-based” and “incipient urbanizers,” according to the cate- gories put forth in WDR 2008 and WDR 2009, respectively. The princi- pal finding—that rural-urban inequalities in average household welfare stem primarily from differences in household endowments—suggests an important role for “spatially blind” policies and the kinds of institutions Summary and Conclusions 193 advocated in WDR 2009. These policies may include the universal pro- vision of services such as education, health, water, and sanitation financed through taxes and transfers, as well as regulations affecting land, labor, and business development that do not discriminate on the basis of location. Because rural-urban welfare inequalities are strongly associated with differences in education attainment, policies must focus on the equal provision of education (both quantity and quality) to narrow pro- ductivity and income gaps between leading and lagging areas. Nonmonetary welfare measures suggest that a significant rural-urban divide exists in more than just access to education. Stunting, an indica- tor of child health as well as the overall socioeconomic status of the household, is consistently found more in rural than in urban areas; ade- quate water and sanitation are consistently found less in rural areas than in urban areas, and in some cases are continuing to diverge. Thus the provision of spatially neutral institutions should also extend to increas- ing access to all basic services. Because rural areas are presently disad- vantaged, the provision of “spatially blind” services might effectively entail a renewed emphasis on lagging parts of the country until these areas catch up. Such investments in lagging areas are not intended to slow down the process of rural-urban transformation. Rather, because education and health are portable assets, expanded provision of them enables people to pursue a better quality of life in whatever geographical space they have chosen to make their home. However, improvements in basic services in rural areas may reduce excess migration to urban areas, because, as indi- cated in the case studies on South Asian countries in part II of this report, rural-urban migration choices are based not only on potential income but also on prospective access to services. In contrast to the main policy thrust of WDR 2009, the findings in chapter 3 suggest that even in “incipient urbanizers” such as Ghana, Mozambique, and Uganda, policy makers must look beyond the provision of spatially blind institutions to the provision of connective infrastructure. As described in WDR 2008, such infrastructure is necessary both to sup- port productivity gains among smallholder farmers and to promote the growth of the nonfarm sector and linkages between rural and urban economies. The effective delivery of services will depend in part on a higher level of infrastructure than currently exists in many rural areas of these countries. For example, a school is unlikely to be effective, or acces- sible, without a road to it. Likewise, the payoff to investments in portable assets depends in part on reducing the costs of trade for rural workers if 194 From Farm to Firm they choose to remain in a rural area, or the costs of migration if they seek to move to an urban area. The case studies of three African capitals in chapter 4 indicate that intraurban inequalities can be even starker than the rural-urban divide. Slums, in which hundreds of thousands of people lack access to even basic amenities, exist in close proximity to highly serviced luxury develop- ments. Colonial legacies, weak land institutions, and poor governance all have a role to play in the kinds of spatial disparities on view in Accra (Ghana), Maputo (Mozambique), and Kampala (Uganda). Local and national governments have recognized the urgent need to address the multiple issues—from inadequate legal frameworks and weak dispute settlement mechanisms to poor land management and corruption—that combine to create tight land and housing markets in each of these cities. As elaborated in greater detail in part II, land reforms are among the most challenging reforms to implement from a political economy perspective. However, intermediary measures such as efforts to improve land titling registration or codify existing laws are important steps forward. Evidence from other African cities suggests that tripartite agreements among governments, landlords, and tenants could help to improve tenure conditions. One example, from Nairobi’s slums, required landlords to relinquish control of some housing units in exchange for for- mal title and assistance in improving retained units (Gulyani and Talukdar 2008). Such incremental but meaningful steps represent a promising approach in view of the entrenched vested interests in land and its ties to economic and political power. The provision of adequate services for all urban residents remains a paramount challenge. Cohen (2007) recently remarked that “most urban analysts are still entering the city through the house and the bathroom, rather than through the workplace, the market, or as a site of creating long-term assets.” Nevertheless, the challenge of urban governance—to provide adequate infrastructure and services—is vital to both human development and job creation. Indeed, as demonstrated in part II, dis- economies caused by overcrowded and underserviced urban living can outweigh agglomeration economies, deterring firms from locating in large cities. The dichotomy between the household and workplace is perhaps a false one: both rely on adequate public services and good governance to flourish. Like the findings in chapter 3, those in chapter 4 indicate that policy makers in “incipient urbanizers” will need to look beyond the broad pol- icy guidance in WDR 2009, which argues they should focus first and Summary and Conclusions 195 foremost on “spatially neutral” institutions. Within cities, targeted inter- ventions such as slum clearance and housing construction may be an important policy step, even in countries that are only incipient urbanizers at the national level.1 Indeed, as discussed in chapter 4, slum upgrading programs in Accra have already yielded important successes that could be replicated on a wider scale. But as indicated in WDR 2009, investments in connective infrastructure or slum upgrading will be effective only if the basic institutions are sound. We hope that the findings of these chapters will spur deeper research and debate into the causes of spatial inequalities at the national and local levels. Land tenure in both rural and urban areas would certainly bene- fit from more in-depth analysis because of the evidence suggesting its profound impact on the rural-urban transformation (Shilpi 2010). Our research also demonstrates a need for more geographically sensitive poverty analysis of the kind generated by small area poverty maps, espe- cially in urban areas. Managing the Rural-Urban Transformation: Examples from South Asia Now the focus shifts back to the national level. South Asian countries are at various stages of the rural-urban transformation. Indeed, these coun- tries differ widely in their stages of agricultural development, urbaniza- tion, and living standards (chapter 5). Among the five major countries featured in part II of this report—Bangladesh, India, Nepal, Pakistan, and Sri Lanka—only Nepal can be classified as an agro-based country in which urbanization is at an incipient stage. The rest of the countries have experienced considerable declines in the importance of agriculture to economic growth and so are rapidly “transforming” (or urbanizing) coun- tries. All countries have experienced substantial declines in the incidence of poverty, although the rapidly urbanizing countries have experienced higher economic growth. Compared with cross-country differences, the pattern of rural-urban transformation within countries (chapter 6) suggests much greater variation in employment structure, urbanization levels, and economic well-being. Sri Lanka, the richest country in South Asia, has regions— such as Uva Province—in which the poverty incidence is comparable to that of poorer countries, and in which most people are still depend- ent on agriculture for their livelihoods. The Kathmandu Valley in Nepal, one of the least urbanized countries in South Asia, is nearly 196 From Farm to Firm fully urbanized with a poverty incidence of only 3.4 percent. These striking intracountry differences in economic structure and well-being are also present in Bangladesh, India, and Pakistan. A survey of the policy environment reveals that most of the South Asian countries implemented broad economic reforms over the last two decades (chapter 7). In Sri Lanka, an early mover, the first wave of trade and industrial deregulation began in 1977. By the early 1990s, the rest of the countries had implemented industrial and trade policy reforms. But in almost every country, agricultural reforms lagged significantly. As for the factor markets, one of the most important conditions shap- ing the trajectory of economic development in rural areas is the land market (chapters 7 and 8). All South Asian countries have attempted to introduce some land and tenancy reform measures, but the actual implementation experience has been uneven at best. The land market in Sri Lanka is characterized by the predominance of public ownership of agricultural land. The distribution of landownership is unequal in most countries, particularly in Pakistan. In the labor market, restrictions on the hiring and firing of workers can create an environment prohibitive to business start-up and expansion. However, enforcement of these restrictions remains uneven, applying primarily to the formal urban labor markets. In assessing the access to connective infrastructure, we find that although road densities are relatively high in South Asian countries (except Nepal), the intracountry differences in access to urban centers are striking (see chapters 7 and 9, which describe geographical linkages). Even in Sri Lanka, a country with the highest road density in South Asia, the relatively remoter regions are more than 15 hours away from the nearest city of 100,000 or more. These intracountry differences in access to urban centers correlate with geographical differences in economic structure and well-being. By contrast, intracountry differences in the provision of health care and education