43738 Reshaping economic geography world development report 2009 Reshaping Economic Geography world development report 2009 Reshaping Economic Geography THE WORLD BANK Washington, DC © 2009 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 E-mail: feedback@worldbank.org All rights reserved 1 2 3 4 5 12 11 10 09 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 boundaries, colors, denominations, and other information shown on any map in this work do not imply any judgement on the part of The World Bank 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. Softcover ISSN: 0163-5085 ISBN: 978-0-8213-7607-2 eISBN: 978-0-8213-7608-9 DOI: 10.1596/978-0-8213-7607-2 Hardcover ISSN: 0163-5085 ISBN: 978-0-8213-7640-9 eISBN: 978-0-8213-7608-9 DOI: 10.1596/978-0-8213-7640-9 Cover design and interior navigational graphics by Zefyr Design, info@zefyr.co.uk. Typesetting, graphs, and page layout by Precision Graphics. Printed in the United States by Quebecor World. Contents Foreword xiii Acknowledgments xv Abbreviations and Data Notes xvi Geography in motion: The Report at a Glance--Density, Distance, and Division xix Overview 1 Place and prosperity 1 The world is not flat 8 Markets shape the economic landscape 12 Putting development in place 20 Navigating This Report 33 Scope 34 Terms 34 Structure 39 Geography in motion: Overcoming Distance in North America 44 Part One Seeing Development in 3-D 47 1 Density 48 Defining density 49 Economic concentration--the richer, the denser 56 Convergence--rural-urban and within cities 62 What's different for today's developers? 70 2 Distance 73 Defining distance 74 Economic concentration in leading areas 81 Divergence, then convergence--between leading and lagging areas 84 What's different for today's developers? 93 v vi CONTENTS 3 Division 96 Defining division 97 Economic concentration 105 Divergence, then convergence 109 Geography, globalization, and development 115 What's different for today's developers? 120 Geography in motion: Overcoming Division in Western Europe 122 Part Two Shaping Economic Geography 125 4 Scale Economies and Agglomeration 126 A guide to scale economies 129 A different realm 132 A portfolio of places 135 Apprehension of market forces 140 5 Factor Mobility and Migration 146 From mercantilism to globalization to autarky, and back again 147 Labor mobility: learning from a generation of analysis 158 Practical policies for managing migration 168 6 Transport Costs and Specialization 170 What has happened: two centuries of experience 173 Transport costs and scale economies: two decades of analysis 181 What to do: transport policies in the developing world 184 Transport: an increasingly important sector 192 Geography in motion: Distance and Division in East Asia 194 Part Three Reframing the Policy Debates 197 7 Concentration without Congestion: Policies for an inclusive urbanization 198 Principles for managing a portfolio of places 200 A framework for integration 202 The framework in action 216 A strategy for inclusive urbanization 229 8 Unity, Not Uniformity: Effective approaches to territorial development 230 People seek opportunities 231 Countries seek unity 234 Contents vii A policy framework for integrating lagging and leading areas 238 The framework in action 245 Avoiding Balkanization: the political benefits of economic integration 258 9 Winners without Borders: Integrating poor countries with world markets 260 Regional integration to scale up supply, global integration to scale up demand 262 Building integrated neighborhoods: a framework 265 The framework in action 273 Geography in motion: Density, Distance, and Division in Sub-Saharan Africa 283 Bibliographical Note 287 Endnotes 291 References 305 Selected Indicators 331 Table A1 Geography and access 332 Table A2 Urbanization 335 Table A3 Territorial development 338 Table A4 International integration 340 Table A5 Other indicators 343 Sources and definitions 346 Selected World Development Indicators 349 Data sources and methodology 349 Classification of economies and summary measures 350 Terminology and country coverage 350 Classification of economies by region and income, FY2009 351 Table 1 Key indicators of development 352 Table 2 Millennium Development Goals: eradicating poverty and improving lives 354 Table 3 Economic activity 356 Table 4 Trade, aid, and finance 358 Table 5 Key indicators for other economies 360 Technical notes 361 Index 369 viii CONTENTS Boxes 1 Three geographic scales: local, national, and 4.4 When sowing and reaping happen in different places: international 3 rising interdependence of cities 139 2 The three dimensions of development: density, distance, 4.5 Cities continue to thrive as telecommunication and division 6 costs fall 140 3 Intraindustry trade and intermediate inputs 20 4.6 Hong Kong, China: market forces led the way, 4 New insights from a generation of analysis 21 government followed 142 5 Concentration without congestion in western China: 4.7 Reinvention and renewal: how New York became a Chongqing and Chengdu 26 great city 143 0.1 What this Report is not about 34 5.1 Regional labor mobility has been falling in Sub-Saharan Africa 152 0.2 This Report's regions are more detailed than the World Bank's 36 5.2 Cross-border migration in the Greater Mekong Subregion 153 0.3 This Report's message is not anti-equity 39 5.3 From facilitating to restricting to (again) facilitating 0.4 Fresh insights from economic geography: labor mobility in China 154 concentration, convergence, and integration 41 5.4 Labor and social policies restrain migration in Eastern 1.1 Two laws and a rule: the empirical regularities of a Europe--not good for growth 160 country's city-size distribution 52 5.5 From Lewis to Lucas: the economic perspective on 1.2 The Republic of Korea's portfolio of places 53 migration has changed 161 1.3 Computing the agglomeration index 55 5.6 Implicit barriers to mobility: place-based entitlement 1.4 Africa's urbanization reflects industrialization 59 and divisions in India 163 1.5 Urbanization and narrowing rural-urban disparities 5.7 Why did the U.S. South take so long to catch up? in the Islamic Republic of Iran 63 Division. 165 1.6 Slums, then and now 68 5.8 Migrating to economic density: rational decisions or 1.7 Many of today's world-class cities were littered bright lights? 167 with slums 69 5.9 Too early to tell? The impact of African emigrants 2.1 Defining an area: impossible or NUTS? 78 on Africa 168 2.2 How developed and developing countries define 6.1 Biggest in the world: size and social obligations of lagging areas: a quick survey 79 Indian Railways 175 2.3 Dangerous disparities: when divisions aggravate 6.2 The jet engine 177 distance 80 6.3 The big box 178 2.4 Correcting geographic disparities in postwar Japan 87 6.4 Italy's intervento straordinario: an unexpected response to 2.5 Spatial inefficiency and the downfall of the falling transport costs 184 Soviet Union 90 6.5 Mobility with density in Hong Kong, China 185 3.1 A country's neighborhood matters: regional integration 6.6 Neighborhoods matter, but so do trade and transport and growth spillovers 102 policies 188 3.2 Bolivia and Chile's border--from wide to narrow? 102 6.7 Unclogging Latin America's arteries: transport costs now 3.3 The benefits of size 103 matter more than tariffs 192 3.4 Artificial states? 105 7.1 Are the policy messages of this Report antirural? 3.5 Market access and per capita incomes 108 No. 200 3.6 Neighborhoods matter: Southern Cone versus 7.2 Land reform to jump-start urbanization: aiding villeins in Southern Europe 115 Denmark 203 3.7 The influence of first-nature geography: is it possible to 7.3 Bombay fights the markets, and more than half of eradicate malaria? 117 Mumbai's residents live in slums 205 3.8 Integration takes a long time, and its benefits do not come 7.4 Widening the reach of New York City 208 overnight 119 7.5 Promoting concentration in Japan between 1860 and 1980: 4.1 Scale economies in an almost unreal world: the story of spatially connective policies for Tokyo-Yokohama and Dongguan, China 127 Osaka-Kobe 209 4.2 Sharing, matching, and learning 129 7.6 Climate change calls for a different urban form, not slower urbanization 211 4.3 Agglomeration economies in Indonesia 132 Contents ix 7.7 Speeded up, but still in sequence: spatial integration in 8.6 Low market access in Mexico's lagging south 244 twentieth-century Hong Kong, China 213 8.7 Universal primary education in Uganda increased access 7.8 Singapore: from slums to world city 215 to schools in the northern areas 248 7.9 Titling land for a sustainable rural-urban 8.8 Improving the spatial progressivity of Nigeria's transformation 219 intergovernmental transfers 249 7.10 Land markets in transition 219 8.9 Exporting by mail in Peru--connecting small producers 7.11 Strengthening land market institutions for rural-urban to markets 251 integration 220 8.10 Special economic zones bring growth if they exploit 7.12 Retrofitting transport infrastructure in Bangkok 223 advantages in natural and economic geographies 254 7.13 New cities: escapes from urban jungles, or cathedrals 9.1 Are the policy messages of this Report anti­global in the desert? 224 integration? No. 261 7.14 Rural-urban integration in Beijing, Guangzhou, and 9.2 Diversifying production through regional Shanghai 227 cooperation 263 7.15 Slum upgrading and prevention: what works? 229 9.3 Economic partnership agreements between the EU and African, Caribbean, and Pacific countries can be 8.1 Are the policy messages of this Report"anti-decentralization"? made better 265 No. 231 9.4 Integration in Central Asia 278 8.2 The German integration: convergence and concentration with mobile labor 235 9.5 Integrating the small and distant Pacific Islands with world markets 278 8.3 Constitutions call for unity, not uniformity 236 9.6 A contract with Africa? The give and take of the world's 8.4 An instrument per objective: Ireland used EU funds for biggest development challenge 282 international convergence 237 8.5 Taxation against spatial inequality? The U.S. federal income tax system 239 Figures G0.1 The Report can be read by part or by policy xxi 1.5 Developing countries have a pace of urbanization similar to 1 At all three geographic scales, the patterns of concentration that of early developers 58 of economic activity are similar 9 1.6 Density intensifies rapidly in the early phase of urbanization 2 At all three geographic scales, the patterns of convergence in before leveling off 59 living standards are similar 11 1.7 Shares of population living in urban agglomerations rise 3 Later developers face a different world 13 with the level of development 60 4 Intraindustry trade is high in North America, Western 1.8 Geographic concentration of population, gross Europe, Oceania, and East Asia 19 product, and household consumption rises sharply with development, then levels off 61 5 In charted waters: the pace of urbanization today has precedents 24 1.9 Rural-urban disparities in GDP per capita tend to be smaller in richer OECD countries 63 6 Quicker geographic convergence in basic living standards in Malaysia 27 1.10 Rural-urban gaps in per capita consumption become smaller with urbanization 65 7 Northeast, Southeast, and South Asia have been catching up to developed nations 30 1.11 Even at the subnational level, rural-urban disparities fall as density increases 66 0.1 A navigational aid for the reader 43 1.12 Slums grow with the pace of urbanization, and fall with its 1.1 From dichotomy to continuum: a portfolio of places 51 level 66 1.2 Almost a law: relative size distributions of settlements 1.13 Urbanization's speed has precedents 70 remain stable over time 51 1.14 The population increment in urban areas of today's 1.3 The agglomeration index helps to compare urbanization developing countries is much larger 71 across regions 57 2.1 Manufacturing activity flourishes in areas with shorter 1.4 The richer a country, the more concentrated its economic economic distance to density 77 mass 58 2.2 Rising density of economic mass accompanies development over decades, even centuries 83 x CONTENTS 2.3 Measures based on national accounts, household surveys, ys, 3.15 Divergence, then convergence in East Asia, 1950­ and geoscaled economic data confirm the historical pattern 2006 112 of a rising concentration of economic mass with the level of 3.16 The East Asian growth experience had two distinct development 84 phases 113 2.4 Spatial inequality rose and remained high before slowly 3.17 Western Asia and Eastern Europe have had little declining as economies approached $10,000 in GDP per integration--and little convergence 114 capita 86 3.18 Mexico and other LAC countries have not been catching up 2.5 Subnational disparities in income and wages persisted for with the United States 114 more than 70 years in Canada and France 86 G2.1 The stairway to success 123 2.6 Spatial disparities have narrowed slowly in Europe since G2.2 Border effects between the European Union and the United een World War II 87 States remain more than twice that within the European 2.7 Contemporary comparisons of countries indicate that Union 123 disparity in welfare among subnational areas fall with 4.1 The urban premium for household consumption can be economic development 88 considerable 141 2.8 Economic growth in East Asia and Eastern Europe is faster 5.1 International capital flows have surged since the than the world's growth 89 1970s 148 2.9 Disparities in per capita gross product have been rising 5.2 A large share of capital now flowsdirect investment large share of capital now as as direct investment 148148 between leading and lagging areas in Southeast Asia een 89 5.3 Migrants from East Asia, Latin America, and the Middle 2.10 Steady rise in inequality of per capita provincial gross East and North Africa go mainly to OECD countries, but product in China since 1990 89 most in South Asia and Sub-Saharan Africa stay close to 2.11 Income disparities between areas widened as Eastern een home 151 European nations moved from plan to market 91 5.4 In Latin America and the Caribbean internal migrants are 2.12 In Malaysia, geographic convergence in basic welfare more educated than those who stay behind 156 accompanied economic growth 92 5.5 Internal migrants are more schooled than workers in the 2.13 Today's developing countries face a more integrated places to which they move 157 world 93 5.6 Migration from developing countries is becomingoming more 3.1 The number of borders between nations tripled in the past een skilled 158 50 years 97 5.7 Internal labor mobility and economic growth often en go 3.2 Rich countries tend to have lower border restrictions 99 together 162 3.3 Tariffs are highest in Africa, South Asia, and Western 5.8 Internal migrants are more likely to work than Asia 99 natives 164 3.4 Capital restrictions are highest in Africa, South Asia, and 6.1 Intraindustry trade is becoming oming more important for all Central Asia 100 types of goods, but not in all world regions 171 3.5 Residents of richer countries face fewer visa er 6.2 Air freight costs are down less than ocean freight requirements 101 costs 179 3.6 Globally, language diversity is highest near the 6.3 The costs of communicating have fallen to a fraction of equator 104 what they were a decade ago 180 3.7 Concentration increases at the global level, then a leveling 6.4 Distance has become more of a deterrent, divisions ome er off 106 less 182 3.8 Only Asia's share in world GDP has risen noticeably since 6.5 Global intraindustry trade is no longer confined to rich 1980 107 countries 183 3.9 The effect of distance between Brazil and its trading een 6.6 Transport accounts for about one-seventh of CO2 partners has remained considerable 109 emissions 191 3.10 Education outcomes have improved ed 110 7.1 The growth of cities has been grossly overestimated 199 3.11 East and South Asia have been the only regions catching 7.2 The dimensions increase with the level of up 110 urbanization 201 3.12 Life expectancy decreased significantly in many African 8.1 France has benefited from increasing concentration of countries 111 economic production and declining spatial disparities in 3.13 Education has become more equal since the 1980s ome 111 disposable income 238 3.14 Slight global divergence in per capita incomes, 1950­ 8.2 Living standards can vary considerably between leading and een 2006 112 lagging areas 247 Contents xi Maps G0.1 Density--why it pays to be close to Tokyo xix 6.1 Busy seafaring in the North, little landfall in the G0.2 Distance--why Americans must be mobile xx South 172 G0.3 Division--what prevents progress in Africa does not in 6.2 Crossing borders or climbing walls? 187 Western Europe xx G3.1 Asia divided: conflict in the middle of the nineteenth G0.4 How markets view the world xxi century 195 1 The biggest development challenges--at the local, national, G3.2 Asia integrated: trade at the end of the twentieth and international geographic scales 4 century 196 2 Settlements of varying size facilitate different scale 7.1 The Republic of Korea--three areas at different stages of economies 14 urbanization 217 3 Migrating to reduce distance to density: Despite 7.2 China--three areas at different stages of the obstacles, Chinese workers have migrated in the urbanization 218 millions 16 7.3 Decentralization and localization economies in 4 Migrating to reduce distance to density: Brazil's young Daegu 221 workers move in thousands to get closer to economic 7.4 Changsha, Zhuzhou, and Xiangtan--spatially connective density 17 infrastructure in a two dimensional area 222 5 Migrating to reduce distance to density: Migration in India 7.5 Bogotá's TransMilenio has helped to integrate the has been less frenetic 17 poor 225 6 As urbanization advances, policies must evolve 25 7.6 Economic density in Seoul with good connections to other 7 Three types of countries, differing challenges for area cities 225 development 28 8.1 The poverty rate is high in China's western interior, but 8 Market access distinguishes world regions 31 most poor people live closer to economic density in the East 232 0.1 Three geographic scales--area, country, and region 35 8.2 The poverty rate is high in Brazil's Northeast and G1.1 The U.S. geographic center of population gravity moved Amazon areas, but the poor are massed in areas along the 1,371 kilometers between 1790 and 2000 44 coast 233 G1.2 America's large cities are in the Northeast and on the two 8.3 Both the poverty rate and poverty mass are high in some of coasts 45 India's lagging states 234 1.1 The landscape of economic mass is bumpy, even in a small 8.4 In the Russian Federation, population densities are country like Belgium 50 highest in the economically vibrant and warmer west, 1.2 Local divisions--spatial disparities within urban but a communist legacy has left some people in the cold settlements can be large 67 interior 242 2.1 Access to markets is not a straight line 75 8.5 In the Lao People's Democratic Republic, the poverty rate 2.2 Distances can be long even in the developed world 76 is high outside the capital region, but poor people are 2.3 Vietnam's poverty rate is higher in lagging inland areas, but scattered in remote communities 243 its poverty mass is greater in leading coastal areas 81 8.6 In the Arab Republic of Egypt, population densities are high 2.4 The poverty rate is high in distant eastern Honduras, but in lagging areas, and connective infrastructure is needed to the poor are concentrated in the two largest metropolitan help spatial integration 245 areas 82 9.1 Mobile phone coverage has spread rapidly in Africa 268 2.5 Exporting industries concentrate in coastal areas to 9.2 Density, distance, and division combine to determine access minimize distance to the global market 94 to markets 271 3.1 Global GDP is concentrated in a few world regions, 9.3 Potential access to major world markets distinguishes the 2006 97 developing world's regions 272 3.2 Some borders are much wider than others 98 9.4 Building regional infrastructure in Southern Africa 277 3.3 Forty-three countries do not have direct access to the 9.5 Twin cities for local integration 279 coast 103 9.6 West Africa has potential for cotton-led industrial 3.4 Language diversity is very high in Africa 104 development 280 G2.1 The division in Western Europe has gradually G4.1 Africa's borders were charted before World War I 285 dissipated 122 xii CONTENTS Tables 1 Density is most important locally, distance nationally, and 5.2 Close to home: the largest international flows of labor are division internationally 7 between neighboring countries 150 2 Agglomeration, migration, and specialization are the most 5.3 Rates of labor mobility vary widely across countries in the important forces--and land, labor, and intermediate inputs developing world 156 the most sensitive factor markets 22 5.4 Most migrants move for economic reasons, but many are 3 "An I for a D?" A rule of thumb for calibrating the policy pushed out by poor services 166 response 23 5.5 What does a practical policy toward migration do? 1.1 The size of urban settlements grows with Recognize agglomeration benefits. 169 development 61 6.1 Prices, costs, and profit margins are all high on Africa's 1.2 Rural-urban disparities in earnings, wealth, and transport corridors 187 consumption characterize development over the last two 6.2 Time costs for crossing borders are highest in Central Asia, centuries 64 Central Africa, East Africa, and Southern Africa 189 1.3 Rural-urban disparity in basic services narrows with 7.1 An instrument per dimension--a simple framework for development 65 urbanization policies 216 2.1 Administrative, statistical, and geographic area measures all 7.2 Interventions to integrate the urban poor 228 point to rising spatial concentrations of economic activity 8.1 An instrument per dimension--a framework for area, with development 85 territorial, or regional development policies 246 2.2 Spatial inequality varied through different phases of 8.2 The OECD's experience with incentives to stimulate lagging development 86 areas has been evolving 256 2.3 Household survey and subnational gross product data 8.3 A range of instruments has been used by governments corroborate the pattern of declining spatial disparities in to create economic mass in lagging areas, with modest welfare with development 88 results 257 3.1 The concentration of GDP and population growth shifted 8.4 Assessing the performance of area development between 1820 and 1998 106 policies 259 3.2 GDP per capita increased tenfold, 1500­1998 109 9.1 Few regional agreements provide for full mobility of 4.1 A dozen economies of scale 128 labor 266 4.2 Internal scale economies are low in light industries and high 9.2 Sub-Saharan Africa, South Asia, and the Middle East and in heavy industries 130 North Africa are most affected by unreliable infrastructure, 4.3 Scale economies amplify with density and attenuate with East Asia the least 267 distance 135 9.3 Regional "club goods" can easily be provided because 4.4 Thirty years of theoretical advance recognize the costless exclusion is possible 269 importance of scale economies 136 9.4 An instrument per dimension--a simple framework for 4.5 Agglomeration economies vary by city size and profile, and regional integration 273 by the industry life cycle 137 G4.1 The most distant and divided regions--trading and 5.1 In the late-nineteenth century most international migrants transport are expensive 285 came from better-off Europe 149 Foreword Production concentrates in big cities, leading provinces, and wealthy nations. Half the world's production fits onto 1.5 percent of its land. Cairo produces more than half of Egypt's GDP, using just 0.5 percent of its area. Brazil's three south-central states comprise 15 percent of its land, but more than half its production. And North America, the European Union, and Japan--with fewer than a billion people--account for three-quarters of the world's wealth. But economic concentration leaves out some populations. In Brazil, China, and India, for example, lagging states have poverty rates more than twice those in dynamic states. More than two-thirds of the developing world's poor live in villages. A billion people, liv- ing in the poorest and most isolated nations, mostly in Sub-Saharan Africa and South and Central Asia, survive on less than 2 percent of the world's wealth. These geographically disadvantaged people cope every day with the reality that develop- ment does not bring economic prosperity everywhere at once; markets favor some places over others. But dispersing production more broadly does not necessarily foster prosperity. Economically successful nations both facilitate the concentration of production and insti- tute policies that make people's living standards--in terms of nutrition, education, health, and sanitation--more uniform across space. Getting the benefits of both economic concen- tration and social convergence requires policy actions aimed at economic integration. Integration should begin with institutions that ensure access to basic services such as primary education, primary health care, adequate sanitation, and clean drinking water for everyone. As integration becomes more difficult, adaptive policies should include roads, rail- ways, airports, harbors, and communication systems that facilitate the movement of goods, services, people, and ideas locally, nationally, and internationally. For places where integration is hardest, for social or political reasons, the response should be commensurately comprehen- sive, with institutions that unite, infrastructure that connects, and interventions that target, such as slum upgrading programs or incentives for producers to locate in certain areas. Using these principles, World Development Report 2009, the 31st in the series, reframes the policy debates on urbanization, territorial development, and regional integration. The report analyzes the early experience of developed countries and draws practical implications for urbanization policies in today's developing countries. For the poorest countries in Africa and Asia that are landlocked or otherwise isolated from world markets, the Report discusses promising approaches to regional integration that combine institutional cooperation, shared infrastructure, and special incentives. In growing middle-income economies, general pros- perity can camouflage areas of persistent poverty. For such countries, the Report outlines strategies to foster domestic integration and help the poor in the least fortunate places. I expect that Reshaping Economic Geography will stimulate a much-needed discussion on the desirability of "balanced growth," which has proved elusive. And by informing some important policy debates, it will point the way toward more inclusive and sustain- able development. Robert B. Zoellick President xiii Acknowledgments This Report has been prepared by a team led by Indermit S. Gill, comprising Souleymane Coulibaly, Uwe Deichmann, Maria Emilia Freire, Chorching Goh, Andreas Kopp, Somik V. Lall, Claudio E. Montenegro, Truman Packard, and Hirotsugu Uchida. Important con- tributions were made by Homi Kharas, Marisela Montoliu Munoz, Andrew Nelson, Mark Roberts, Sebastian Vollmer, and Fang Xu. The team was assisted by Eduardo S.F. Alves, Brian Blankespoor, Maximilian Hirn, Siobhan Murray, and Catalina Tejada. Bruce Ross-Larson was the principal editor. The maps were created by the World Bank's Map Design Unit under the direction of Jeff Lecksell. Book production and printing were coordinated by the World Bank's Office of the Publisher, under the supervision of Stephen McGroarty, Susan Graham, Rick Ludwick, and Andres Ménèses. World Development Report 2009 is co-sponsored by the Development Economics Vice Presidency (DEC) and the Sustainable Development Network (SDN). The work was conducted under the joint guidance of François Bourguignon, Alan H. Gelb, and Justin Yifu Lin, DEC Senior Vice Presidents at various stages of the production and dissemination of this Report, and Katherine Sierra, SDN Vice President. Jean-Jacques Dethier, Jeffrey Lewis, Claudia Paz Sepulveda, Laszlo Lovei, and Antonio Estache also provided valuable comments. World Bank Presidents Robert B. Zoellick and Paul D. Wolfowitz provided guidance and advice, and Managing Directors Ngozi Okonjo-Iweala and Graeme Wheeler have been invaluable sources of encouragement to the team. Two panels of advisers provided excellent advice at all stages of the Report. The Aca- demic Panel was chaired by the Chief Economist and DEC Senior Vice President and con- sisted of François Bourguignon, Paul Collier, Masahisa Fujita, Vernon Henderson, Philippe Martin, Ravi Kanbur, Lord Nicholas Stern, and Anthony Venables. The Policymaker Panel was chaired by the SDN Vice President and consisted of Lobna Abdellatif Ahmed, Newai Gebreab, Jerzy Kwiecinski, Shantong Li, Katharina Mathernova, Charbel Nahas, Enrique Peñalosa, Carolina Renteria, Kamal Siddiqui, Jorge Wilheim, and Natalia Zubarevich. Many others outside and inside the World Bank contributed with comments and sug- gestions. The team benefited greatly from many consultations, meetings, and regional workshops held locally and in Côte d'Ivoire, Finland, France, Germany, India, Japan, Nor- way, Russia, South Africa, Sweden, Tanzania, Turkey, and the United Kingdom. The team wishes to thank participants in these and other workshops, videoconferences, and on-line discussions, which included academics, policy researchers, government officials, and staff of nongovernmental, civil society, and private sector organizations. The team would like to acknowledge the generous support of the United Kingdom's Department for International Development, the multi-donor Knowledge for Change Pro- gram, the