tend to be small. Sri Lanka ranks highest in terms of the over- all provision of health care and education and has also managed to pro- vide them equitably across geographical space. Even though the overall levels of health care and education are lower in Nepal and Bangladesh, the intracountry differences are still small. By contrast, Pakistanis are sub- ject to high inequality in health care and education along socioeconomic and gender lines. Migration rates among adult male workers also vary considerably across countries (see chapter 10, which discusses labor mobility across areas). In Summary and Conclusions 197 the mountains and hills of Nepal where earning opportunities are limited, rural-to-urban migration is an important part of the livelihood strategy. By contrast, much of the work migration in India takes place within rural areas. However, the pace of migration, particularly rural-to-urban migra- tion, has picked up in recent years in all of the South Asian countries. The evidence on the patterns of rural-urban transformation thus demonstrates clearly that South Asian countries are not only experienc- ing different stages of structural transformation but also following differ- ent paths of transition. The uneven progress in the implementation of policy and institutional reforms also suggests that, despite their common colonial histories, these countries now face different constraints to rural- urban transformation. The in-depth case studies reported in part II, while reinforcing some of the principal policy messages of WDR 2008 and WDR 2009, yield addi- tional policy insights for the rural-urban transformation. Here we summa- rize the major policy implications. Land reforms that improve the access of the poor to agricultural land could ensure a more inclusive rural-urban transformation. As revealed by the case study, in Pakistan, inequality in the distribution of land has forced a large portion of the population to seek off-farm employment, thereby depress- ing wages, particularly in agriculture and manufacturing. The effect is not confined to rural areas. Self-employment in smaller cities and towns and manufacturing in metropolitan areas are also affected. These findings bol- ster the case for land reform that would increase the access of the poor to land. Even though drastic land reform remains politically infeasible in Pakistan as in most other countries, several policy measures can increase the access of the poor to land. These include increased access to credit to enable poor people to purchase land, land taxation to minimize the hold- ing of land for speculative purposes, and measures to improve the effi- ciency of land sales and rental markets. Flexibility in land market transactions is essential for ensuring a faster and more inclusive rural-urban transition. WDR 2008 emphasizes the need for land reform to both spur the growth of agricultural productivity and reduce rural poverty. WDR 2009 highlights the need for flexibility in converting land from one use to another. According to the case study on Sri Lanka, restrictions on sales and subdivisions of agricultural land increase the cost of migration, resulting in slower urbanization and a higher incidence of poverty. Thus flexible land markets are needed not only at the urban fringe to allow physical expansion of urban areas, but also in rural areas beyond the urban fringe. 198 From Farm to Firm Investment in human capital remains a powerful instrument for under- taking a rural-urban transformation. The regression results confirm that educated people have better access to nonfarm jobs and can reap consid- erable positive returns in the labor market. The need to invest in human capital is much more urgent in countries such as Pakistan, where inequality in landownership may have led to underinvestment in human capital. When some of the institutional reforms (such as land reform) are politically difficult to implement, investment in human capital is an alter- native strategy for promoting a smoother rural-urban transformation. Investments in connective infrastructure are needed to strengthen farm- nonfarm and rural-urban linkages. WDR 2008 and WDR 2009 give high priority to investments in connective infrastructure as a step toward agri- cultural transformation, urbanization, and regional convergence in living standards. The results from the case studies reinforce this policy recom- mendation. Improved connectivity can reduce migration costs substan- tially, facilitating better integration of labor markets across regions. In addition, the case studies indicate the benefits of targeting infrastructure investments to areas with high potential agricultural growth, where improved connectivity would have the greatest impact on productivity and farm-nonfarm linkages. Similarly, targeting investments in infrastruc- ture and services in periurban and rural areas with existing concentrations of activities and access to markets could facilitate the further agglomera- tion of activities in these areas. Combining institutional reforms and infrastructure investment would have a much stronger positive impact on rural-urban transformation than pursu- ing these goals separately. Sectoral employment and wage patterns over geographical space are shaped by complex interactions of access to mar- kets and agrarian institutions. In Bangladesh, for example, the impact of access to large urban markets on farm-nonfarm linkages is particularly pronounced in areas with high agricultural potential. In Sri Lanka, the impact of land market restrictions is greater in villages closer to urban centers. The presence of interaction effects implies that countries may have to implement a package of complementary policies and investments to get a bigger bang out of their investments. Investment in services combined with infrastructure investment can ensure better integration of labor and product markets in rural and urban areas, facilitating the clustering of activities and workers. Together, the Bangladesh and Nepal case studies demonstrate the importance of services in attract- ing firms and workers to a location. Firms in Bangladesh tend to move to periurban and rural areas that already offer agglomeration economies and Summary and Conclusions 199 better access to large markets. Meanwhile, migrants in Nepal value serv- ices (such as schools, health care centers, and access to urban areas) just as they value income prospects. These results indicate scope for investing further in infrastructure outside of metropolitan areas—whether rural, periurban, or small towns—subject to a location-specific cost-benefit analysis. The relationship of the urbanization strategy to smaller towns and cities should be reconsidered. The empirical results suggest that smaller towns and cities in Bangladesh have not been able to attract manufacturing. The low clustering of manufacturing activities can be traced to these cities’ and towns’ dearth of infrastructure and services. Urban primacy, whereby Dhaka and Chittagong account for more than half of the urban population and receive the majority of public investments, is estimated to cost about 1 percent of economic growth (World Bank 2004). Connecting smaller towns and cities to national and international mar- kets and providing better services could encourage firms and people to move to these locations, creating alternative engines of growth and sec- toral transformation. The fact that migrants tend to relocate close to their regions of origin implies that investment in smaller towns will enable a faster convergence of living standards across areas. Possible spillovers across sectors and areas in policy making merit atten- tion. As revealed by the case studies, policies intended for one area may have substantial effects on others. For example, in Sri Lanka the produc- tive inefficiency arising from the restrictions posed by the Land Development Ordinance in agriculture was to a large extent mitigated by the massive public investment in irrigation schemes. Such policies may have addressed the immediate concerns in agriculture, but they resulted in slower employment diversification, slower migration, and lower overall living standards. Similarly, the adverse effects of land inequality in Pakistan extend far beyond rural areas and into the smaller towns and cities. In some cases, political economy considerations will be paramount. A political economy analysis was beyond the scope of our own research, but it is clear that the political environment can impose binding con- straints on reform. This finding is particularly true of land reform because of the role of land in ensuring access to productive assets and political influence. With the exception of India, none of the South Asian countries has been able to implement marginal tenancy reforms, let alone broader land reforms, in recent decades. The country case studies demonstrate that the costs of inaction are likely to be high, whether for 200 From Farm to Firm the rural-urban transformation, economic growth, or the convergence of living standards. These reforms will require a change in the mindset of policy makers and recognition of the importance of urbanization and rural-urban transformation in reducing rural and urban poverty. Management of the rural-urban transition in a way that preserves growth and promotes equity is one of the major challenges facing policy makers in today’s developing countries. The design and implementation of an effective policy mix rely in turn on having accurate information on the causes and manifestations of spatially based inequalities. We hope that the information conveyed in this report represents a modest step in the right direction. Note 1. In “incipient urbanizers,” the distribution of population is less than 25 percent urban and more than 75 percent rural. In “intermediate urbanizers,” the distribution is between 25 and 75 percent urban (or 75 percent urban and 25 percent rural). In “advanced urbanizers,” the population distribution is above 75 percent urban (below 25 percent rural). References Cohen, M. 2007. “Thinking Outside the Bubble: The Urban Crises of Land, Labor, and Capital.” Paper prepared for Lincoln Institute for Land Policy Symposium, Rio de Janeiro, October 1–2, 2007. http://www.gpia.info/files/u1/wp/2007-05 .pdf. Gulyani, S., and D. Talukdar. 