Government of Norway, the Japan Policy and Human Resources Development Trust Fund, the Institute for Environment and Sustainability in the Joint Research Centre of the European Commission, Germany's Federal Ministry for Economic Cooperation and Development, and the William and Flora Hewlett Foundation. Rebecca Sugui served as senior executive assistant to the team, Ofelia Valladolid as pro- gram assistant, and Jason Victor and Maria Hazel Macadangdang as team assistants. Evan- geline Santo Domingo served as resource management assistant. xv Abbreviations and Data Notes ACP Africa, Caribbean, and Pacific AMU Arab Maghreb Union ASEAN Association Southeast Asian Nations BELDES Municipal Infrastructure Support Project (Turkey) CACM Central American Common Market CARICOM Caribbean Community CEFTA Central European Free Trade Agreement CESIN Center for International Earth Science Information Network CKLN Caribbean Knowledge and Learning Network DR-CAFTA Dominican Republic­Central America Free Trade Agreement EAC East African Community EAP East Asia and the Pacific ECA Europe and Central Asia ECOWAS Economic Community of West African States ECSC European Coal and Steel Community ECTEL Eastern Caribbean Telecommunications Authority EEC European Economic Community EMU European Monetary Union ENEA École Nationale d'Économie Appliquée (Dakar) ENSEA École Nationale de Statistique et d'Economie Appliquée (Abidjan) EPA economic partnership agreements EU European Union FDI foreign direct investment FEU forty-foot equivalent units GATS General Agreement on Trade in Services GDP gross domestic product GIS geographic information system GNI gross national income GRP gross regional product GRUMP Global Rural-Urban Mapping Project IBRD International Bank for Reconstruction and Development ICT information and communication technology IDA International Development Association IIED International Institute for Environment and Development (UK) INEGI Instituto Nacional de Estadística y Geografía IOM International Organization of Immigration ISSEA Institut Sous-Régional de Statistique et d'Économie Appliquée (Yaoundé) IT information technology KÖYDES Village Infrastructure Support Project (Turkey) xvi Abbreviations and Data Notes xvii MERCOSUR Southern Common Market (Latin America) NAFTA North American Free Trade Agreement NEPAD New Partnership for Africa's Development NSDP National Slum Development Program (India) NUTS Nomenclature of Territorial Units for Statistics OECD Organisation for Economic Co-operation and Development OEEC Organization for European Cooperation PAFTA Pan-Arab Free Trade Area PPS purchasing power standard R&D research and development RASCOM Regional African Satellite Communication Organization SADC Southern African Development Community SAR South Asia region SASEC South Asia Sub-regional Economic Cooperation SEZ special economic zone SIC Standard Industrial Classification. SPARTECA South Pacific Regional Trade and Economic Cooperation Agreement TEU twenty-foot equivalent units TFP total factor productivity UPE universal primary education VAMBAY Valmiki Ambedkar Awas Yojana WAEMU West African Economic and Monetary Union WTO World Trade Organization Data notes The countries included in regional and income groupings in this Report are listed in the Classification of Economies table at the beginning of the Selected World Development Indicators. Income classifications are based on gross national income (GNP) per capita; thresholds for income classifications in this edition may be found in the Introduction to Selected World Development Indicators. Group averages reported in the figures and tables are unweighted averages of the countries in the group, unless noted to the contrary. The use of the word countries to refer to economies implies no judgment by the World Bank about the legal or other status of a territory. The term developing countries includes low- and middle-income economies and thus may include economies in transition from central planning, as a matter of convenience. The terms industrialized countries or devel- oped countries may be used as a matter of convenience to denote high-income economies. Dollar figures are current U.S. dollars, unless otherwise specified. Billion means 1,000 million; trillion means 1,000 billion. Geography in motion The Report at a Glance--Density, Distance, and Division Growing cities, ever more mobile people, and increasingly specialized products are integral to development. These changes have been most noticeable in North America, Western Europe, and Northeast Asia. But countries in East and South Asia and Eastern Europe are now experiencing changes that are similar in their scope and speed. World Development Report 2009: Reshaping Eco- nomic Geography concludes that such transformations will remain essential for economic success in other parts of the developing world and should be encouraged. Seeing development in 3-D the start of winter in some parts of the cans change states every year, migrating These transformations bring prosper- country, so flights often are canceled. to reduce their distance to economic ity, but they do not happen without But Americans put up with the pain of opportunity. The most striking feature risk and sacrifice. Look at three of the leaving friends and family, because eco- of this map is distance. world's most prosperous places: nomic activity is concentrated in a few Across the Atlantic, in Western parts of the country (see map G0.2). To Europe, another massive movement · The first is Tokyo, the largest city in get a part of this wealth, you have to get takes place every day--not of people the world with 35 million people, closer to it. That is why 8 million Ameri- but of products. One example is Airbus, a quarter of Japan's population, packed into less than 4 percent of its Map G0.1 Density--why it pays to be close to Tokyo Economic production per square kilometer in Japan land. · The second is the United States, the largest economy in the world and perhaps also the most mobile, where Sapporo about 35 million people change resi- JAPAN dences each year. · The third is Western Europe, the most connected continent in the world today, where countries trade Tokyo about 35 percent of their gross domestic product (GDP), more than half among neighbors. Visitors to Tokyo can see people Nagoya being crushed into trains by profes- Osaka sional train-packers. Millions of peo- ple willingly subject themselves to the unpleasantness of such a crush. A map of Japan's economic density shows why. Hiroshima Tokyo generates a big part of Japan's wealth--to get a share of it, people Kitakyushu have to live close by (see map G0.1). Fukuoka The most striking feature of this map is density--the concentration of wealth in Tokyo and Osaka. In the United States, each year in the days before the Thanksgiving holi- day, about 35 million people try to get Source: WDR 2009 team and World Bank Development Research Group based on subnational GDP estimates for back to their families and friends. It is 2005. See also Nordhaus (2006). xx WORLD DEVELOPMENT REPORT 2009 Geography in Motion Map G0.2 Distance--why Americans must be mobile tage of economic density. Despite the Economic production per square kilometer in the United States crush among commuters and in such UNITED STATES slums as Dharavi, Mumbai's popula- tion has doubled since the 1970s. Since the 1990s, millions of Chinese work- Portland Seattle Chicago New ers have migrated to get closer to eco- Washington, DC York nomic opportunity concentrated along the coast. Just as Americans travel dur- ing Thanksgiving, more than 200 mil- Denver Boston lion people in China travel during the Chinese New Year. Regional produc- Phoenix Dallas- tion networks in East Asia are spread San Ft. Worth Francisco far wider than Airbus sites in Western Miami Europe. East Asian countries may not Los Houston trade airplane parts, but nations that Angeles once were enemies now trade parts of cars and computers with the same fre- Source: WDR 2009 team and World Bank Development Research Group based on subnational GDP estimates for 2005. See also Nordhaus (2006). quency and speed. And what is the payoff? We can which makes parts of planes and assem- a region that was divided not so long again recognize the shapes of China, bles them in France,Germany,Spain,and ago now trade with former enemies to India, and Southeast Asian countries the United Kingdom as well as in other become an ever-more-integrated Euro- on the map of the world's economic countries. Huge sections of aircraft are pean Union (EU). As this integration geography (see map G0.4). Contrast loaded onto ships and planes, as places has increased, economic divisions have these shapes with that of the mighty specialize in making different parts and decreased, making specialization and continent of Africa, which shows up as producing them in scale. Countries in scale possible (see map G0.3). a slender peninsula. What is the payoff for this pain? A The World Development Report Map G0.3 Division--what prevents progress map of economic geography, which argues that some places are doing well in Africa does not in Western Europe resizes the area of a country to reflect because they have promoted transfor- Border restrictions to flows of goods, capital, people, and ideas its GDP, shows the benefits of big cit- mations along the three dimensions of ies, mobile people, and connected economic geography: countries. The United States, Western · Higher densities, as seen in the Europe, and Japan dominate the world's growth of cities. economy (see map G0.4). · Shorter distances, as workers and Cities, migration, and trade have businesses migrate closer to density. been the main catalysts of progress in · Fewer divisions, as countries thin the developed world over the past two their economic borders and enter centuries. These stories are now being world markets to take advantage of repeated in the developing world's most scale and specialization. dynamic economies. The United States and Japan · Mumbai is not the largest city in reshaped their economic geography the world, but it is the most densely along these lines in the past. China is populated. And it keeps growing. reshaping its economic geography now. · China is not the largest economy in This Report proposes that these will be the world, but it is the fastest growing the changes that will help developing and may be among the most mobile. nations in other parts of the world, · Southeast Asia may not have formed most notably Africa. a political union like Europe, but it trades parts of goods back and forth Unbalanced growth,inclusive Source: WDR 2009 team (see chapter 3 for details). as the EU does. development Note: The width of borders is proportional to a sum- mary measure of each country's restrictions to the People risk loss of life or limb on That is what this World Development flow of goods, capital, people, and ideas with all other countries. Mumbai's packed trains to take advan- Report proposes, and the Report is Geography in Motion xxi Map G0.4 How markets view the world A country's size shows the proportion of global gross domestic product found there Source: WDR 2009 team using 2005 GDP (constant U.S. dollars). Note: The cartogram was created using the method developed by Gastner and Newman (2004). This map shows the countries that have the most wealth when GDP is compared using currency exchange rates. This indicates international purchasing power--what someone's money is worth if spent in another country. structured to bring out the message (see interest all countries--small and large, Africa is how to make a better economic figure G0.1). low income and middle income. The union that benefits both the landlocked chapters on distance, factor mobility, and the coastal countries, the poorest · Part one describes the changes along and regional development may be of and the more prosperous. the dimensions of density, distance, most interest to larger middle-income As the geographic scale increases and division--taking up each in countries. And the chapters on division, from local to national to international, turn. It summarizes the experience transport costs, and regional integration the specific policy issue changes. But of the past century or so. may be of most interest to low-income the underlying problem is the same-- · Part two analyzes the drivers of these and smaller economies. some places do well, others do not. And transformations--the market forces Four spotlights on Geography in it is difficult for anyone to accept this as of agglomeration, migration, and Motion examine the interplay between inevitable. specialization and trade. It distills market forces and government policies The Report's main message is that the findings of policy research dur- in North America, Western Europe, economic growth will be unbalanced. ing the past generation or so. East Asia, and Sub-Saharan Africa. By To try to spread out economic activ- · Part three discusses the policy impli- highlighting the interactions among the ity is to discourage it. But development cations of the experience and analy- three dimensions, they also connect the can still be inclusive, in that even people sis in the first two parts. It provides Report's different parts. who start their lives far away from eco- a common framework for reframing Seen another way, the Report exam- nomic opportunity can benefit from three policy debates--on urbaniza- ines the most important policy issues the growing concentration of wealth in tion, on lagging areas within coun- of economic geography, from local, to a few places. The way to get both the tries, and on regional integration national, to international. Locally, the benefits of uneven growth and inclu- and globalization. policy issue in areas such as Lagos state sive development is through economic The Report is structured and written in southern Nigeria is how to manage integration. in such a way that people interested in urbanization. Nationally, the policy only one of these debates can read just issue in Nigeria is how to manage the Economic integration--local, some of it. That is, it can be read verti- disparities in resources and living stan- national,and international cally. The chapters on density, agglom- dards in the north and the south. And The Report makes it clear what eco- eration, and urbanization should internationally, the policy issue in West nomic integration means. It means one xxii WORLD DEVELOPMENT REPORT 2009 Geography in Motion aph Figure G0.1 The Report can be read by part or by policy 1 DENSITY 2 DISTANCE GIM 1: 4 3 NORTH AMERICA AGGLOMERATION DIVISION AMERICA 5 FACTS MIGRATION 7 6 GIM 2:WESTERN WESTERN URBANIZATION SPECIALIZATION EEUROPE UROPE 8 TERRITORIAL DEVELOPMENT ANAL ANALYSIS YSI 9 GIM GIM S REGIONAL 3:EAST INTEGRATION ASIA POL POLICIES ICIES GIM GIM 44::SUB-SAHA SUB-SAHARAN RANAFRICA