2008. “Slum Real Estate: The Low-Quality High- Price Puzzle in Nairobi’s Slum Rental Market and Its Implications for Theory and Practice.” World Development 36 (10): 1916–37. Shilpi, F. 2010. “The Effect of Land Market Restrictions on Employment Pattern and Wages: Evidence from Sri Lanka.” World Bank, Washington, DC. World Bank. 2004. Bangladesh: Promoting the Non-Farm Sector in Bangladesh. Report 29719-BD. World Bank, Washington, DC. ———. 2007. World Development Report 2008: Agriculture for Development. Washington, DC: World Bank. ———. 2008. World Development Report 2009: Reshaping Economic Geography. Washington, DC: World Bank. APPENDIX A Countries and Surveys Used for Consumption and Poverty Analysis 201 202 Table A.1 Countries and Surveys Used for Consumption and Poverty Analysis GDP per capita GDP per capita Country (by World LSMS survey, LSMS survey, (2005 US$, PPP), (2005 US$, PPP), DHS survey, DHS survey, Bank Region) Abbreviation year 1 year 2 year 1 year 2 year 1 year 2 Africa Burkina Faso BFA 1994 2003 729 982 1992–93 2003 Cameroon CMR 1996 2001 1,678 1,870 1991 2004 Côte d’Ivoire CIV 1998 2002 1,885 1,672 1994 1998–99 Ghana GHA 1991–92 2005–06 886 1,209 1993 2003 Madagascar MDG 1997 2001 823 884 1992 2003–04 Mali MLI 1994 2001 737 933 1995–96 2006 Mozambique MOZ 1996–97 2002–03 435 596 1997 2003 Nigeria NGA 1996–97 2003 1,454 1,558 1990 2003 Senegal SEN 1994–95 2001 1,293 1,429 1992–93 2005 South Africa ZAF 1995 2000 7,319 7,466 — — Tanzania TZA 1991–92 2000–01 837 880 1992 2004 Uganda UGA 1992–93 2002–03 563 791 1995 2006 Zambia ZMB 1996 2003 1,055 1,107 1992 2001–02 East Asia and Pacific Cambodia KHM 1994 2004 769 1,291 2000 2005 Indonesia IDN 1993 2002 2,487 2,873 — — Vietnam VNM 1992 2002 997 1,780 — — Europe and Central Asia Armenia ARM 1998–99 2002 2,109 2,870 2000 2005 Kazakhstan KAZ 1996 2002 4,591 6,748 1995 1999 Kyrgyz Republic KGZ 1998 2002 1,408 1,556 — — Moldova MDA 1997 2002 1,519 1,711 — — Uzbekistan UZB 1998 2002 1,545 1,725 — — Latin America and the Caribbean Boliviaa BOL 1997 2002 3,508 3,565 1994 2003 a BRA 1990 2002 7,235 8,017 — — Brazil Colombiaa COL 1996 2003 5,596 5,498 1995 2005 Dominican Rep. DOM 1992 2003 3,493 5,014 1991 2002 Ecuadora ECU 1994 1998 5,665 5,862 — — El Salvadora SLV 1995 2002 4,707 5,018 — — Guatemalaa GTM 1998 2002 3,863 4,014 1995 1998–99 Hondurasa HND 1992 2003 2,744 3,044 — — Mexico MEX 1992 2002 9,550 10,815 — — Nicaraguaa NIC 1993 2001 1,734 2,145 1997–98 2001 Paraguaya PRY 1998 2003 4,069 3,706 — — Perua PER 1994 2003 4,978 5,894 1992 2000 Middle East and North Africa Egypt, Arab Rep. EGY 1995 2000 3,558 4,178 1992 2005 Morocco MAR 1990–91 1998 2,825 2,977 1992 2003–04 Tunisia TUN 1995 2000 4,422 5,444 — — (continued next page) 203 204 Table A.1 Countries and Surveys Used for Consumption and Poverty Analysis (continued) GDP per capita GDP per capita Country (by World LSMS survey, LSMS survey, (2005 US$, PPP), (2005 US$, PPP), DHS survey, DHS survey, Bank Region) Abbreviation year 1 year 2 year 1 year 2 year 1 year 2 South Asia Bangladesh BGD 1991–92 2000 712 901 1996–97 2004 India IND 1993–94 2005 1,303 2,230 1998–99 1992–93 Nepal NPL 1995–96 2003–04 818 938 1996 2006 Pakistan PAK 1992–93 2002 1,799 1,937 1990–91 2006–07 Sri Lanka LKA 1995–96 2002 2,520 3,092 — — Source: GDP per capita data from World Development Indicators, World Bank (2009). Note: DHS = Demographic and Health Surveys; LSMS = Living Standards Measurement Study; PPP = purchase power parity. a. Consumption and poverty measures are based on income instead of on expenditure or consumption. APPENDIX B Data, Methodology, and Results for Decomposition Analysis To analyze the factors behind the rural-urban welfare inequalities, we used variations of the Oaxaca-Blinder decomposition to estimate the rel- ative contributions of (1) the differences in household and community endowments or characteristics, and (2) the differences in the returns to these characteristics in accounting for differences in household welfare. These decompositions are simple descriptive tools that provide a useful way of summarizing the correlation among welfare differentials, endow- ments, and returns. Data The Ghana data are drawn from the 1991–92 and 2005–06 rounds of the Ghana Living Standards Surveys (GLSS3 and GLSS5, respectively). These nationally representative cross-sectional surveys are conducted by the government-run Ghana Statistical Service. The multipurpose household surveys provide detailed information on the household con- sumption of purchased and home-produced goods. The surveys are based on a two-stage (nonstratified) sample design. The Ghana Statistical Service used an amended sampling procedure to produce a self-weighting sample in GLSS3 and computed weights for GLSS5 (GLSS 2007). After the data 205 206 From Farm to Firm were cleaned, the sample size was 4,498 households in GLSS3 and 8,652 households in GLSS5. The Mozambique data are drawn from the 1996–97 and 2002–03 rounds of the national household living standards survey (Inquérito aos Agregados Familiares sobre as Condições de Vida, or IAF96 and IAF02). These nationally representative cross-sectional surveys are conducted by the National Statistics Institute. The multipurpose household and community surveys are similar to the World Bank’s Living Standards Measurement Study (LSMS) surveys, including detailed information on household consumption of purchased and home-produced goods. A three-stage stratified cluster sample design was used for these sur- veys, with a sample of 8,250 households in IAF96 and 8,700 households in IAF02. The Uganda data are drawn from the 1992–93 Uganda National Integrated Household Survey and the 2005–06 Uganda National Household Survey (UNHS) produced by the Uganda Bureau of Statistics. These standard LSMS-type surveys covered 9,923 households in 1992–93 and 7,421 households in 2005–06. Model Specification For intertemporal and spatial comparability of welfare, the log welfare ratio is the measure of welfare used as the dependent variable in the regression analysis. The welfare ratio is defined here as the nominal household consumption expenditure per adult equivalent divided by the applicable regional poverty line for the same survey year. Because of the spatial price variation across regions and rural and urban areas, the use of regional poverty lines in the welfare ratio improves comparabil- ity across these areas. The poverty lines are calculated to represent the same standard of living across different survey years, and thus dividing nominal consumption by the poverty line ensures comparability across years. Anyone interpreting the log welfare ratio should note that a value of zero implies that a household is living exactly at the poverty line. The regression analysis was conducted at the household level. The right-hand variables in the regressions include: household size or adult equivalents, age and age squared of household head (as a proxy for expe- rience), urban and regional dummy variables and their interactions, educa- tional attainment of the household head or maximum educational attainment of the entire household, and the employment sector of the household head. In the Uganda regressions, we also used community-level Data, Methodology, and Results for Decomposition Analysis 207 variables indicating the presence of a secondary school, hospital, or tarmac trunk road. Community-level variables are not used in the Mozambique regressions because in Mozambique the community survey was conducted only in rural areas. The regression results are shown in tables B.1, B.2, and B.3. Oaxaca-Blinder Decomposition Given any two subpopulations A and B (e.g., rural and urban areas of a given year or two different points in time), we assume that the logarithm of the welfare ratio for each group, denoted by lnC, can be summarized by the linear regression lnCj = bjXj + ej, where j represents subpopulations A and B of the sample, and e is a random disturbance term with a normal distribution and mean equal to zero. In this specification, the coefficients bj summarize the influence of a variety of factors Xj on welfare. The dif- ference in mean welfare ratios between subpopulations A and B can then be expressed as lnC A − lnCB = β A X A − β B X B , where the bar over the relevant variables denotes the sample mean val- ues of the respective variables. As proposed by Oaxaca (1973) and Blinder (1973), we can add and subtract the term β B X A or the term β A X B and express the difference as either ( ) lnC A − lnCB = X A − X B β B + ( β A − β B ) X A , or ( ) lnC A − lnCB = X A − X B β A + ( β A − β B ) X B . These decompositions produce two components: one that consists of the differences in average characteristics and another that consists of the differences in the coefficients. The only difference between these two expressions lies in how the differences in the characteristics X − X ( ) and the differences in coefficients ( β A − β B ) are weighted. A B Since the original decomposition by Oaxaca and Blinder, numerous scholars (e.g., Reimers 1983; Cotton 1988; Neumark 1988) have extended the method by proposing alternative weights for the differences in the ( ) characteristics X A − X B and the differences in returns ( β A − β B ) .1 Because of the path dependence of the decompositions, a common prac- tice is to average the results of the different “paths.” We follow Reimers 208 From Farm to Firm (1983) and use as weights the average of the coefficients and the average of the characteristics so that ⎛ ( β + βB ) ⎞ ( ⎛ X A + XB ) ⎞. ( ) lnC A − lnCB = X A − X B ⎜ A 2 ⎟ + ( βA − βB ) ⎜ ⎜ 2 ⎟ ⎟ ⎝ ⎠ ⎝ ⎠ Detailed decompositions can indicate which variable or set of variables is driving the endowment or coefficient effects. Quantile Decomposition By constructing counterfactual distributions using quantile regression estimates, we decompose differences in welfare distributions into a com- ponent based on differences in covariates and a component based on dif- ferences in coefficients. The counterfactual distributions are estimated by first obtaining quantile regression estimates at each percentile for two mutually exclusive subpopulations, A and B (e.g., rural and urban), and then predicting log welfare ratios using the covariates of the other sub- population. These predicted values are then randomly sampled with replacement—with probability proportional to the survey weight of the observation—to obtain the desired self-weighting counterfactual distribu- tions. Differences in log welfare ratios between