AFRI CA Source: WDR 2009 team. thing to integrate rural and urban areas, · In places urbanizing rapidly, gov- integrate areas within nations, while and slums with other parts of cities. It ernments must put in place, in reducing poverty everywhere. means another to integrate lagging addition to institutions, connective Lagging areas have one thing in and leading provinces within a nation. infrastructure so that the benefits common--they are economically dis- And it means yet another to integrate of rising economic density are more tant from places doing well. But besides isolated and well-connected countries. widely shared. A good example is this, the economic geography of differ- These notions of economic integration Chongqing, China. ent areas is not the same: are central to three debates in develop- · In places where urbanization has ment--urbanization, territorial devel- advanced, in addition to institutions · In some countries, such as China, opment, and international integration. and infrastructure, targeted inter- lagging areas are sparsely populated. ventions may be necessary to deal It does not make a lot of sense to Urbanization with slums. But these interventions spread expensive infrastructure into The arguments and evidence in World will not work unless institutions for these places--or to give firms incen- Development Report 2009 can set pri- land and basic services are reason- tives to move to them. What makes orities for policies at different stages of ably effective and transport infra- much more sense is to provide basic urbanization, essentially providing the structure is in place. A good example services everywhere, even if it costs e, elements of an urbanization strategy. is Bogotá, Colombia. more to reach these distant areas. Each territory or area within a nation Encouraging mobility of people is has a specific geography. But the prin- Territorial development the priority, and institutions that ciples are quite universal. The principles also can reshape the make land markets work better and er · In places mostly rural, governments debate on territorial or regional develop- provide security, schools, streets, and should be as neutral as possible and ment. The tools of geography can iden- sanitation should be the mainstay of should establish the institutional tify which places are poor--the lagging integration policy. foundation for possible urbaniza- areas--and where most of the poor live. · In other countries,such as Brazil,lag- other countries, such as Brazil, tion in some places. Good land poli- Often, the two are not the same, because ging areas are densely populated. But lagging areas are densely populated. cies are central, and so are policies to the poor have the most reason to move in China, poor people have moved As in China,poor people have provide basic services to everyone. A from poor places. The Report discusses in the millions from the northeast good example is Costa Rica. how governments can tailor policies to to the southeast. Everyone speaks Geography in Motion xxiii the same language, and domestic possibility is providing incentives to markets attractive to investors, face a mobility is not difficult. But many agriculture and allied activities that more difficult challenge. Good insti- poor people still live in the north- are appropriate for states that are tutions and regional infrastructure east. Encouraging mobility of people still mostly rural. can help them access these markets. from the northeast is important, but Examples include East Asia and, Regional integration so is enabling access to markets in increasingly, South Asia. Southern Finally, the principles developed in this the dynamic southeast. In such cases, Africa and South America can also Report inform the debates on how to both institutions and infrastructure integrate globally by making their make globalization work for all coun- to connect the two coastal areas are home markets bigger and more spe- tries. The same logic applied at the necessary for economic integration. cialized through regional institu- local and national levels can be used at · In a third group of countries, such as tions and infrastructure. the international level to classify world India, lagging areas are densely pop- · Integration is hardest for countries regions by the difficulty of economic ulated--almost 60 percent of India's in regions that are divided, are dis- integration in these regions. The com- poor live in these poor places--and tant from world markets, and lack mon problem is division--thick eco- people can find it difficult to migrate the economic density provided by a nomic borders. Aside from this, the task to places doing well, such as the cap- large local economy. These countries of integration varies in different parts ital area and the south. Language include those nicknamed the "bot- of the developing world: and cultural differences within tom billion"--East, Central, and West some areas can be considerable. In · Countries in regions close to world Africa; Central Asia; and the Pacific such cases, institutions and infra- markets, such as Central America, Islands. For these countries, all three structure could be complemented by North Africa, and Eastern Europe, instruments are needed--regional incentives to producers to locate in face a relatively straightforward task institutions that thin borders,regional these lagging states. But these incen- of integration. Common institutions infrastructure that connects countries, tives should be carefully designed to can help them become extensions of and such incentives as preferential avoid offsetting the unifying effects these large markets. access to world markets, perhaps con- of common institutions and con- · Countries in regions distant from ditioned on ensuring that all countries nective infrastructure. A promising world markets, but with large home strengthen regional cooperation. One thing is common to the policy debates on urbanization, area development, and globalization. In their current form, they over- emphasize geographic targeting--what to do in rural areas or in slums, what to do in lagging states or remote areas, and what to do in the most poor or landlocked countries. The Report reframes these debates in a way that better conforms to the reality of growth and development. The reality is that the interaction between leading and lagging places is the key to economic development. The reality is that spatially targeted interventions are just a small part of what governments can do to help places that are not doing well. The reality is that, besides place-based incentives, governments have far more potent instruments for integration. They can build institutions that unify all places and put in place infrastructure that connects some places to others. The Report calls for rebalancing these policy discussions to include all the instruments of integration--institutions that unify, infrastructure that connects, and interventions that target. And it shows how to use the three dimensions of density, distance, and division to tailor the use of these policy instruments to address integration challenges that range from the relatively straightforward to the most complicated. Overview Economic growth will be unbalanced, but development still can be inclusive--that is the mes- sage of this year's World Development Report. As economies grow from low to high income, production becomes more concentrated spatially. Some places--cities, coastal areas, and con- nected countries--are favored by producers. As countries develop, the most successful ones also institute policies that make living standards of people more uniform across space. The way to get both the immediate benefits of the concentration of production and the long-term benefits of a convergence in living standards is economic integration. Although the problems of economic integration defy simple solutions, the guiding principle does not have to be complex. The policy mix should be calibrated to match the difficulty of the development challenge, determined by the economic geography of places. Today, policy discus- sions about geographic disparities in development often start and end with a consideration of spatially targeted interventions. The Report reframes these debates to include all instruments for economic integration--institutions, infrastructure, and incentives. The bedrock of integration efforts should be spatially blind institutions. As the challenges posed by geography become more difficult, the response should include connective infrastructure. In places where integration is hardest, the policy response should be commensurately comprehensive: institutions that unite, infrastructure that connects, and interventions that target. Place and prosperity would earn about three times as much in the Place is the most important correlate of a United States. A Nigerian with nine years of person's welfare. In the next few decades, a education would earn eight times as much in person born in the United States will earn a the United States than in Nigeria. This "place hundred times more than a Zambian, and live premium" is large throughout the develop- three decades longer. Behind these national ing world.1 The best predictor of income in averages are numbers even more unsettling. the world today is not what or whom you Unless things change radically, a child born know, but where you work. in a village far from Zambia's capital, Lusaka, will live less than half as long as a child born Bumps, curves, and spills in New York City--and during that short These disparities in incomes and living stan- life, will earn just $0.01 for every $2 the New dards are the outcome of a striking attribute Yorker earns. The New Yorker will enjoy a of economic development--its unevenness lifetime income of about $4.5 million, the across space. Somewhat unfairly, prosper- rural Zambian less than $10,000. ity does not come to every place at the same A Bolivian man with nine years of time. This is true at all geographic scales, schooling earns an average of about $460 from local to national to global. Cities per month, in dollars that reflect purchas- quickly pull ahead of the countryside. Liv- ing power at U.S. prices. But the same person ing standards improve in some provinces 1 1 2 WORLD DEVELOPMENT REPORT 2009 while others lag. And some countries grow Less widely appreciated is the fact that to riches while others remain poor. If eco- places near prosperous provinces, coun- nomic density were charted on a map of tries, and regions have invariably benefited. the world, the topography at any resolution Prosperity produces congestion and causes would be bumpy, not smooth. economic activity to spill over, but only to Location remains important at all stages places that are well connected to these pros- of development, but it matters less for living perous parts. The detrimental effects of standards in a rich country than in a poor poverty, instability, and conflict spill over one. Estimates from more than 100 living as well. To prosperous places, proximity is standard surveys indicate that households a blessing, to poor places, a curse. in the most prosperous areas of developing These three attributes of development-- countries--such as Brazil, Bulgaria, Ghana, geographic unevenness, circular causation, Indonesia, Morocco, and Sri Lanka--have and neighborhood effects--have not always an average consumption almost 75 percent received much attention. They should, higher than that of similar households because they have radical implications for in the lagging areas of these countries. public policy. Compare this with less than 25 percent for such developed countries as Canada, · Geographic unevenness--the first attri- Japan, and the United States. In contrast, bute of development--implies that as a country grows richer, location becomes governments generally cannot simulta- more important for economic production. neously foster economic production and Ghana, Poland, and New Zealand--three spread it out smoothly. medium-size countries with land areas of · Circular causation--the second attri- about 250,000 square kilometers--have bute--provides hope for policy makers vastly different per capita gross national wishing to pursue progressive objectives. incomes of about $600, $9,000, and $27,000, Rising concentrations of economic pro- respectively. The most economically dense duction are compatible with geographic 5 percent of the country's area produces convergence in living standards. And the about 27 percent of gross domestic product market forces of agglomeration, migra- (GDP) in Ghana, 31 percent in Poland, and tion, and specialization can, if combined 39 percent in New Zealand. with progressive policies, yield both a Put another way, as countries develop, concentration of economic production location matters less for families and more and a convergence of living standards. for firms. Development seems to give a place · Neighborhood effects--the third attri- the ability to reap the economic advantages bute--come with a principle for policy of rising concentrations of production, and making: promote economic integration. to obtain the social benefits that come from Unevenness and circularity imply that it a convergence in consumption. Economic is more difficult for places left behind to development thus brings with it the condi- catch up. But spillovers point to the prom- tions of even greater prosperity, in a virtu- ise for surmounting this handicap. Eco- ous circle. nomic integration is an effective and the Another stylized fact: neighborhoods most realistic way to harness the immedi- matter. A prosperous city seldom leaves its ate benefits from concentration to achieve periphery mired in poverty. A province's the long-term benefits of convergence. prosperity is sooner or later shared with those nearby. And neighboring countries Putting this principle of economic inte- share not just political borders but eco- gration into practice requires identifying nomic destinies. North America, Western the market forces and government poli- Europe, and East Asia are now prosperous cies that best support the concentration neighborhoods. Within these regions, all of economic mass and the convergence of countries did not grow in lockstep. Within living standards across different locations. countries, some provinces did better, and It also requires recognizing that these mar- within each province, prosperity came at ket forces can be strong or weak depend- different times to cities, towns, and villages. ing on economic geography. Earlier World Overview 3 Development Reports have studied these phe- manifest in urbanization. In East Asia, for nomena. This Report advances the influence example, if current trends continue, the of geography on economic opportunity by urban population is expected to increase elevating space and place from