the two subpopulations are then decomposed by comparing the predicted distributions with the constructed counterfactual distribution—see Machado and Mato (2005) for additional details. Data, Methodology, and Results for Decomposition Analysis 209 Table B.1 Regression Results, Ghana Dependent variable: log welfare ratio 1991–92 2005–06 Adult equivalents –0.134*** –0.135*** (0.008) (0.007) urban*adult equivalents –0.022 –0.025** (0.014) (0.011) Age of head –0.007* 0.006 (0.004) (0.004) urban*age of head 0.012 0.003 (0.009) (0.006) age of head squared (1/1,000) 0.076* –0.049 (0.040) (0.036) urban*age of head squared (1/1,000) –0.075 –0.005 (0.094) (0.058) Education of head primary 0.003 0.092*** (0.036) (0.032) urban*primary –0.015 0.043 (0.065) (0.051) junior secondary 0.092*** 0.153*** (0.028) (0.033) urban*junior secondary 0.115** 0.097 (0.047) (0.049) some senior secondary 0.186** 0.296*** (0.081) (0.060) urban*some senior secondary –0.031 0.136* (0.101) (0.076) completed SSS or higher 0.236*** 0.498*** (0.068) (0.055) urban * SSS+ 0.270*** 0.121* (0.093) (0.072) Employment sector of head secondary sector 0.153*** 0.133*** (0.047) (0.043) urban*secondary –0.089 0.007 (0.065) (0.064) tertiary sector 0.277*** 0.209*** (0.037) (0.036) urban*tertiary –0.130** –0.019 (0.054) (0.055) other (unemployed) 0.148** –0.056 (0.062) (0.047) urban*other –0.092 0.043 (0.090) (0.072) (continued next page) 210 From Farm to Firm Table B.1 (continued) Dependent variable: log welfare ratio 1991–92 2005–06 Region/ecological zone forest –0.084* –0.02 (0.049) (0.055) urban*forest 0.163** 0.092 (0.071) (0.073) savannah –0.190*** –0.314*** (0.058) (0.069) urban*savannah 0.263** 0.184* (0.106) (0.108) Urban –0.020 0.090 (0.198) (0.160) Constant 0.612*** 0.664*** (0.111) (0.100) N 4,498 8,651 R2 0.399 0.413 F 54.975 71.663 Source: Authors’ calculations from Ghana Living Standards Surveys. *p < .10. **p < .05. *** p < .01. Table B.2 Regression Results, Mozambique Dependent variable: log welfare ratio 1996–97 2002–03 Adult equivalents –0.178*** –0.131*** (0.008) (0.009) urban*adult equivalents 0.052*** 0.023 (0.013) (0.013) Age of head –0.007 –0.007** (0.005) (0.004) urban*age of head 0.025*** 0.021** (0.009) (0.009) age of head squared (1/1,000) 0.074 0.099*** (0.048) (0.036) urban*age of head squared (1/1,000) –0.221** –0.229** (0.100) (0.099) Education of head primary 1 0.143*** 0.081** (0.025) (0.037) urban*primary 1 0.114*** 0.104 (0.043) (0.066) primary 2 0.281*** 0.279*** (0.067) (0.071) (continued next page) Data, Methodology, and Results for Decomposition Analysis 211 Table B.2 (continued) Dependent variable: log welfare ratio 1996–97 2002–03 urban*primary 2 0.278*** 0.125 (0.097) (0.095) secondary 1st cycle 0.504 0.596*** (0.314) (0.122) urban*secondary 1st cycle 0.298 0.077 (0.324) (0.148) secondary (2nd cycle) or higher 0.848 0.570*** (0.550) (0.204) urban*secondary (2nd cycle) or higher 0.333 0.660 (0.566) (0.242) Employment sector of head secondary sector –0.019 0.173** (0.057) (0.083) urban*secondary 0.178** –0.072 (0.077) (0.100) tertiary sector 0.264*** 0.314*** (0.046) (0.056) urban*tertiary –0.023 –0.045 (0.076) (0.072) other (excludes primary) 0.020 0.019 (0.042) (0.076) urban*other 0.051 0.239** (0.082) (0.094) Region northern (region1) 0.053 0.297*** (0.049) (0.049) urban*northern –0.515*** –0.139 (0.118) (0.097) central (region2) –0.069 0.518*** (0.045) (0.054) urban*central –0.159* –0.345*** (0.083) (0.074) Urban –0.743*** –0.426** (0.199) (0.197) Constant 0.568*** 0.162* (0.118) (0.089) N 8,241 8,700 R2 0.266 0.244 F 45.747 30.393 Source: Authors’ calculations from Mozambique Household Living Conditions Surveys. *p < .10. **p < .05. *** p < .01. 212 From Farm to Firm Table B.3 Regression Results, Uganda Dependent variable: 1992–93 2005–06 1992–93 2005–06 log welfare ratio Adult equivalents –0.091*** –0.073*** –0.091*** –0.076*** (0.007) (0.004) (0.006) (0.004) urban*adult equivalents –0.001 –0.020* (0.013) (0.011) Age of head 0.001 0.006** 0.002 0.010*** (0.003) (0.003) (0.003) (0.003) urban*age of head 0.008 0.023*** (0.008) (0.007) age of head sq. (1/1,000) –0.017 –0.021 –0.025 –0.055 (0.032) (0.031) (0.030) (0.029) urban*age sq. (1/1,000) –0.078 –0.200*** (0.090) (0.066) Education of head some primary 0.105*** 0.132*** 0.106*** 0.137*** (0.019) (0.024) (0.019) (0.024) urban*some primary 0.094 0.021 0.085 –0.023 (0.059) (0.073) (0.060) (0.071) completed primary 0.152*** 0.268*** 0.154*** 0.276*** (0.031) (0.029) (0.030) (0.029) urban*comp primary 0.083 0.018 0.073 –0.031 (0.079) (0.087) (0.082) (0.083) some secondary 0.276*** 0.429*** 0.277*** 0.436*** (0.032) (0.029) (0.031) (0.028) urban*some secondary 0.119 0.080 0.113 0.027 (0.072) (0.077) (0.073) (0.073) completed sec or higher 0.608*** 0.789*** 0.608*** 0.793*** (0.052) (0.065) (0.052) (0.065) urban*comp sec + 0.134 0.329*** 0.134 0.295*** (0.095) (0.111) (0.094) (0.106) Employment sector of head secondary sector 0.132*** 0.054* 0.132*** 0.057* (0.044) (0.032) (0.045) (0.032) urban*secondary –0.152** 0.078 –0.154** 0.063 (0.076) (0.069) (0.078) (0.069) tertiary sector 0.182*** 0.277*** 0.182*** 0.279*** (0.029) (0.024) (0.029) (0.024) urban*tertiary 0.057 –0.055 0.055 –0.070 (0.055) (0.056) (0.055) (0.055) other (excludes primary) –0.027 –0.008 –0.026 –0.007 (0.039) (0.037) (0.039) (0.037) urban*other 0.325*** 0.196** 0.311*** 0.160* (0.107) (0.095) (0.097) (0.091) (continued next page) Data, Methodology, and Results for Decomposition Analysis 213 Table B.3 (continued) Dependent variable: 1992–93 2005–06 1992–93 2005–06 log welfare ratio Community secondary school 0.194*** 0.126*** 0.194*** 0.124*** (0.055) (0.029) (0.055) (0.029) tarmac trunk road 0.107** 0.123*** 0.107** 0.122*** (0.054) (0.028) (0.054) (0.028) Region eastern (region2) –0.073** –0.260*** –0.073** –0.261*** (0.035) (0.028) (0.035) (0.029) urban*eastern –0.199*** –0.030 –0.197*** –0.029 (0.064) (0.070) (0.065) (0.070) northern (region3) –0.277*** –0.571*** –0.277*** –0.571*** (0.042) (0.032) (0.042) (0.032) urban*northern –0.056 0.037 –0.052 0.032 (0.096) (0.068) (0.096) (0.068) western (region4) 0.026 –0.019 0.026 –0.018 (0.037) (0.029) (0.037) (0.029) urban*western –0.122 –0.073 –0.122 –0.071 (0.082) (0.062) (0.082) (0.062) Urban –0.073 –0.363** 0.102 0.160* (0.181) (0.168) (0.069) (0.093) Constant 0.171** 0.341*** 0.149** 0.258*** (0.069) (0.076) (0.065) (0.071) N 9,115 7,291 9,115 7,291 R2 0.263 0.422 0.263 0.421 F 34.688 94.690 37.122 103.335 Source: Authors’ calculations from Uganda National Household Surveys. *p < .10. **p < .05. *** p < .01. Table B.4 Average Household Characteristics: Ghana, Mozambique, and Uganda, Selected Years 214 Ghana Mozambique Uganda 1991 2005 1996 2002 1992 2005 rural urban rural urban rural urban rural urban rural urban rural urban welfare ratio 1.70 2.22 1.98 3.46 welfare ratio 1.08 1.39 1.29 1.77 welfare ratio 1.19 2.07 1.84 3.79 equivalent 3.38 3.20 3.30 2.72 equivalent 3.31 3.91 3.27 3.75 equivalent 3.64 3.14 4.01 3.59 age of head 45.19 42.54 46.30 43.27 age of head 42.37 41.58 42.89 42.32 age of head 42.24 35.33 43.02 38.04 hd educ: primary 0.49 0.27 0.42 0.17 hd educ: primary 0.79 0.42 0.83 0.47 hd educ: primary 0.35 0.12 0.19 0.08 hd educ: jr. hd educ: jr. hd educ: jr. secondary 0.12 0.08 0.15 0.10 secondary 0.18 0.36 0.12 0.25 secondary 0.43 0.32 0.45 0.26 hd educ: hd educ: hd educ: some sr. sec. 0.34 0.46 0.35 0.43 some sr. sec. 0.02 0.15 0.03 0.15 some sr. sec. 0.09 0.13 0.14 0.14 hd educ: comp. sr. hd educ: comp. sr. hd educ: comp. sr. sec. + 0.02 0.04 0.02 0.09 sec. + 0.00 0.05 0.01 0.07 sec. + 0.11 0.30 0.17 0.35 hd_edlev4a 0.03 0.14 0.05 0.21 hd_edlev4a 0.00 0.02 0.00 0.06 hd_edlev4a 0.02 0.14 0.04 0.17 hd: primary sector 0.74 0.20 0.72 0.17 hd: primary sector 0.87 0.29 0.88 0.39 hd: primary sector 0.77 0.13 0.71 0.15 hd: secondary sector 0.07 0.17 0.07 0.18 hd: secondary sector 0.03 0.15 0.02 0.09 hd: secondary sector 0.04 0.19 0.06 0.17 hd: tertiary sector 0.16 0.53 0.15 0.50 hd: tertiary sector 0.05 0.39 0.08 0.45 hd: tertiary sector 0.12 0.61 0.16 0.59 hd: other 0.04 0.10 0.06 0.15 hd: other 0.05 0.17 0.02 0.08 hd: other 0.06 0.08 0.07 0.09 coastal region 0.24 0.57 0.22 0.49 Northern (region1) 0.37 0.32 0.36 0.37 Central (region1) 0.28 0.62 0.26 0.59 forest region 0.47 0.31 0.48 0.39 Central (region2) 0.46 0.25 0.46 0.27 Eastern (region2) 0.28 0.21 0.26 0.13 savannah region 0.29 0.12 0.30 0.12 Southern (region3) 0.17 0.44 0.19 0.36 Northern (region3) 0.16 0.06 0.20 0.15 Western (region4) 0.28 0.11 0.28 0.13 comm: sec. school 0.05 0.05 0.14 0.40 comm: tarmac road 0.05 0.63 0.09 0.46 Source: Authors’ calculations based on household survey data. Table B.5 Decompositions of Rural-Urban Welfare Differences: Ghana, Mozambique, and Uganda, Selected Years Ghana 1991–92 2005–06 Variable Endowment SE Coefficient SE Endowment SE Coefficient SE equivalents 0.027 0.016* –0.073 0.045 0.086 0.011*** –0.076 0.033** agehd 0.004 0.012 0.510 0.391 -0.021 0.009** 0.132 0.262 agehd^2 (1/1,000) –0.011 0.014 –0.161 0.202 0.015 0.009* –0.010 0.130 hd educ: primary (P1-5) 0.000 0.001 –0.002 0.007 -0.006 0.002*** 0.006 0.007 hd educ: jr. secondary 0.018 0.005*** 0.046 0.019** 0.016 0.004*** 0.038 0.019** hd educ: some sr. secondary 0.003 0.001** –0.001 0.003 0.023 0.003*** 0.008 0.004 hd educ: comp. sr. secondary + 0.042 0.008*** 0.022 0.008*** 0.086 0.009*** 0.016 0.009* hd: secondary sector 0.011 0.004*** –0.010 0.008 0.015 0.004*** 0.001 0.008 hd: tertiary sector 0.079 0.011*** –0.045 0.019** 0.070 0.010*** –0.006 0.018 hd: other 0.006 0.003** –0.006 0.006 –0.003 0.003 0.004 0.007 forest zone 0.000 0.006 0.063 0.028** –0.002 0.004 0.040 0.032 savannah zone 0.010 0.009 0.054 0.023** 0.041 0.012*** 0.039 0.023 constant –0.020 0.198 0.090 0.160 Total 0.190 0.029*** 0.378 0.038*** 0.321 0.028*** 0.281 0.036*** 33% 67% 53% 47% 215 (continued next page) 216 Table B.5 (continued) Mozambique 1996–97 2002–03 Variable Endowment SE Coefficient SE Endowment SE Coefficient SE equivalents –0.091 0.016*** 0.188 0.046*** -0.057 0.019*** 0.080 0.046* agehd –0.004 0.005 1.050 0.374*** -0.002 0.003 0.909 0.379** agehd^2 (1/1,000) 0.004 0.006 –0.434 0.196** 0.001 0.004 –0.468 0.202** hd educ: primary (1GR) 0.036 0.006*** 0.031 0.012*** 0.017 0.005*** 0.019 0.012 hd educ: primary (2GR) 0.051 0.009*** 0.024 0.009*** 0.040 0.008*** 0.011 0.008 hd educ: secondary (1CIC) 0.028 0.009*** 0.007 0.008 0.038 0.007*** 0.003 0.006 hd educ: secondary (2CIC) + 0.020 