mere under- by about 450 million people over the next currents in policy to a major focus. two decades, as countries in the regions grow, adding the equivalent of a Paris The problem--at three geographic scales every month. In South and Central Asia, Depending on the "geographic scale," the the increase is expected to be almost 350 market forces to be harnessed or supported million. And in Sub-Saharan Africa--if differ. At a smaller scale--say, an area within economies continue to grow--the urban a country (a province or state)--geography population could increase by 250 million poses different challenges than at a larger between 2005 and 2025. In other parts of geographic scale--say, a country. At an even the developing world, within-urban trans- larger geographic scale--say, a group of formations will be as important. countries that form a geographic region--the The question is whether growing con- market forces that work toward integration centrations of humanity will increase can be blocked by even greater geographic prosperity, or produce congestion and and political obstacles (see box 1). squalor. Another concern is the diver- Locally, the concentration of eco- gence in living standards between those nomic production as countries develop is who benefit most from this geographic BOX 1 Three geographic scales: local, national, and international Consider the "neighborhoods" of Lagos agglomeration economies and rural- of religious and linguistic di erences. State, Nigeria, and West Africa (see the urban migration. The sharing of oil wealth is a source of maps below). · The second geographic scale is the coun- tension. · The rst geographic scale is the area. try. With its 36 states and capital area · The third geographic scale is the region. The state of Lagos in southwestern and covering 924,000 square kilome- Nigeria's West African neighbors Nigeria has the ve districts of Badagry, ters, Nigeria is the world's 32nd larg- include Cameroon, the Central African Epe, Ikeja, Ikorodu, and Lagos, cover- est country. The distance from Lagos Republic, Côte d'Ivoire, Equatorial ing about 3,500 square kilometers. Its to the northeastern tip of Nigeria is Guinea, Ghana, Niger, and Togo. The estimated population density--with almost 1,500 kilometers. The southern region covers more than 6 million the smallest land area but among the states have seaports and oil elds. The square kilometers, divided by some of two most populous in the nation--is northern part, once a seat of ancient the world's thickest borders. about 2,600 persons per square kilome- empires, now has higher poverty. ter. Metropolitan Lagos has a density Migration between the north and the more than three times that, fueled by south is not an easy matter because Source: WDR 2009 team. Three geographic units: area, country, and region Lagos State, Nigeria, and West Africa represent the local, national, and international scales The first geographic scale The second geographic scale The third geographic scale The area around Lagos State The country of Nigeria The West African region Lagos LAGOS STATE LAGOS STATE Source: WDR 2009 team. 4 WORLD DEVELOPMENT REPORT 2009 Map 1 The biggest development challenges--at the local, national, and international geographic scales a. A billion in slums b. A billion in remote areas c. The bottom billion Population (millions) WDR regions > 250 100­250 50­100 25­50 < 25 Sources: Panel a: United Nations 2006a; panel b: WDR 2009 team, based on household survey data; panel c: Collier 2007. Overview 5 concentration--essentially urbanites in is that they also had more than half of the prosperous neighborhoods--and those left world's population; the European Union behind in villages and those living in slums, (EU), Japan, and the United States have less estimated to number about 1 billion in the than one-sixth. developing world (see map 1, panel a). The Today, the worry at the international (ineffective) policy responses so far have level is the high poverty, illiteracy, and been to try to slow down urbanization. mortality in some parts of the world, set At the national scale, economic growth against the prosperity, literacy, and longev- displays a similar unevenness, as places close ity in others. The policy responses include to large markets prosper sooner than places foreign aid and multilateral efforts to ease more distant. In China the coastal provinc- international trade and investment flows. es--mainly in the three areas known as the But barriers to the agricultural exports of Bohai Basin, the Pearl River Delta, and the developing countries remain considerable, Yangtze River Delta--accounted for more and apathy for people distant or distinct than half of the country's GDP in 2005, renders aid flows miniscule. Aid will be a with less than a fifth of its area. In Brazil small part of the solution. Even in the Euro- the south-central states of Minas Gerais, pean Union, with a combined GDP of about Rio de Janeiro, and São Paulo account for 8 trillion, annual aid through the struc- more than 52 percent of the country's GDP, tural and cohesion funds will average less with less than 15 percent of its land area. than 50 billion between 2007 and 2013. Greater Cairo produces 50 percent of the Foreign aid is less than 0.5 percent of the Arab Republic of Egypt's GDP, using just gross national income of giving countries, 0.5 percent of its land area. and not even a large fraction of the GDP Politicians generally view this economic of countries home to the "bottom billion" imbalance disapprovingly. In communist who have 12 percent of the world's popula- Russia the government labored to reduce tion, but less than 1 percent of its GDP (see the economic share of the old industrial map 1, panel c).2 area of St. Petersburg, the Center, and the A billion slum dwellers in the developing mid-Urals from 65 percent to 32 percent, world's cities, a billion people in fragile lag- forcibly shifting production to the eastern ging areas within countries, a billion at the areas. It boosted the share of the east in eco- bottom of the global hierarchy of nations-- nomic production from 4 percent in 1925 these overlapping populations pose today's to 28 percent at the end of communism, biggest development challenges. Seemingly whose demise probably was hastened by the disparate, they share a fundamental feature: spatial inefficiency that these efforts engen- at different spatial scales, they are the most dered. Because governments care so much visible manifestation of economic geogra- about domestic disparities, they jeopardize phy's importance for development. competitiveness and risk collapse. Policies Concern for these intersecting 3 billion to reduce interstate or provincial dispari- sometimes comes with the prescription that ties in production and living standards are economic growth must be made more spa- commonplace--but largely ineffective. tially balanced. The growth of cities must About 1 billion people continue to live in be controlled. Rural-urban gaps in wealth these inhospitable lagging areas (see map 1, must be reduced quickly. Lagging areas and panel b). provinces distant from domestic and world At the international scale, economic markets must be sustained through territo- growth has concentrated global produc- rial development programs that bring jobs tion in a few regions, with commensurate to the people living there. And growing differences in incomes. In 2000 about gaps between the developed and developing three-quarters of world GDP was concen- world must be addressed through interven- trated in North America, Western Europe, tions to protect enterprises in developing and Northeast Asia. This concentration is countries until they are ready to compete. not new. Three centuries ago, China and World Development Report 2009 has India accounted for about two-thirds of a different message: economic growth the world's wealth. What was different then is seldom balanced. Efforts to spread it 6 WORLD DEVELOPMENT REPORT 2009 BOX 2 The three dimensions of development: density, distance, and division This Report uses three geographic dimen- area's economic density (GDP per square Division. But the journey is not yet sions to describe the transformation of kilometer) was 35 times that of Nigeria complete. The goods must surmount economies as they develop (part one) in 1990. The capacity of the tanneries in the division caused by di erences in and the conditions to keep in mind when and around the city even makes it worth- currencies and conventions between formulating policies (part three). The while to illegally import live animals--the Nigeria and Europe. Between December words are easy metaphors, since density, most important intermediate input-- 2007 and March 2008, Nigeria's currency distance, and division summon images of from neighboring countries. But Kano is depreciated from 170 naira to 1 to human, physical, and political geography. neither large enough, nor rich enough, 180 naira, but appreciated from 246 naira But they can be measured. Consider this to consume more than a little of what is to the British pound in November 2007 illustration. produced. The goods must be exported to 235 naira in March 2008. Buyers and In 2003 Nigeria had 45 million goats and to people willing to pay enough to make sellers of leather goods have to contend kids, 28 million sheep and lambs, and 15 production worthwhile. with these uctuations. They must also million cattle. In a typical year 8 million Distance. Wealthy Europeans want deal with two sets of laws and customs. sheep, 7 million goats, and 0.5 million goods made with "Morocco leather," a The United Kingdom has 30 procedures cattle are slaughtered, mostly in ve north- lot of which comes from Kano. To get to for enforcing a contract, Nigeria 39. These ern states including Kano. More than half Europe, Kano's bulky exports must travel divisions multiply the costs of doing the hides are consumed as pomo. The rest through Lagos, which along highways business. Few cargo ships make landfall are sold to tanneries. The demand from and railways is about 1,000 kilometers in Lagos, so it costs much more to trans- tanneries exceeds local supply, so animals away. It might as well be 4,000 kilome- port goods from Lagos than from busier are imported from nearby Chad, Niger, ters. A railway goes to Lagos through places such as Shanghai. It costs less than and Cameroon. Goat and sheep skins are the cities of Kaduna and Ibadan, but it is $400 to ship a container to the United good business--in 2001 Nigeria produced narrow gauge and poorly maintained. Kingdom from China, more than $1,000 30 million to 35 million of them, exporting Most commerce is by road, obstructed by from Nigeria. almost all to Europe. roadblocks and piracy. Shipping compa- Low local density, costly internal dis- Density. Consider the market condi- nies charge more than $1,200 for a 30-ton tances, and international divisions conspire tions for a tannery that produces leather trailer from Kano to Lagos. Once the against Kano. Making matters more di cult in the city of Kano in Northern Nigeria. goods get to Lagos, there are port fees, are religious and other divisions within O cially, the population of Kano State is pilferage, and delays. It takes 26 days to Nigeria. about 9 million, large enough to provide get the goods onto a ship. The economic the skilled labor and infrastructure for distance from Kano to Lagos, measured Sources: World Bank 2007; Phillips, Taylor, its tanneries. Due to the concentration as money, is several times the Euclidean Sanni, and Akoroda, (FAO 2004); Govern- of people in and around Kano city, the (straight-line) 829 kilometers. ment of Nigeria 2003. prematurely will jeopardize progress. Two convergence. Part three proposes the prin- centuries of economic development show ciple of economic integration--between that spatial disparities in income and pro- places that producers prefer and places duction are inevitable. A generation of eco- where people live--to guide policy making. nomic research confirms this: there is no Using this principle, it reframes the debates good reason to expect economic growth to on urbanization, territorial development, spread smoothly across space. The experi- and international integration, calling for a ence of successful developers shows that change in orientation of policies away from production becomes more concentrated geographic targeting toward integration. spatially. The most successful nations also By using a well-calibrated blend of insti- institute policies that make basic living tutions, infrastructure, and interventions, standards more uniform across space. Eco- today's developers can reshape their eco- nomic production concentrates, while liv- nomic geography. When they do this well, ing standards converge. they will experience unbalanced growth Part one of the Report describes the and inclusive development. geographic transformations that are neces- sary for development. Part two analyzes the The three dimensions of development drivers of these changes and identifies the The geographic transformations for eco- markets that deliver both concentration and nomic development can be characterized in Overview 7 three dimensions--density, distance, and and workers reduce their distance from division. These three words are not just met- density. The main mechanisms are the aphors for the policy challenges just outlined. mobility of labor and the reduction of They conform closely to the more technical transport costs through infrastructure notion of "market access" (see box 2). And investments. Divisions within coun- they represent the dimensions of economic tries--differences in language, currency, geography that have to be reshaped if the and culture--tend to be small, though development challenges are to be met. large countries such as India and Nigeria Understanding the transformations may be geographically divided because along the dimensions of density, distance, of religion, ethnicity, or language. and division helps to identify the main · Division is the most important dimen- market forces and the appropriate policy sion internationally. But distance and responses at each of the three geographic density are also relevant. Economic pro- scales--local, national, and international duction is concentrated in a few world (see table 1). regions--North America, Northeast · Density is the most important dimension Asia, and Western Europe--that are locally. Distances are short, and cultural also the most integrated. Other regions, and political divisions are few and shal- by contrast, are divided. While dis- low. The policy challenge is getting den- tance matters at the international level, sity right--harnessing market forces to for access to world markets, divisions encourage concentration and promote associated with the impermeability of convergence in living standards between borders and differences in currencies villages and towns and cities. But dis- and regulations are a more serious bar- tance can be important as rapid urban- rier than distance. Having a large and ization leads to congestion, and divisions dynamic economy within the neigh- within cities can be manifest in slums borhood can help smaller countries, and ghettos. especially in regions distant from world markets. For economies in other regions · Distance to density is the most impor- such as Central Africa and Central Asia, tant dimension at the national geo- international integration is hardest. graphic scale. Distance between areas where economic activity is concentrated But the potential problem at each of and areas that lag is the main dimension. these geographic scales is the same--people The policy challenge is helping firms in one place, production in another. Places Table 1 Density is most important locally, distance nationally, and division internationally Geographic scales Local National International Unit Area Country Region Examples Guangdong (178,000 km2) China (9.6 million km2) East Asia (15.9 million km2) Rio de Janeiro State (44,000 km2) Brazil (8.5 million km2) South America (17.8 million km2) Lagos State (3,600 km2) Nigeria (933,000 km2) West Africa (6.1 million km2) Greater Cairo (86,000 km2) Egypt, Arab Rep. of (995,000 km2) North Africa (6.0 million km2) Most important Density Distance Division dimension Of rural and urban settlements Between lagging and leading areas Between countries Second-most important Distance Density Distance dimension Because of congestion Of population and poverty in lagging To major world markets areas Third-most important Division Division Density dimension Between formal settlements and slums Between areas within countries Absence of large country in the neighborhood Source: WDR 2009 team. Note: Throughout the Report, "areas" are within-country economic neighborhoods or administrative units such as states or provinces, and "regions" are groupings of countries based on geographic proximity. 8 WORLD DEVELOPMENT REPORT 2009 attract production and people at different $10,000­$11,000, about the threshold for speeds, and these differences determine crossing into high incomes. This is the geographic disparities in income. Across experience of successful developers. The provinces, nations, and the world, develop- implication is that developing countries ment comes in waves and leaves behind a should expect rising subnational dispari- bumpy economic landscape--prosperity in ties in income and production when they some places, poverty in others. still have underdeveloped infrastructure and institutions. The world is not flat · Concentration is slowest internation- Development is neither smooth nor lin- ally, and it continues longer. Production ear--at any geographic scale. Growth and wealth continue to concentrate in comes earlier to some places than to others. countries beyond per capita incomes of Geographic differences in living standards $25,000, the upper reaches of the inter- diverge before converging, faster at the local national income distribution. Neighbor- scale and slower as geography exercises its hoods of nations seem to grow or stagnate influence. These are the stylized facts, based together--nearness to prosperity helps, on the experiences of successful developers while nearness to poor nations hurts. The over the last two centuries. implication is that growth strategies for later developers are not the same as the Economic production becomes strategies that worked for those who have more concentrated already grown to high-income levels; for As countries develop, people and economic today's developing countries, economic activities become more concentrated. But integration with the rest of the world-- the speed varies, depending on the spatial neighbors and distant countries--is even scale--economic forces do not operate in more essential. a geographic vacuum. The concentration of people and production is fastest locally, Local concentration (in towns and cit- slowest internationally. ies)happensquickly. Consider first the rising concentration of people in towns and cities. · Concentration is fastest locally. Economic As countries develop, the economic density concentration at the local scale is most in some places increases as more people move conveniently measured by the rate of to live in or near towns and cities (see figure urbanization--the growth of economic 1, panel a). The urban share of the population and population density in towns and cit- rises sharply--from about 10 percent to 50 ies. A large part of this geographic trans- percent--as countries grow from low income formation has been completed when to lower-middle incomes of about $3,500. countries reach per capita incomes of (It is difficult to make international com- about $3,500, roughly the threshold for parisons because countries define "urban" crossing into upper-middle incomes. The differently.3) Between 2000 and 2005, the speed of this transformation is no differ- average urban population growth for low- ent from what was seen in today's devel- income countries was 3 percent a year, more oped countries when they transformed. than twice the rate for middle-income coun- The implication is that all nations must tries and more than three times the rate for manage a rapid growth of cities when high-income countries. Sometimes, this can they still have low incomes and nascent mean rapid growth of a single city, such as institutions. Bangkok, Thailand, producing even greater · Concentration is steadier nationally. concentration. Here, it can best be measured by area The share of urban residents in total development indicators--the accumula- household consumption rises too. Urbanites tion of production and people in leading in Malawi, Jordan, and Panama--countries areas. A large part of this transforma- with per capita GDPs of about $160, $1,600, tion generally is completed when coun- and $5,600 respectively--account for 36, 63, tries reach per capita incomes of about and 80 percent of aggregate consumption. Overview 9 These spatial transformations are Figure 1 At all three geographic scales, the patterns of concentration of economic activity are closely related to the sectoral transfor- similar mation of countries from agrarian to a. As nations start to develop, people concentrate in towns and cities industrial and then, in a postindustrial Agglomeration index economy, to services. Today's high-in- 1.0 come countries experienced a similar rush to urbanize as they industrialized 0.8 (see chapter 1). All the evidence indicates that the shift from farming to industry is helped, not hurt, by healthy agriculture, 0.6 which helps towns and cities prosper.4 People move to make their own lives bet- 0.4 ter. But when agriculture is doing well, migration makes not just them better off, but also the villages they leave and the 0.2 cities in which they settle. National concentration (in leading areas) continues for longer. What is true of 0 0 5 10 15 20 25 30 35 cities is also true of areas within countries, GDP per capita (PPP, constant 2000 US$, thousands) but at a slower speed. With development, people and production become concen- b. Nationally, production concentrates in leading areas trated in some parts of countries, called Concentration Index 10 "leading" areas. Economic density grows in France, 1801­1999 9 these parts--Marmara in Turkey, for exam- 8 ple--while incomes in places economically 7 Spain, 1850­2000 Japan, 1900­2000 distant--such as southeastern Anatolia in 6 the east--can lag far behind. This concen- 5 Canada, 1890­2006 tration is hard to quantify, but it appears to 4 Netherlands, 1850­2006 3 slow or stop at per capita incomes between 2 $10,000 and $15,000 (see figure 1, panel b). 1 Initially, the concentration increases 0 2 10 20 30 rapidly. The share of total consumption of GDP per capita (constant 2000 US$, thousands) the leading areas in countries with incomes ranging between $500 and $7,500--Tajik- c. Internationally, wealth concentrates in some regions istan, Mongolia, El Salvador, and Argen- % share of global GDP (log scale) tina--increases from 30 percent to 65 percent. Comparing GDP concentrations 20 in countries with the same land area-- Lao Western "offshoots" Western Europe (United States; Canada) PDR, Ghana, Poland, and Norway--but (EU12) 10 with incomes from $600 to $27,000 shows concentration rising as incomes increase. Northeast Asia 5 (Japan; Rep. of Korea; This is nothing new. Production in Taiwan, China) today's developed economies grew more 2 concentrated until they reached high incomes. Concentration in France's leading 1 area quadrupled between 1800 and 1960, 0 5 10 15 20 25 and French incomes grew from $1,000 to GDP per capita (1990 int'l $, thousands) $6,000. But at some point, nations continue Sources: Panel a: WDR 2009 team (see chapter 1 for details); panel b: WDR 2009 team (see chapter 2 for to grow wealthier but not more concentrat- details); panel c: WDR 2009 team (see chapter 3 for details). ed--about when they enter the ranks of high-income countries. There are no rea- sons to expect greatly different patterns today (see chapter 2). 10 WORLD DEVELOPMENT REPORT 2009 International concentration (in some ging and leading areas of a country as it world regions and leading countries) con- grows through low and lower-middle tinues for a while. A similar concentration incomes, the same range of per capita of economic mass has occurred internation- incomes needed for territorial concentra- ally. Today, a quarter of the world's GDP tion to increase. And global divergence can fit into an area the size of Cameroon, in wages and wealth appears to go on and a half into one the size of Algeria. In for much longer. East Asia saw per capita 1980 the shares of the EU15, North Amer- incomes diverge between 1950 and 1970 ica, and East Asia added up to 70 percent; as Japan pulled ahead. Then, Japan's pros- in 2000 the sum was 83 percent.5 Within perity spilled over into the neighborhood, these regions, economic activity became and incomes converged as countries in more concentrated in a few countries over the region that integrated internationally time before it became more dispersed. The prospered. Among the countries of West shares of France, Germany, and the United Asia, by contrast, there was no divergence Kingdom in the EU15 regional GDP rose to in incomes--nor was there rapid growth. about two-thirds by 1940, before falling to about half today. In East Asia, the share of Convergence in living standards, like con- Japan in the region's GDP rose to 83 percent centration of economic activity, takes place in 1975 and then fell to 62 percent by 2000. faster at the local geographic scale and slow- There is no reason to expect that, when est at the international. But this happens only they prosper, other parts of the world will not in prosperous neighborhoods. Even in such experience the same patterns--a rising con- places, some measures of living standards centration in some countries, before over- (such as per capita consumption, income, flowing to their neighbors (see chapter 3). or earnings) take a long time to converge, sometimes even with an initial divergence Living standards diverge before (see figure 2). For others, such as education converging and health indicators, it can be quicker. As incomes increase, living standards con- Locally, convergence in basic living stan- verge between places where economic mass dards sets in early. Urban-to-rural gaps in has concentrated and where it has not, but consumption levels rise until countries reach not before diverging. upper-middle-income levels (see figure 2). But they fall soon after, and become small · Essential household consumption converges even before they get to high-income levels of soonest. Rural-urban gaps in essential around $10,000 per capita. Access to water household consumption diminish quite and sanitation in urban areas is more than rapidly.Even for countries that have urban 25 percent higher in urban areas for the less shares of about 50 to 60 percent, these dif- urbanized countries. For countries with ferences can be small. Area differences in urbanization rates of about 50 percent, such poverty rates are more persistent, inter- as Algeria, Colombia, and South Africa, the national differences even more so. But as disparity in access is about 15 percent. For the world has developed, these gaps have such countries as Brazil, Chile, Gabon, and diminished at all geographic scales. Jordan, the disparity is less than 10 percent. · Access to basic public services converges This pattern is also seen within countries. next. Rural-urban gaps in basic educa- Provinces that are more prosperous and tion, health, drinking water, and sani- urbanized have smaller rural-urban gaps in tation persist until countries reach living standards. This is true even in coun- upper-middle incomes. But within-city tries at low levels of income, such as China, disparities in these services--most vis- India, and the Philippines. But within highly ible as slums--persist well past high lev- urbanized areas, gaps in basic living stan- els of urbanization and upper-middle dards such as sanitation and schools tend to incomes. persist. Despite the best efforts of govern- · Wages and incomes converge last. Indeed, ments, for example, slums mark the urban wages and incomes diverge between lag- landscape in countries well after they reach Overview 11 high-income levels. It is common for one- Figure 2 At all three geographic scales, the patterns of convergence in living standards are third of a developing city's population to live similar in slums. a. Locally, first divergence, then convergence, in rural-urban gaps Nationally, divergence in living stan- Ratio of urban to rural per capita consumption dards happens quickly, but convergence is slower. At early levels of income, provin- 3 cial or interarea disparities in basic living standards can be small. But they increase quickly as countries grow. In low-income Cambodia, for example, the gap between 2 leading and lagging areas in consumption of otherwise-similar households is almost 90 percent. In middle-income Argentina, the gap is 50 percent; but in contemporary 1 0 2 4 6 8 Canada, it is just 20 percent. In the rapidly GDP per capita (2000 US$, thousands) growing East Asian and Eastern European countries, for example, these gaps have b. Nationally, divergence, then convergence, in incomes between leading and lagging areas increased rapidly. Coefficient of variation of regional wages or income A few countries such as Chile have been 0.6 exceptions. Between 1960 and 2000, it Sweden, 1920­61 0.5 Spain, experienced geographic convergence while 1860­1975 its GDP per capita more than doubled to 0.4 about $10,500. In Colombia, the ratio of United States, 1840­1960 GDP of leading Bogotá to lagging Choco 0.3 Japan, 1955­83 fell from 10 to 3 between 1950 and 1990. 