0.007*** 0.003 0.006 0.054 0.012*** 0.021 0.009** hd: secondary sector 0.008 0.005* 0.016 0.007** 0.009 0.004*** –0.004 0.005 hd: tertiary sector 0.086 0.015*** –0.005 0.017 0.107 0.019*** –0.012 0.019 hd: other 0.005 0.005 0.006 0.009 0.007 0.003*** 0.012 0.005** northern region 0.011 0.011 –0.177 0.043** 0.003 0.012 –0.050 0.035 central region 0.032 0.010*** –0.057 0.029* –0.065 0.012*** –0.126 0.028*** constant –0.743 0.199 –0.426 0.197 Total 0.186 0.030*** –0.090 0.052* 0.151 0.025*** –0.029 0.037 194% –94% 124% –24% Uganda 1992–93 2005–06 Variable Endowment SE Coefficient SE Endowment SE Coefficient SE equivalents 0.046 0.007*** –0.003 0.045 0.035 0.007*** –0.077 0.043* agehd –0.038 0.029 0.317 0.325 –0.088 0.020*** 0.928 0.277*** agehd^2 (1/1,000) 0.037 0.029 –0.139 0.156 0.059 0.017*** –0.383 0.124*** hd educ: some primary –0.018 0.004*** 0.037 0.022* –0.026 0.007*** 0.006 0.026 hd educ: comp. primary 0.008 0.003*** 0.009 0.008 –0.002 0.004 0.001 0.012 hd educ: some secondary 0.063 0.008*** 0.026 0.015* 0.083 0.012*** 0.020 0.020 hd educ: comp. secondary + 0.082 0.011*** 0.012 0.008 0.119 0.016*** 0.034 0.012*** hd: secondary sector 0.008 0.005 –0.017 0.009* 0.010 0.004** 0.009 0.008 hd: tertiary sector 0.102 0.014*** 0.023 0.020 0.105 0.013*** –0.021 0.022 hd: other 0.002 0.002 0.023 0.007*** 0.002 0.001 0.016 0.008** com: sec. school 0.000 0.002 –0.007 0.005 0.029 0.010*** –0.020 0.016 com: trunk tarmac 0.060 0.029** –0.010 0.033 0.046 0.013*** 0.006 0.017 eastern region 0.013 0.006** –0.047 0.016*** 0.035 0.005*** –0.008 0.013 northern region 0.032 0.006*** –0.006 0.010 0.030 0.007*** 0.004 0.012 western region 0.006 0.007 –0.024 0.016 0.007 0.005 –0.014 0.013 constant –0.070 0.180 –0.335 0.171* Total 0.401 0.036*** 0.123 0.037*** 0.443 0.030*** 0.165 0.031*** 77% 23% 73% 27% Source: Authors’ calculations based on household survey data. Note: SE = standard error. *p < .10. **p < .05. ***p < .01. 217 218 Table B.6 Decompositions of Welfare Differences over Time for Rural and Urban Areas: Ghana, Mozambique, and Uganda, Selected Years Ghana Rural (1991–92 to 2005–06) Urban (1991–92 to 2005–06) Variable Endowment SE Coefficient SE Endowment SE Coefficient SE equivalents 0.011 0.011 –0.002 0.035 0.076 0.017*** –0.010 0.041 agehd –0.001 0.003 0.589 0.254** 0.005 0.005 0.180 0.390 agehd^2 (1/1,000) 0.001 0.003 –0.292 0.127** –0.003 0.005 –0.111 0.198 hd educ: primary (P1–5) 0.002 0.001 0.012 0.007 0.001 0.001 0.014 0.006** hd educ: jr. secondary 0.001 0.002 0.021 0.015 –0.007 0.005 0.019 0.023 hd educ: some sr. secondary 0.000 0.001 0.003 0.002 0.013 0.003*** 0.018 0.005*** hd educ: comp. sr. secondary + 0.010 0.002*** 0.010 0.003*** 0.038 0.012*** 0.020 0.014 hd: secondary sector 0.001 0.001 –0.001 0.004 0.001 0.002 0.014 0.011 hd: tertiary sector –0.003 0.003 –0.011 0.008 –0.006 0.004 0.022 0.029 hd: other 0.001 0.001 –0.010 0.004** 0.001 0.002 –0.009 0.011 forest zone –0.001 0.002 0.030 0.035 0.006 0.005 –0.003 0.025 savannah zone –0.003 0.010 –0.037 0.027 0.000 0.001 –0.024 0.015 constant 0.052 0.149 0.163 0.206 Total 0.019 0.021 0.365 0.031*** 0.127 0.024*** 0.293 0.034*** 5% 95% 30% 70% Mozambique Rural (1996–97 to 2002–03) Urban (1996–97 to 2002–03) Variable Endowment SE Coefficient SE Endowment SE Coefficient SE equivalents 0.006 0.008 0.156 0.040*** 0.018 0.021 0.070 0.052 agehd –0.004 0.004 0.002 0.259 0.012 0.012 –0.151 0.463 agehd^2 (1/1,000) 0.006 0.005 0.051 0.124 –0.013 0.012 0.034 0.248 hd educ: primary (1GR) –0.007 0.002*** –0.010 0.007 –0.025 0.007*** –0.022 0.020 hd educ: primary (2GR) 0.001 0.001 –0.000 0.003 –0.001 0.011 –0.023 0.014 hd educ: secondary (1CIC) 0.004 0.002** 0.001 0.002 0.018 0.009** –0.007 0.007 hd educ: secondary (2CIC) + 0.001 0.001* –0.000 0.001 0.050 0.014*** 0.002 0.008 hd: secondary sector –0.001 0.001 0.005 0.003* –0.008 0.003** –0.007 0.009 hd: tertiary sector 0.009 0.003*** 0.003 0.005 0.015 0.014 0.012 0.032 hd: other –0.001 0.001 –0.000 0.003 –0.015 0.005*** 0.023 0.011 northern region –0.002 0.004 0.088 0.025*** –0.008 0.011 0.212 0.051 central region –0.001 0.006 0.270 0.033*** –0.001 0.001 0.104 0.024 constant –0.406 0.148 –0.089 0.237 Total 0.013 0.009 0.161 0.028*** 0.043 0.033 0.157 0.052 7% 93% 21% 79% (continued next page) 219 220 Table B.6 (continued) Uganda Rural (1992–93 to 2005–06) Urban (1992–93 to 2005–06) Variable Endowment SE Coefficient SE Endowment SE Coefficient SE equivalents –0.030 0.004*** 0.070 0.031** –0.042 0.009 –0.003 0.053 agehd 0.003 0.002 0.204 0.187 0.053 0.019 0.716 0.365 agehd^2 (1/1,000) –0.001 0.001 –0.011 0.093 –0.037 0.015 –0.197 0.157 hd educ: some primary 0.002 0.001 0.012 0.014 –0.010 0.005 –0.016 0.026 hd educ: comp. primary 0.012 0.002*** 0.013 0.005*** 0.002 0.004 0.005 0.014 hd educ: some secondary 0.020 0.003*** 0.021 0.006*** 0.022 0.011 0.034 0.031 hd educ: comp. secondary + 0.014 0.003*** 0.005 0.003** 0.024 0.019 0.055 0.019 hd: secondary sector 0.001 0.001** –0.004 0.003 –0.001 0.002 0.026 0.015 hd: tertiary sector 0.009 0.002*** 0.014 0.005** –0.004 0.007 –0.014 0.042 hd: other 0.000 0.000 0.001 0.003 0.003 0.003 –0.010 0.011 com: sec. school 0.016 0.005*** –0.006 0.007 0.026 0.016 0.000 0.020 com: trunk tarmac 0.005 0.003* 0.000 0.007 –0.018 0.010 0.024 0.040 eastern region 0.005 0.001*** –0.050 0.012*** 0.023 0.009 –0.006 0.014 northern region –0.018 0.003*** –0.053 0.010*** –0.041 0.008 –0.022 0.011 western region 0.000 0.000 –0.012 0.013 –0.002 0.002 0.001 0.011 constant 0.168 0.103 –0.097 0.228 Total 0.038 0.011*** 0.372 0.017 –0.002 0.034 0.497 0.037 9% 91% 0% 100% Source: Authors’ calculations based on household survey data. Note: SE = standard error. *p < .10. **p < .05. ***p < .01. Data, Methodology, and Results for Decomposition Analysis 221 Note 1. The twofold decompositions are carried out using the Stata command “oaxaca” written by Ben Jann (2008) using the “weight (0.5)” option. References Blinder, A. 1973. “Wage Discrimination: Reduced Form and Structural Estimates.” Journal of Human Resources 8: 436–55. Cotton, J. 1988. “On the Decomposition of Wage Differentials.” Review of Economics and Statistics 70: 236–43. GLSS (Ghana Living Standards Surveys). 2007. “Pattern and Trends of Poverty in Ghana.” Ghana Statistical Service, Accra. Jann, Ben. 2008. “A Stata implementation of the Blinder-Oaxaca decomposition.” ETH Zurich Sociology Working Papers No. 5. ETH Zurich, Zurich, Switzerland. http://repec.ethz.ch/ets/papers/jann_oaxaca.pdf. Machado, J. A. F., and J. Mata. 2005. “Counterfactual Decomposition of Changes in Wage Distributions Using Quantile Regression.” Journal of Applied Econometrics 20: 445–65. Neumark, D. 1988. “Employers’ Discriminatory Behavior and the Estimation of Wage Discrimination.” Journal of Human Resources 23: 279–95. Oaxaca, R. 1973. “Male-Female Wage Differentials in Urban Labor Markets.” International Economic Review 14: 673–709. Reimers, C. 1983. “Labor Market Discrimination against Hispanic and Black Men.” Review of Economics and Statistics 65: 570–79. United Nations. 2008. World Urbanization Prospects: The 2007 Revision. Department of Economic and Social Affairs, Population Division. New York: United Nations. Index b, f, n, and t denote box, figure, note, and table. A colonial legacy, 91, 92 governance, 96, 98 access to services growth, 80, 80f, 81 across rural-urban continuum, 161, health status in, 87 161fr, 162f land disputes in and around, 80 in cities of Sub-Saharan Africa, 82, 83f, land policies and institutions, 92–94, 95 84–87, 85f, 86t, 88 poverty patterns and trends, 81, 81t developmental significance, 10 slum upgrading effort, 98, 99b energy, 136–137, 161 welfare distribution patterns in, 88, 90t human capital endowments in South welfare inequality in, 84, 100 Asian countries, 10, 131–133, achievement index, 31b, 45, 46t 132t, 133t, 196 Africa location decisions of firms and, 167, Gini coefficient, 82–84 198–199 see also Sub-Saharan Africa; specific as migration incentive, 185–186, country or city 187, 188b age, as factor in rural-urban inequality, policy recommendations, 10, 193, 68, 69 198–199 agglomeration economies, 3, 15–17, 16b, rural-urban inequalities, 10, 82, 83f, 101b, 110, 119, 163, 165, 168t, 85f, 86t 170–171, 185, 198–199 sanitation, 131 agriculture sector in slum settlements, 89b access of poor to agricultural land in see also educational services; health South Asia, 197 services age effects on productivity, 68 Accra contributions to economic growth in access to services in, 82, 83f, 84–87, South Asia, 116–117, 117f, 119f 85f, 86t 223 224 Index agriculture sector (continued) educational access and utilization, economic performance in South Asia, 35, 131 115–118 findings of World Development employment in South Asia, 115, 118, Reports, 6b 119f, 150–153 geographic variation in living standards geographic inequality within country, 69 within, 115, 115f global comparison, 52, 54f growth of nonfarm activities, 163, land ownership distribution and self- 165–166, 167f employment in, 144 health care delivery in, 131, 132 nominal rate of assistance in South Asia, land ownership distribution, 129 127, 128f land policies and institutions, 128 policy reforms and outcomes in South life expectancy, 131 Asia, 126 per capita income, 122n1 poverty reducing effects of growth in, population density, 163 115–116 poverty patterns across rural-urban in poverty reduction strategies, 109 continuum, 159–160, 160f productivity factors in growth of poverty rate, 122n2 nonfarm activities, 163 rationale for investment in periurban recommendations for infrastructure areas, 171, 198–199 investment, 198 sources of rural-urban inequalities, recommendations from WDR reports, 64, 72, 74 192–193 stage of developmental transformation, role in improving rural welfare, 74 110 in Sub-Saharan countries, 52 transport infrastructure, 134, 135, 135f subsidies, 127–128 urbanization effects on rural tariff rates, 126, 126f, 127f poverty, 120 urban agriculture, 68–69 urbanization patterns, 118t, 119, WDR country classification, 2, 5b 161, 162f Armenia, 33, 49n7 See also South Asia barriers to development B access to land ownership, 141–142 differences in rural-urban Bangladesh endowments, 22 access to electricity and water, 136–137, political economy, 11, 199–200 161, 161f in South Asian case studies, 111 agricultural tariffs, 126, 126f, 127f World Development Report agriculture’s share of economic growth, classification, 3 116–117, 117f Bolivia, 47 communications infrastructure, 137 Brazil, 2, 3, 48, 187, 188b concentration of employment and firms, Burkina Faso, 35, 47 165–166 consumption patterns, 114 C contribution of nonfarm activities to income and employment, China 120–121, 121f development challenges, 3 determinants of rural-urban poverty patterns, 27b transformation in, 111 rural-urban transition, 2 distribution of nonfarm activities colonial legacy, 91–92, 142 along rural-urban continuum, communications technology, 137 163–165, 166f competition economic policy reforms and outcomes, among service providers in urban 126 areas, 82 Index 225 location decisions of firms and, E 167–168, 169–170 educational attainment productivity and, 170 characteristics of rural-to-urban consumption migrants, 175, 178, 179, 181f, 184 calculation, 32 as factor in rural-urban inequality, 66, evolution of income inequality patterns 67, 68, 69, 74, 193 and, 3, 25–26, 27b labor market composition and, geography effects, 75 144, 154 internal migration outcomes, 184, 185t policies to increase, 74 as measure of rural-urban welfare, wage premium, 146–147 65, 77n8 see also school enrollment and rural-urban divergence–convergence attendance patterns, 36, 40, 69 educational services rural-urban gap, 25–26, 26f, 33, 34f, intracountry differences in South Asia, 35–36 131–132, 133, 133t, 196 rural-urban ratio, 31b land ownership inequalities and, 142 in South Asian countries, 114 policy recommendations, 193 corruption, 9, 97b significance of, in development, 10 Czech Republic, 27b in South Asia, 131–133, 132t, 133t see also educational attainment; school D enrollment and attendance data sources and analysis, 7 employment Demographic and Health Surveys, 29 in and around cities, 144–146, density of economic activity 165–166 policy recommendations, 5–6 contribution of nonfarm activities in in South Asia, 116b, 122 South Asia, 120–121, 121f WDR classification of developing educational attainment and, 144, 154 countries, 3, 5b, 113 intracountry differences in distance, development challenges South Asia, 118 related to land regulation effects, 150–156, 152t policy recommendations, 6 manufacturing sector, 162–163 in South Asia, 122 migration incentives, 68, 178, 188b WDR classification of developing policies to increase, 74, 75 countries, 3, 5b, 113 poverty reduction strategies, 109 distortions and spillovers proximity to large metropolitan areas findings of World Development and, 163–164 Reports, 7 sectoral, as factor in rural-urban policy considerations, 199 inequality, 66–68 policy outcomes in South Asia, 110, in South Asian agricultural sector, 125–128, 159 115, 118, 119f in rural-urban transformations, 110 termination regulation, 130 urbanization effects on rural poverty, see also labor market 120 energy supply, 136–137, 161 divisions within country as development exchange rate policies, 125–126 challenge expropriation, 128–129 policy recommendations for, 6 in South Asia, 122 G spatial inequality and, 192 WDR classification of developing gender differences countries, 3, 5b, 113 access to education, 131 Dominican Republic, 47 migration motivation, 177–178 226 Index geographic variation in development current inequalities in rural-urban in explanation of rural-urban welfare, 25, 33–36, 34f inequality, 64 current patterns, 15 intracountry living standard differences improvement index, 40–44, 41–43t in South Asia, 115 inequality divergence–convergence within large countries, 2–3 patterns, 4f, 36–44, 37–39t, in Sub-Saharan countries, 64, 75 41–43t, 47–48 World Development Report country national income per capita as factor classification, 3, 7, 54f in inequality outcomes, 44–47, see also global-level rural-urban 46t, 48 transformation; local economic nutritional status differences in, disparities; national-level rural-urban 35, 40, 44 transformation poverty rate differences in, 33–35, 34f, geography effect 36–40 analytical methodology, 65–66, 73 school enrollment patterns, 35, 40, 44 definition, 64 scope of analysis, 8, 21 explanatory power, 64 slum population distribution, 19, 20f implications for policy making, variation among developing countries, 64, 65 9, 17–18, 21 in Sub-Saharan Africa, 64, 66, 75 governance, urban, 96–100, 102n6 Ghana access to services, 82 H access to water, 61, 63f agricultural sector, 52, 53f health and nutritional status changes in rural-urban welfare data sources, 29 differences over time, 69–70, 75 global rural-urban patterns, 34f, 35 child health in, 61 measurement, 32 findings of World Development rural-urban divergence–convergence Reports, 5b patterns, 40, 44 growth, 52, 52t rural-urban inequalities, 87, 87f, 193 lessons from rural-urban transition in South Asian countries, 133 experience, 73 in Sub-Saharan countries, 61, 62–63, local governance, 102n6 62f, 87, 193 as model for analysis, 19, 51 health services poverty patterns and trends, 57–60, 59t, access and outcomes in South Asian 61, 81, 81t countries, 131–133, 132t, 133t, 134f regional distribution of industry and intracountry differences in South Asia, agriculture, 54 131–132, 133, 133t, 196 rural-urban disparities, 17–18, 54, 97 migration for access to, 10 school attendance rates, 61, 62, 62f policy recommendations, 193 sectoral shares of gross domestic see also health and nutritional status product, 55f Honduras, 40 sources of rural-urban inequality, 51, 66, housing costs, migration patterns and, 67f, 68–69, 75 182, 186 urbanization trends, 54–56, 56f, human capital 57, 58t, 93b access to endowments in South Asia, see also Accra; Sub-Saharan Africa 131–133, 133t global-level rural-urban transformation developmental significance, 10 consumption differences in, investments to reduce regional 33, 34f, 36, 40 inequalities, 64 current distribution by World policy recommendations for South Asian Development Report classification, 54f countries, 198 Index 227 source of rural-urban inequality in, rural-urban inequalities, 117, 120, 133 66–68, 74 stage of developmental transformation, Hungary, 27b 2, 3, 110 transport infrastructure, 134–135, 135f urbanization effects on rural I poverty, 120 improvement index, 30, 32b, 40–44, urbanization patterns, 118, 118t, 41–43t 119, 172n3 income See also South Asia contribution of nonfarm activities in Indonesia, 2 South Asia, 120, 121f inequalities, rural-urban global trends, 36 access to electricity, 137, 161, land market restrictions and, 153–154, 161fr, 162f 153t access to water, 61 land ownership inequality and, 142 analytical methodology, 30–32, 48n1, rural-urban convergence-divergence 65–66 patterns and, 4f, 9, 19, 44–47, consumption patterns, 33, 34f 46t, 48 consumption ratio, 26, 26f in South Asia, 114, 122n1 continuum conceptualization of, 10–11, in Sub-Saharan countries, 52 16b, 159–163. see also periurban areas upper-middle-income level as develop- and smaller towns/cities ment threshold, 3, 18, 25–26, 27b convergence effects of rural-to-urban urbanization outcomes, 16, 16f migration, 175 see also consumption data sources, 28–30, 48n2 India development patterns in South Asia, 113 access to electricity, 136–137 in development process, 1, 191–192 agricultural subsidies, 127–128 divergence–convergence patterns, 3, 4f, agricultural tariffs, 126, 126f, 127f 9, 17–18, 21, 26, 27b, 28, 36–44, agriculture’s share of economic growth, 37–39t, 41–43t, 47–48, 69–70, 192 116–117, 117f global patterns, 33–36, 34f communications infrastructure, 137 health and nutritional status, 35, 61, 87, contribution of nonfarm activities to 87f, 193 income and employment, 121, 121f household endowments as factor in, economic growth, 114 65–66, 70–72 economic policy reforms and intracountry differences in South Asia, outcomes, 126 115, 115f, 122, 196 educational access and utilization, in living standards, 25, 115 131, 133 national income per capita correlated Gini coefficient, 114 with, 44–47, 46t, 48 health care delivery in, 133 patterns across rural-urban continuum, internal migration in, 176b, 177, 159–163 178, 197 poverty rates, 27b, 28–29, 33–35, 34f, labor market regulation, 130 57–63, 109, 117, 120, 159–160 land ownership distribution, 129, 142 research needs, 195 land policies and institutions, 128, 129 returns to endowments as factor in, per capita income, 114, 122n1 65–66, 69 poverty patterns, 117, 120 school enrollment and attendance poverty patterns across rural-urban patterns, 35, 61 continuum, 160–161 scope of analysis, 26–28 poverty rate, 122n2 sectoral employment of household head poverty reducing effects of agricultural as factor in, 66–68 growth, 115–116 social tensions and, 18, 113, 192 228 Index inequalities, rural-urban (continued) J sources of, 3, 21–22, 64–65, 65–69, Japan, 147–148 67f, 73, 192–193 in South Asia, 7, 117, 120, 122 in Sub-Saharan Africa, 7, 18–19, 54, K 57–63, 193 Kampala welfare distribution by income group, access to services in, 82, 83f, 84–87, 85f 70–73, 71f colonial legacy, 91, 92 infrastructure corruption in, 97b access to services in Sub-Saharan evolution of land policies and cities, 85f institutions, 94 communications, 137 governance, 96, 98, 100, 102n6 connective, 64, 65, 134–136, 164, growth, 80, 80f, 81 193–194, 198 health status in, 87 economic advantages of urban areas, land policies and institutions, 94, 95 3, 15–16, 82 poverty patterns and trends, 81–82, 81t electricity, 136–137 welfare distribution patterns in, 88–90 as factor in firm start-up decisions, welfare inequality in, 84, 100 166–167, 198–199 Kazakhstan, 35, 49n7 investment to reduce regional inequalities, 64, 65 L policy recommendations, 6, 193–194, 198 labor market rural development patterns, 3 economic advantages of urban rural-urban inequalities, 82 concentration, 15–16, 165 in South Asian case studies, 111, land market restrictions and, 147–148 134–137, 136t, 196 land ownership distribution and innovation, 82 transitions in, 142–147, 143b internal migration in Nepal policies and outcomes in South Asia, access to services as factor in, 129–131 185–186, 187 regulatory interventions, 74–75, consumption outcomes, 184, 185t 130–131, 196 cost of living factors in, 186 rural-urban transformation and, 3, 129 demographic patterns, 179, 180f in Sub-Saharan countries, 52, 53f destination characteristics, see also employment 182–183, 183t land ownership distribution destination choice, 184–188, 187f colonial legacy, 142 educational attainment of migrants, 178, labor market outcomes of, 142–147, 179, 181f 143b, 154–155 employment incentives, 178 rural-urban transformation and, 155 