0.2 Habsburg Empire, 1756­1910 Less exceptional is convergence in poverty, basic health, nutrition, and education lev- 0.1 United Kingdom, 1871­1955 els between areas within countries. Fast- growing countries everywhere have been 0 0 2 4 6 8 10 12 14 16 able to quickly translate economic progress GDP per capita (constant int'l Geary-Khamis $, thousands) into spatial equity in these more basic liv- ing standards. c. Internationally, divergence, then convergence--but only in growing regions Internationally, divergence in incomes Coefficient of variation of GDP per capita continues a while, and convergence 2.0 is slowest. Global GDP per capita has 1.9 increased almost tenfold since 1820. Life 1.8 expectancy has doubled. Literacy rates have 1.7 increased from less than 20 to more than 1.6 80 percent. But these gains have not been shared equally. Europe and its offshoots-- 1.5 Australia, Canada, New Zealand, and the 1.4 United States--and more lately Japan and 1.3 its neighbors have seen enormous increases 1.2 in income and living standards. 1950 1960 1970 1980 1990 2000 2010 For incomes, the convergence has hap- Year pened only in the fastest-growing regions Source: Panel a: WDR 2009 team estimates from more than 120 household surveys for more than 75 countries; Panel b: WDR 2009 team (see chapter 2); Panel c: WDR 2009 team (see chapter 3). of the world. The pattern has been uneven within these countries--a few countries lead, resulting in divergence within the these regions have been falling behind neighborhood, and then growth appears Europe, the European offshoots, and to spill over into their neighbors. In other Japan. The importance of neighborhoods regions such as Western Asia, there is is shown most graphically by a compari- no divergence--cold comfort because son of the southern cone nations of Latin 12 WORLD DEVELOPMENT REPORT 2009 America--Argentina, Brazil, Chile, and and persist for longer in today's develop- Uruguay--with Italy, Portugal, and Spain ing countries. Not all parts of a country in southern Europe. Between 1950 and are suited for accessing world markets, and 2006, convergence within southern Europe coastal and economically dense places do took place at 1 percent per year, but in South better. China's GDP per capita in 2007 was America at just 0.3 percent. the same as that of Britain in 1911. Shang- In contrast to incomes, global inequal- hai, China's leading area, today has a GDP ity in access to basic living standards--life per capita the same as Britain in 1988, while expectancy and education--has been fall- lagging Guizhou is closer to Britain in 1930. ing since 1930. These improvements have China's size, the openness of coastal China picked up pace since 1960 and have been to world trade, and Shanghai's location are shared across all regions. the reasons. Moreborders.While markets are becom- The world is different today, ing more international because of better but the past provides useful lessons transport and communications, the world The general patterns of concentration and has become more politically fragmented. convergence are likely to remain the same In 1900 there were about 100 international for today's developing countries as they borders (see figure 3, panel c). Today, there were for early developers. But there are are more than 600, as nations in Asia and some differences because of reasons that Africa gained independence from Euro- are technological and political. pean colonizers, and the Soviet Union and Bigger cities. Thanks to better medicine other communist countries broke up into and transport, the world is now more pop- smaller nations. The fragmentation of the ulated and cities are much larger. Between world into more nations means smaller 1985 and 2005, the urban population in domestic markets. But at the same time, the developing countries grew by more than potential for accessing foreign markets has 8.3 million a year, almost three times the been growing. In any case, thinner borders annual increase of 3 million for today's between countries now bring greater pay- high-income countries between 1880 and offs for producers and workers. 1900, when their incomes were comparable. Do such differences in technology mean If China and India are excluded, though, that the past provides no lessons? Are cit- the annual increase is less than 4.5 million, ies in developing countries too large, and about 50 percent more than a century ago. would these countries be better off if The big difference is that the world's larg- urbanization were slowed? Should today's est cities are today much larger. London developing countries be more concerned had fewer than 7 million people in 1900; about regional disparities in production the largest city among low-income coun- and income than developed countries were tries today (Mumbai) is three times that at a comparable stage of development? Is it size. So is Mexico City, the largest city in easier today for all developing countries to middle-income countries. The average size access global markets and offset the disad- of the world's largest 100 cities has grown vantages associated with greater fragmenta- to almost 10 times their size in 1900 (see tion? This Report shows why the answer to figure 3, panel a), and almost two-thirds of all these questions is no. these cities are in developing countries. Wider markets. Because of advances in Markets shape the economic communications and transport technol- landscape ogy, the notion of markets is more global. Rising densities of human settlements, Global trade as a share of production is now migrations of workers and entrepreneurs to more than 25 percent, almost five times shorten the distance to markets, and lower more than in 1900 (see figure 3, panel b). divisions caused by differences in curren- The openness to trade and capital flows cies and conventions between countries are that makes markets more global also makes central to successful economic develop- subnational disparities in income larger ment. The spatial transformations along Overview 13 Figure 3 Later developers face a different world changing the economic landscape of today's most successful developing countries, in a. Cities are more populous Millions of people, world's largest cities ways similar in scope and speed. Growing 7.0 cities, mobile people, and vigorous trade 6.3 6.0 have been the catalysts for progress in the developed world over the last two centuries. 5.0 Now these forces are powering the develop- 4.0 ing world's most dynamic places. 3.0 2.0 The realm of "agglomeration 2.0 economies" 1.0 0.7 0.2 A trip on National Highway 321 east from 0 1820 1900 1950 2000 Chengdu in Sichuan province to Shenzhen Year in Guangdong is a journey through eco- b. Markets are more international nomic development. Migrating workers Global trade as share of global GDP (%) who travel these highways often leave their 30 families behind. But they also help their 26 25 families escape poverty and propel China through the ranks of middle-income coun- 20 tries. As they travel eastward, they leave 15 an agrarian realm in which they receive 12 few benefits from working in proximity 10 6 to others. Instead, they enter the realm of 5 "agglomeration economies," in which being 1 near other people produces huge benefits. 0 1820 1900 1950 2000 Shenzhen attracts young workers--90 Year percent of its 8 million residents are of c. But the world is more fragmented working age. It specializes in electronic Number of borders goods. But it makes them in enormous 700 quantities. In 2006 its exports exceeded 600 600 India's, making its seaport the fourth busi- 500 est in the world. Propelled by the forces of agglomeration, migration, and specializa- 400 tion, and helped by its nearness to Hong 300 Kong, China, Shenzhen has grown the fast- 200 200 est of all cities in China since 1979, when it 104 100 75 was designated a special economic zone. This story is being replayed in India. In 0 1820 1900 1950 2000 1990 Sriperumbudur was known mostly as Year the place where Prime Minister Rajiv Gan- Source: WDR 2009 team. dhi was assassinated. In 2006 his widow, Sonia Gandhi, watched as Nokia's tele- these three dimensions--density, distance, phone plant churned out its 20-millionth and division--have been most noticeable handset.6 The plant had begun produc- in Japan, North America, and Western tion just earlier that year. With neither Europe. Fast and frequent movements of Shenzhen's favored administrative status people and products have helped North nor its infrastructure, Sriperumbudur America, Western Europe, and Northeast may be on its way to becoming a national, Asia account for about three-fourths of perhaps even regional, hub for electronic global production with less than a sixth of goods. The key is the town's proximity to the world's people. Chennai, just as Shenzhen's proximity to The same market forces of agglomera- Hong Kong, China, was instrumental in tion, migration, and specialization are its growth. 14 WORLD DEVELOPMENT REPORT 2009 In 1965, when independence was thrust first three chapters of the most influential on Singapore, it was not near any prosper- economics text ever written, Adam Smith's ous or peaceful place. Instead, it lay between The Wealth of Nations. Malaysia and Indonesia, two poor countries The economies of scale emphasized by that had been ravaged by war between colo- Smith can be categorized into three types-- nizers. Three-quarters of Singapore's popu- those exclusive to firms, those shared by lation lived in tenements. By 1980 it had firms in the same industry and location, industrialized, specializing in electronics, and those more generally available to pro- much as Shenzhen is doing now. By 1986 it ducers in a larger urban area. was the world's busiest container port and · With fewer than 17,000 people, Sriper- Southeast Asia's financial hub. Along the umbudur was large enough for Hyun- way, by instituting land markets, building dai to set up a big plant there in 1999. efficient transport infrastructure, and inter- By 2006 the town had helped Hyundai vening to improve housing, it cleaned up its produce its millionth automobile. Basic slums. Prosperity spilled over into neighbor- education and health services, proximity ing Malaysia. Malaysia's manufacturing-led to a port, and basic infrastructure were prosperity in turn helped more than 2 mil- all it needed to facilitate plant-level scale lion Indonesians who streamed in to fill jobs economies. The evidence is that internal in construction and services. Singapore's scale economies are high in such heavy businessmen jet around Asia, fueling growth industries as shipbuilding, and low in in places farther than Shenzhen and Sriper- such light industries as garments. The umbudur. The "little red dot" on a map--as town has enough workers to enable reportedly derided by a neighboring pres- matching workers and jobs in big plants. ident7--has transformed itself, integrated So towns like Sriperumbudur are large its neighborhood, and overtaken Britain, its enough to facilitate internal economies. former colonizer (see map 2). Singapore, Shenzhen, and Sriperum- · Shenzhen Special Economic Region-- budur show how scale economies in pro- with an area of just 300 square kilometers duction, movements of labor and capital, but a population of almost 3 million--is and falling transport costs interact to pro- home to a bustling electronics industry. duce rapid economic growth in cities and With a ready supply of skilled and semi- countries both large and small. These are skilled young workers, the area is invest- the engine of any economy, with a role so ing in better education and research fundamental in prosperity and poverty facilities to ensure that the city supplies reduction that they are the subject of the what the industry needs. Its port ships in intermediate inputs and ships out final products. It shares expensive facilities, Map 2 Settlements of varying size facilitate different scale economies such as top-notch container ports and convention centers, and matches work- Shenzhen ers to the growing number of jobs as HONG KONG, firms rapidly expand their operations. CHINA Proximity to Hong Kong, China, pro- vides access to finance, though Shenzhen is home to a rapidly expanding financial Bangalore sector. And competition for customers Chennai Sriperumbudur among the multiple suppliers of inputs produces cost savings. The area excels in providing, in economic jargon, localiza- tion economies. · Singapore has passed through these Singapore stages and is now one of the world's top centers of commerce. By providing Source: WDR 2009 team. a stable economic environment, excel- Overview 15 lent transport links, livability, and effi- of all inputs to production--land--must cient finance, it provides services to the become mobile between uses. Access to entire Asia-Pacific region. These ser- oceans and rivers might be the reason a vices are used by a wide range of indus- place is settled, but the nimbleness of its tries, from shipping to manufacturing, land markets will largely determine how to education, and to finance, insurance, much it will grow. Governments may not be and real estate. They thrive on eco- good at picking places that will prosper. But nomic density. With fewer than 5 mil- how well they institute regulations, build lion people packed into less than 700 infrastructure, and intervene to make land square kilometers of space, Singapore use efficient will decide the pace of prosper- is the world's most densely populated ity for the entire neighborhood. country. In 2006 its exports of $300 Depending on what type of agglomera- billion approached those of the Russian tion economies they deliver, places can be Federation, which has more than 16 large or small. Function is far more impor- million square kilometers. Singapore's tant than size. But locating farther away diversity facilitates sharing, matching, from economic density generally reduces and learning, providing what econo- productivity. Doubling this distance in mists call urbanization economies. Brazil apparently reduces productivity by 15 percent and profits by 6 percent. Better In most countries, such towns and cit- infrastructure reduces economic distance. ies coexist. Brazil's Rio de Janeiro state has But in a developing country, the most nat- about 14.5 million people. Volta Redonda, ural way for workers and entrepreneurs to not too far from Rio city, originally supplied close this distance is to move closer. goods and services just to meet the needs of CSN, the largest steel plant in Latin Amer- Migrating to profit from proximity ica. Duque de Caixas, about 15 kilometers Agglomeration economies attract people from Rio, meets the needs of an industry and finance. Today, capital tends to move producing petrochemicals. And the diversi- quickly over long distances to exploit fied Rio de Janeiro metropolis, with about opportunities for profit. People also move, 6 million people, supplies financial services but they move more quickly to nearby to settlements that surround it. And with agglomerations than to those far away. other metropolises like São Paulo, Rio con- Once plants and people come to a place, nects Brazil to the rest of Latin America and others follow. the world. The pattern is so familiar that it is almost a law of urban economics. · Locally, the move toward density is quick The functions and fortunes of settle- in fast-growing economies, manifest ments are linked. Industrialized places are in a rapid rural-urban migration that different from their agrarian predecessors accompanies the shift from agriculture to not just because they are more concentrated industry. As the Republic of Korea grew but also because they are more specialized. between 1970 and 1995, the urban share The largest cities may be well suited for of population quadrupled to 82 percent, startup enterprises; the smaller ones may be with migration accounting for more than better suited for those more established. In half the increase in the 1960s and 1970s. agriculture, sowing and reaping must hap- · Nationally, workers move to reduce dis- pen in the same place. Not so for industry tance to markets in parts that are pros- and business services. Falling transport and pering. About 3 million people moved communications costs allow firms to spa- in the second half of the 1990s from tially separate sowing and reaping. Prod- the lagging Indian states of Bihar and ucts may be designed and financed in large Uttar Pradesh to leading Maharashtra cities--and produced in small towns. and prosperous Punjab (see map 5). In As firms adjust to changing market con- Vietnam, a much smaller country, more ditions, places have to perform different than 4 million people migrated inter- functions or risk decay. The most immobile nally during the same period. 