employment outcomes, 179, 181f significance of, for economic ethno-linguistic characteristics of development, 141–142 migrants and, 182, 182t, 186 in South Asia, 129 gender differences in, 177–178 status of rural elites, 142 geographic patterns, 179, 179f, 180f strategies for improving, 155 housing costs and, 182, 186 land policies and institutions income incentives, 184 access of poor to agricultural land, 197 motivation for, 177, 199 ceiling legislation, 128–129 remittance patterns and outcomes, 178 colonial legacy, 92 size of migrant population, 177, 178 effects on rural-urban transitions, travel distance, 184, 186 148–154 trends, 177, 178, 197 employment outcomes, 155–156 Index 229 expropriations, 128–129 convergence-divergence patterns in, intermediate measures in absence of 26, 27b, 28, 48 reform, 9, 194 poverty distribution, 33 as obstacle to rural-urban transition, 111 rural-urban welfare differences, 28–29, policy recommendations for 33, 35–36, 47, 48 South Asia, 197 school enrollment patterns, 35 policy recommendations for Sub- Luyiga, Bernard, 97b Saharan African countries, 194 postindependence evolution in M Sub-Saharan cities, 92–96 reforms and outcomes in South Asia, Madagascar, 33 128–129, 196 manufacturing sector employment, 144, sales market interventions, 129, 146, 147, 151, 162–163 149b, 197 location decisions of firms in, 166–167, significance of, in rural-urban transition, 198–199 9, 194, 196 Maputo spillover effects, 199 access to services in, 82, 83f, 84–87, in Sri Lanka, 148–154 85f, 88 strategies for improving rural-urban colonial legacy, 91–92 transition, 155 governance, 96, 98–100, 102n6 in Sub-Saharan Africa, 75–76, 194 growth, 80, 80f, 81 tenancy laws, 129, 149b, 194 health status in, 87 titling programs, 149b, 194 land policies and institutions of, 94–95 see also land ownership distribution poverty patterns and trends in, 81t, 82 Latin America and Caribbean countries, welfare distribution patterns in, 40, 44, 74, 82–84 84, 88, 100 literacy, in South Asian countries, 131 Mberazze, Zachel, 97b living standards Mexico, 27b, 40, 176b access to services, 10 middle-income countries divergence-convergence patterns, 9, 26, consumption patterns, 33 27b, 28, 63, 192 determinants of rural-urban effects of rural-to-urban migration, inequalities, 48 175–176, 199 national income per capita and geography effect, 64 rural-urban convergence- intracountry differences in South Asia, divergence patterns in, 48 115, 115f nutritional status patterns, 35 measurement, 28, 36 poverty distribution, 33 patterns of variation in South Asia, 113 rural-urban divergence patterns, rural-urban inequalities, 9, 18, 25, 27b, 33, 35 33–36, 34f, 47, 64, 100 rural-urban welfare differences, see also access to services; income 35–36, 47 Living Standards Measurement upper-middle-income level as Study, 29 development threshold, 3, 18, local economic disparities 25–26, 27b access to services and, 10 migration, rural-to-urban causes, 8, 22 benefits for rural areas, 175 development patterns, 3 cost of, 9 divergence–convergence patterns, 4f employment incentives, 68, 178, 188b scope of analysis, 8, 22, 79 implications for urban investment see also Accra; Kampala; Maputo policies, 187–188 low-income countries motives of migrants, 175–177, 188b consumption patterns, 33 network effects in, 176b 230 Index migration, rural-to-urban (continued) sources of inequality in, 21–22 push-pull incentives, 175 variation among developing countries, remittances, 176b, 178 73–75, 110 returns to, 176b Nepal in South Asia, 177–178, 196–197 access to electricity in, 136–137 wage incentives, 65, 68, 81, 188b agricultural employment in, 115, 118 welfare gain incentives, 65–66, 68, 81 agricultural tariffs in, 126f, 127f see also internal migration in Nepal agriculture’s share of economic growth, Morocco, 49n7 116, 117f Mozambique communications infrastructure of, 137 access to services in, 82, 83f, 86t contribution of nonfarm activities to access to water in, 61, 63f income and employment in, 120, agricultural sector, 52, 53f 121, 121f changes in rural-urban welfare economic growth, 114 differences over time, 69–70 economic policy reforms and child health in, 61, 62–63, 62f outcomes, 126 findings of World Development Reports, 6b educational access and utilization, growth, 52, 52t 131–132 lessons from rural-urban transition findings of World Development experience, 73 Reports, 6b local governance, 102n6 geographic variation in living standards as model for analysis, 19, 51 within, 115, 115f per capita income, 52, 52t Gini coefficient, 114 policy recommendations for, 74 health care delivery in, 131, 196 poverty patterns and trends, 57–60, 59t, infrastructure weaknesses, 111 60t, 61, 81t, 82 labor market regulation, 130 recent structural transformations, 52 pattern of development within, 7, 110 regional distribution of industry and per capita income, 114, 122n1 agriculture, 54 poverty patterns, 117, 120, 178, rural-urban disparities, 7, 18, 54, 77n10 195–196 school attendance rates, 61, 62, 62f poverty rate, 122n2 sectoral shares of gross domestic rural-urban inequalities, 48, 117, 120 product, 55f stage of developmental sources of rural-urban inequality, 51, 66, transformation, 110 67f, 68 transport infrastructure, 135–136, 135f urbanization trends, 7, 54, 56, 56f, 57, travel time, 136, 137f 58t, 80, 93b urbanization patterns, 118t, 119–120 welfare distribution by income group, see also internal migration in Nepal; 70, 71f, 72 South Asia see also Maputo; Sub-Saharan Africa Nicaragua, 40 Muganzi, Charles, 97 nutrition. see health and nutritional status O N Oaxaca-Blinder decomposition, national-level rural-urban transformation 65, 67f, 73 case studies, 8. see also specific country within country variation, 9, 110 P inequality distribution, 17–18 inequality divergence–convergence Pakistan patterns, 4f access to electricity and water in, rationale for case studies based on, 110 136–137, 161, 162f scope of analysis, 8, 21 agricultural tariffs in, 126f, 127f Index 231 agriculture’s share of economic growth, Poland, 27b 116–117, 117f policies to manage urbanization communications infrastructure challenges, 199–200 of, 137 constraints to effective urban contribution of nonfarm activities to governance, 96–100 income and employment in, 120, determinants of rural-urban 121, 121f inequalities, 48 economic policy reforms and educational and health services outcomes, 126 investment, 193 educational access and utilization in, findings from Sub-Saharan case studies, 131, 196 192–193 effects of land ownership distribution in, findings of World Development Reports, 142–147, 154–155 3–6, 5–6b findings of World Development goals, 191 Reports, 6b growth of nonfarm activities outside of geographic variation in living standards urban areas, 171 within, 115, 115f human capital investment, 64, 65, health care delivery in, 131, 132, 196 74, 198 human capital investment in, 198 implications of internal migration internal migration in, 177 patterns for, 187–188 labor market regulation in, 130 implications of welfare gain distribution land ownership distribution in, 129, for, 72 142–147, 196, 197 infrastructure investment, 64, 65, land policies and institutions, 128 193–194, 198 life expectancy, 131 investment in periurban areas and occupational transformation patterns, smaller cities, 198–199 142–147 job creation, 74 per capita income, 122 n1 land market interventions, 155, 197 poverty patterns across rural-urban lessons from transition experiences of continuum, 160, 160f South Asian countries, 155, 195–200 poverty rate in, 122 n2 lessons from transition experiences of rural land inequality in, 10–11 Sub-Saharan countries, 73–76 stage of developmental mix of policies, 10, 198–199 transformation, 110 political economy factors, 11, 194, transport infrastructure, 134, 135, 135f 199–200 urbanization patterns in, 118, 118t, 119, postindependence attitudes toward 120, 161, 162f urban development, 93b wage patterns in, 146–147, 154–155 rationale for regional integration, 101b See also South Asia recommendations for, 22, 100–102, periurban areas and smaller towns/cities 192–195 agglomeration benefits, 165, 170–171 regulatory interventions in labor market, attraction of new manufacturing firms 74–75 to, 162–163, 166–170 research needs, 22 growth of nonfarm activities in, rural bias in Africa, 93b, 96–98 165–166 second-round impacts, 10–11 lessons from case studies, 10–11, sequence of implementation, 10 198–199 significance of, in development nonfarm employment and wages in, outcomes, 9, 192 163–164 spillover considerations, 199 poverty distribution, 159–161 stage of developmental transformation rationale for investment in, 171, and, 3–6 198–199 see also land policies and institutions 232 Index political economy country classification, 2, 2f, 54f, 110, 113 implications for reform, 11, 199–200 determinants of, in South Asia, intermediate measures in absence of 111, 125, 138 reform, 9, 194 development outcomes in Sub-Saharan land reform and, 155, 194, 197 Africa, 18–19 poverty rates distortions and spillovers in, 110, 199 calculation of, 28–29 effects of land policies and institutions data sources, 29 on, 128, 147–154 economic density and, 116b findings of World Development Reports, effects of agricultural growth in India 1–6, 7, 109 on, 115–116 future prospects, 15 nonfarm employment in South Asia and, geographic variation in, 3, 109 120–121, 121f Gini index, 60 remittance effects on, 178 global. see global-level rural-urban rural-to-urban migration effects on, 176b transformation in rural-urban continuum, 159–160, 160f interactions between rural and urban rural-urban differences in, 27b, 28–29, areas in, 7, 110 33–35, 34f, 57, 109, 117, 120 labor market linkage, 129 rural-urban divergence–convergence land ownership inequality and, 142, 155 patterns, 36–40 local level. see local economic disparities rural-urban population shifts and, 60 in middle-income countries, 27b in South Asia, 114, 114t, 116b, 117, national level. see national-level 120, 122n2, 195–196 rural-urban transformation strategies for reducing, 109 pace of, 2–3 in Sub-Saharan Africa, 19, 57–63, 59t, patterns and trends in South Asia, 113, 60t, 81–82 118–120, 118t, 121–122, 125, urbanization effects, 19, 120 161–162, 195, 197 productivity patterns and trends in Sub-Saharan across urban-rural continuum, 170–171 countries, 54–57, 56f, 58t advantages of urban concentration, 3, policies to manage. see policies to 15–16, 165 manage urbanization in agglomeration economies, 170–171, positive feedback effects in, 15–16 170t postindependence attitudes, 93b agricultural, as source of rural welfare poverty trends and, 60 improvement, 74 predictability of development competition and, 170 outcomes, 21 labor market regulation and, 130–131 research needs, 22 land ownership distribution and, rural poverty reduction and, 120 141–142 significance of, in economic purchasing power parity exchange rates, development, 16, 191 28–29 stages of transformation in, 2, 3, 5b, 110 trade policy effects, 125–128 variation among developing countries, R 9, 21, 73–75, 110, 192, 197 remittances, 176b, 178 see also inequalities, rural-urban; road density in South Asia, 134–135 migration, rural-to-urban rural-urban transition Russian Federation, 27b challenges and opportunities, 1 connective infrastructure and, 134, 198 S contribution of nonfarm activities in South Asian economies, school enrollment and attendance 120–121, 121f measurement, 32 Index 233 rural-urban patterns, 35, 40, 44 patterns of rural-urban transformation, in Sub-Saharan countries, 61, 62, 62f 113, 118–120, 118t, 121–122, 125, see also educational attainment; 161–162, 162f, 195, 197 educational services population growth and distribution service economy employment, patterns in, 114, 119 144, 147, 151 poverty patterns in, 27b, 114t, 116b, Singapore, 102 117, 120, 122n2 size of cities, 101b rural-urban convergence-divergence investment in smaller cities, 199 patterns in, 44, 117 migration patterns and, 187 scope of analysis, 8, 110, 113–114 size of country, 2–3 transportation infrastructure in, slum settlements, 19, 20f, 89b, 98, 134–136, 136t, 196 99b, 194, 195 WDR classification of development social tensions status, 5b, 110, 117, 122, 195 development patterns in South Asia see also specific country and, 113 Sri Lanka problems of within-country inequalities, access to electricity in, 136–137 18, 192 agricultural employment in, 115, 118, South Africa, 48 150–153 South Asia agricultural subsidies in, 127–128 access of poor to agricultural land in, 197 agricultural tariffs in, 126, 126f, 127f access to electricity in, 136–137, 136t agriculture’s share of economic growth communications infrastructure of, 116–117, 117f of, 136t, 137 communications infrastructure in, 137 consumption patterns in, 114 determinants of rural-urban density of economic activity in, 116b transformation in, 111 development patterns in, 7, 121–122 economic growth of, 114 economic growth in, 114, 114t economic policy reforms and outcomes economic performance of agricultural in, 125, 148–150, 196 sector in, 115–118, 119f educational access and utilization in, economic policy reforms and outcomes, 131, 132, 133 125–128, 196 employment patterns in, 118, 150–154, findings from country-level case studies, 152t, 155 111, 195–200 findings of World Development Reports, Gini coefficient, 114, 114t 6b growth of nonfarm activities in, Gini coefficient, 114 120–121, 121f health care delivery in, 131, 132 human capital factors in development internal migration in, 177, 178, 187 of, 10 irrigation investment outcomes in, 11 internal migration in, 177–178, 196–197 labor market regulation in, 130 intracountry differences in access to land policies and institutions of, 9, 111, services, 131–132, 196 129, 148–154, 155, 196, 199 intracountry differences in connective life expectancy in, 131 infrastructure, 135–136 living standards, geographic differences intracountry differences in within, 115, 115f employment, 118 per capita income, 114, 122n1 labor market regulation, 129–131, 196 poverty rate, 122n2, 195 land ownership distribution, 129, 196 stage of developmental transformation land policy reforms and outcomes in, of, 110 128–129 transport infrastructure in, 134, 135, literacy rates, 131 135f, 136 living standards, 114 travel times in, 136, 138f 234 Index Sri Lanka (continued) access to water in, 61, 63f urbanization patterns in, 118t, 119 agricultural sector, 52, 53f within-country variation, 9, 115, 115f, causes of rural-urban inequality in, 118, 151, 195 66, 67f, 68, 73–74 See also South Asia changes in rural-urban welfare start-up and location decisions of firms, differences over time in, 69–70 166–170, 198–199 child health in, 61 stunted growth. see health and findings of World Development Reports, 6b nutritional status growth of, 52, 52t Sub-Saharan Africa land policies and institutions of, 95–96 development patterns in, 7, 18–19 lessons from rural-urban transition human capital factors in development experience of, 73–74 of, 10 local governance in, 102n6 lessons from case studies in, 192–195 as model for analysis, 19, 51 policies to manage rural-urban per capita income in, 52, 52t transformation in, 19 policy recommendations for, 74 population trends, 19 poverty patterns and trends in, 57–60, poverty patterns, 27b 59t, 60t, 61–62, 81–82, 81t rural-urban convergence-divergence recent structural transformations in, 52 patterns in, 44 regional distribution of industry and scope of analysis, 8 agriculture in, 54 sources of rural-urban inequalities in, rural-urban disparities in, 17–18, 48, 54 64–65 school attendance rates in, 61, 62f urbanization trends in, 15, 21 sectoral shares of gross domestic product WDR classification of development of, 55f status, 5b sources of rural-urban inequalities in, 51 see also Accra; Ghana; Kampala; urbanization trends in, 54, 56–57, 56f, Maputo; Mozambique; Uganda 58t, 93b welfare distribution by income group in, T 70–72, 71f, 72–73 see also Kampala; Sub-Saharan Africa Tanzania, 33, 47, 176b United States, 176b tariff rates, 126, 126f, 127f upper-middle-income levels, 3, 18, 25–26, taxation, 98 27b technology investment, land ownership urban areas distribution and, 141–142 access to services in, in Sub-Saharan tenancy laws, 129, 149b Africa, 84–87, 85f, 86t trade policies agriculture in, 68–69 liberalization patterns in South Asia, causes of welfare disparities, 90–91 125–126 characteristics of slums, 89b rural-urban transformation in South Asia constraints to effective governance in, and, 126–126 96–100 tariff rates, 126, 126f, 127f development variations among transforming economies (WDR developing countries, 9 classification), 2 diseconomices of congestion, definition, 5b 165–166, 194 global distribution, 54f economic advantages of concentration, policy recommendations for, 3 3, 165 in South Asia, 110, 122 employment patterns in and around, 144–146, 165–166 U global population distribution, 15 Uganda global slum population distribution, access to services in, 82, 83f, 86t 19, 20f Index 235 location decisions of firms in, 167–170, land ownership inequality and, 198–199 146–147 number of firms in, 165–166 migration incentives, 65, 68, 81, 184, policy recommendations to improve, 188b 3, 100–102, 194–195 proximity to large metropolitan areas proximity effects on wages and and, 163–164 employment, 163–164 returns to age, 68 rationale for regional integration, 101b sources of inequalities in rural-urban slum upgrading efforts, 98, 99b transition, 3 sources of growth, 15 waste management, 86 sources of welfare increases in, 70 water supply and access in Sub-Saharan Africa, 19, 21 in cities of Sub-Saharan Africa, 84, 88 taxation in, 98 rural-urban differences in Sub-Saharan undernutrition in, 35 Africa, 61, 63f, 82 WDR classification of country WDR 2008, Agriculture for Development, economices, 5b, 54f 1–3, 7, 52, 109, 113, 191, welfare distribution patterns in, 82–87, 192–193 88–90, 100, 194 WDR 2009, Reshaping Economic see also Accra; inequalities in Geography, 1, 2, 3–6, 7, 18, 21, 25, urbanization; Kampala; Maputo; peri- 79, 109, 113, 191 urban areas and smaller towns/cities; World Development Reports, 7, 18, 21, 25, rural-urban transition 52, 79, 109, 113, 191. see WDR 2008, urbanization. see rural-urban transition Agriculture for Development; WDR 2009, Reshaping Economic Geography W World Institute for Development Economics Research of the United wages Nations University, 27b educational attainment and, 146–147, 154 Z effects of land regulation on, 153–155 individual-level variables in Pakistan, 147t Zamindari system, 142 ECO-AUDIT Environmental Benefits Statement The World Bank is committed to preserving Saved: endangered forests and natural resources. • 3 trees The Office of the Publisher has chosen to • 1 million BTUs of print From Farm to Firm: Rural-Urban total energy Transition in Developing Countries on recy- • 240 pounds of net cled paper with 50 percent postconsumer fiber greenhouse gases in accordance with the recommended stan- • 1,081 gallons of dards for paper usage set by the Green Press waste water Initiative, a nonprofit program supporting • 69 pounds of solid publishers in using fiber that is not sourced waste from endangered forests. For more informa- tion, visit www.greenpressinitiative.org. Around the world, countries are becoming urbanized at an astonishing pace. As countries develop economically, their economies shift from mainly rural and agrarian to increasingly urban and nonagricultural. This rural-urban transformation presents both opportunities and challenges for development. When managed effectively, the transformation spurs growth and reduces poverty. When managed poorly, however, the process can result in stark welfare disparities, the marginalization of entire regions, and poorly functioning cities that fail to realize the potential gains from agglomeration economies. From Farm to Firm: Rural-Urban Transition in Developing Countries investigates the rural-urban transformation underway in Sub-Saharan Africa and South Asia, emphasizing the influence of country conditions as well as the potential of good policies to minimize disparities and ensure that everyone shares in the benefits of urbanization. The first part of this book investigates urbanization and rural-urban welfare inequalities on three geographic scales—global, national, and local—featuring countries and cities in Sub-Saharan Africa on the national and local levels. The second part of the book sheds light on the texture of transformation in five countries in South Asia, each at a different stage in the process: Bangladesh, India, Nepal, Pakistan, and Sri Lanka. Policy makers and development experts, as well as anyone concerned with the impact of rural-urban transition on growth and equity, will find this book thought-provoking and informative. ISBN 978-0-8213-8623-1 SKU 18623