16 WORLD DEVELOPMENT REPORT 2009 · Internationally, regional migration is among the top 10 sending countries. But a big part of labor mobility. Migration interregional migration is sluggish. Fewer among neighbors is considerable. Côte than 200 million of the world's 6.7 billion d'Ivoire, India, and the Islamic Republic people live outside their region of birth. of Iran have been among the top desti- And just 2 million people move from nations for their neighbors. Germany, poorer countries to the developed every Italy, and the United Kingdom still rank year, half of them to the United States. Map 3 Migrating to reduce distance to density: Despite the obstacles, Chinese workers have migrated in the millions Migration totals Migration from western region to 2,000,000 coastal region 1,000,000 Migration from central region to 500,000 coastal region HEILONGJIANG 250,000 C E N T R A L 100,000 Migration within the coastal region R E G I O N JILIN W E S T E R N R E G I O N LIAONING N E I M O N G O L BEIJING Beijing TIANJIN HEBEI XIA COASTAL G SHANXI REGION IN SHANDONG QINGHAI N GANSU HENAN SHAANXI JIANGSU W E S T E R N ANHUI R E G I O N SHANGHAI HUBEI NG SICHUAN ZHEJIANG CHONGQI COASTAL HUNAN JIANGXI REGION GUIZHOU FUJIAN YUNNAN TAIWAN GUANGXI GUANGDONG HONG KONG MACAO CHINA HAINAN Source: Huang and Luo 2008, using data from the population census of China. Overview 17 This sum is not likely to increase, even Map 4 Migrating to reduce distance to density: Brazil's young workers move in thousands to though the gains from greater migration get closer to economic density from developing to developed countries are considerable.8 International migration has been high in the past: fully 20 percent of BRAZIL Europeans emigrated to new lands in the Americas, Australia, and South Africa. Today, these movements have slowed. Just 500,000 Chinese emigrated abroad in 2005. N O R T H But internal migration has picked up in the developing world. More than 150 million 762 (1991) people moved internally in China despite 788 (2000) restrictions (see map 3). In Brazil's high- NORTHEAST growth years during the 1960s and 1970s, almost 40 million people left the coun- CENTER-WEST 45,672 (1991) 57,131 (2000) tryside for cities; even today, young work- ers migrate in large numbers (see map 4). 2,839 (1991) Brasília Vigorous internal migration is not new. 4,117 (2000) Between 1820 and 2000 per capita incomes in the United States multiplied 25-fold, SOUTHEAST and Americans earned the reputation of Population that did not migrate being among the most footloose of people. 1991 2000 In Japan internal migration peaked in the 300,000 14,697 (1991) 1960s, as it grew to become the world's 17,228 (2000) 200,000 1991 2000 second-largest economy. SOUTH 100,000 15,000 Despite aggressive area development 10,000 30,000 policies, 1.7 million people--more women 5,000 2,000 than men--have left East Germany for the West, helping to make incomes more Source: WDR 2009 team, based on census data from the Instituto Brasileiro de Geografia e Estatística. equal. Since the transition to market economies, firms and people have picked places better suited for production. More than a million people--about 12 percent of Map 5 Migrating to reduce distance to density: Migration in India has been less frenetic HIMACHALHIMACHAL MAHARASHTRA HIMACHALHIMACHAL DELHI HIMACHALHIMACHAL PUNJAB PRADESH PRADESH PRADESH PRADESH PRADESH PRADESH PUNJABPUNJAB UTTARANCHALUTTARANCHAL PUNJABPUNJAB UTTARANCHAL UTTARANCHAL UTTARANCHALUTTARANCHAL 1,712,627PUNJAB PUNJAB UTTARANCHAL HARYANADELHIDELHI HAHARYANRYANA HAHARYANRYANA HARYANA SIKKIMSIKKIM SIKKIMSIKKIM HARYANA HAHARYANRYANA SIKKIMSIKKIM 131,895 DELHIDELHI DELHIDELHI RAJASTHANRAJASTHAN UTUTTARTAR UTTAR RAJASTHAN RAJASTHAN UTTAR UTUTTARTAR RAJASTHAN RAJASTHAN UTUTTARTAR UTTAR PRADESHPRADESH PRADESH PRADESH PRADESHPRADESH BIHARBIHAR BIHARBIHAR BIHARBIHAR JHARKHANDJHARKHAND JHARKHAND JHARKHAND JHARKHANDJHARKHAND GUJARGUJARAT GUJARAT WESTWEST MADHMADHYA PRADES MADHYA PRADESH YA PRADESH MADHYA PRADESH MADHMADHYA PRADES YA PRADESH WESTWEST MADHMADHYA PRADES MADHYA PRADESH YA PRADESH WESTWEST BENGALBENGAL BENGAL BENGALBENGAL BENGAL BENGALBENGAL BENGAL GUJARAT GUJARGUJARAT GUJARGUJARAT GUJARAT CHHATTISGARH TTISGARH CHHATTTISGARH TISGARH ORISSAORISSA ORISSAORISSA CHHATTISGARH TTISGARH ORISSAORISSA CHHA CHHA CHHA MAHARASHTRA MAHARASHTRA MAHARASHTRA MAHARASHTRA MAHARASHTRA 12,505,916 ANDHRAANDHRA ANDHRA ANDHRA ANDHRAANDHRA PRADESHPRADESH PRADESH PRADESH PRADESHPRADESH GOA KARNKARNATAKATAKA KARNATAKA GOA KARNATAKA KARNKARNATAKATAKA GOA KARNKARNATAKATAKA KARNATAKA INDIA 12,505,916 Number of within-state migrants Migration totals 1,000,000 500,000 100,000 50,000 Source: WDR 2009 team, based on census data from the Census of India. 18 WORLD DEVELOPMENT REPORT 2009 residents--have left Siberia and the Russian should facilitate labor mobility. For North and Far East for the western parts of decades since independence, India treated Russia. its 40 million emigrants as "not required West Africa has sustained regional labor Indians." Encouraged by a change in atti- mobility through institutional coopera- tude since the 1990s, expatriate Indians are tion. But independent Africa is generally pulling distant places like Bangalore and less integrated. Africans--especially the Hyderabad closer to world markets, just as most skilled--have been leaving the conti- the overseas Chinese did for Shanghai and nent, seeking and getting higher rewards in Guangzhou more than a decade earlier. the North. Other parts of the world show Falling costs of transport and communica- how to deal with this brain drain. Educated tions have helped greatly. workers will be pulled toward places where other skilled people agglomerate. This is Specializing and trading as transport beneficial for both places. But when people costs fall are pushed out by the lack of security or Transport and communication costs have basic services, migration is beneficial for indeed fallen rapidly over the last century, the migrant but not always for the nation. especially in the last 50 years. Since the Pull migration is better than push, but both 1970s, railroad freight costs are down by are hard to stop or slow. Policy makers are half. Road transport costs, despite higher realizing that the challenge is not how to energy and wage costs, are down by about keep people from moving, but how to keep 40 percent. For worldwide air freight, the them from moving for the wrong reasons. price has fallen to about 6 percent of its 1955 China illustrates the benefits. Except for level. The price for tramp shipping services a brief period during the Cultural Revolu- is half that in 1960. A three-minute phone tion, China has treated its diaspora well, call from New York to London was almost according them both rights and respect.9 $300 in 1931. Today, the same call can be Internally, its policies have gone back and made for just a few cents. forth, but now they are shifting from trying With falling domestic transport costs, to discourage people from moving to deliv- economic production should have become ering basic services to people wherever they more evenly spread within countries. With live. The policies are paying dividends. As lower costs of transporting and commu- Chinese migrants are moving to the coast nicating internationally, countries should by the million, many of the 57 million have traded more with distant partners. overseas Chinese are bringing finance and What happened was the opposite. Falling expertise back to some of the same places. transport costs have coincided with greater Internal and international migrants are economic concentration within countries. coming together in a way that is not acci- And while countries now trade more with dental. The willingness of the Chinese to everyone--exports as a share of world pro- move--leaving the country for other parts duction quadrupled to 25 percent over the of the world to escape war and squalor in the last three decades--trade with neighbors first part of the twentieth century and then became even more important. bringing finance and know-how to coastal Why did this happen? The answer lies in China during the last quarter--promises the growing importance of scale economies to bring to southeast China a "reversal of in production and transport (see chapter fortune" rivaling the U.S. Northeast (see 6). As transport costs have fallen, they have "Geography in Motion: Overcoming Dis- allowed greater specialization and radically tance in North America"). altered the location of firms and the nature Countries do not prosper without of trade. With high transport costs, firms mobile people. Indeed, the ability of people had to be near consumers. But as transport to move seems to be a good gauge of their costs fall, they can avail of internal, local, economic potential, and the willingness to and urban economies of scale, and trans- migrate appears to be a measure of their port the product to consumers. Internation- desire for advancement. Governments ally, the same thing. With high transport Overview 19 costs, England imported only what it could Figure 4 Intraindustry trade is high in North America, Western Europe, Oceania, and East Asia not grow or produce at reasonable cost-- spices from India and beef from Argentina Southern Africa in exchange for British textiles and china. Western Africa 1962 As transport costs fell, it imported more 1975 Central Africa 1990 spices and beef. But it also traded more with 2006 Northern Africa France and Germany--Scotch whisky for French wine, English ale for German beer. Central Asia, Caucasus & Turkey Trade to fulfill basic needs was joined and Eastern Africa soon overtaken by trade to satisfy a variety Western Asia of wants. South Asia Falling costs of transportation and com- munication have made the world smaller. Central America & Caribbean But they have also made economic activity South America more geographically concentrated. Eastern Europe & Russian Federation · Locally, with falling costs of commut- Northeast Asia ing and a greater potential for exploit- Southeast Asia & Pacific ing scale economies, towns and cities can Western Europe grow bigger and denser. Australia & New Zealand · Nationally, as leading and lagging areas within countries are connected through North America better modes of transport, production 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 is more concentrated in the more eco- Grubel-Lloyd Index, 3-digit, intraregional trade nomically dense areas to take advantage Source: Brülhart 2008 for this Report. of agglomeration economies. Note: The Grubel-Lloyd Index is the fraction of total trade that is accounted for by intraindustry trade. · Internationally, countries that have lowered the costs of transport more world, explaining why the friction of bor- have benefited most from greater trade. ders on trade has fallen. Aided by a deep- Greater specialization has made these ening integration, the intraregional share countries more competitive still, con- of trade in the EU has risen above 60 per- centrating trade and wealth in a few cent (see "Geography in Motion: Overcom- parts of the world. ing Division in Western Europe"). In East Asia, the fastest-growing region, the share of Scale economies are evident in the trans- regional trade is now more than 55 percent port sector, too. More trade means lower (see "Geography in Motion: Distance and costs of transportation, which in turn Division in East Asia"). means more trade. This is especially true Development in a world of greater spe- for intraindustry trade, which has been the cialization and concentration is even more most rapidly growing part of international challenging. Developing countries have trade during the last half-century. Since higher transport costs and small markets, 1960 the share of intraindustry trade in the which do not support specialization. But world's total has doubled from 27 percent several countries--mainly in East Asia-- to 54 percent. Within-region intraindustry have shown that these markets are acces- trade is low in most regions, and high in a sible for low-income countries. The answer few. It is close to zero for Central Africa, lies in the fastest-growing component of Central Asia, East Africa, Northern Africa, intraindustry trade: trade in "intermediate South Asia, and Southern Africa. It is high- inputs" of production (see box 3). est for Australia, East Asia, New Zealand, In agriculture, industry, and services, North America, and Western Europe (see the potential for fragmenting production figure 4). is almost without limit. Thailand may not Regional cooperation has advanced be able to make a television set better than much faster and further in these parts of the Japan, but it could make parts of televisions 20 WORLD DEVELOPMENT REPORT 2009 BOX 3 Intraindustry trade and intermediate inputs More than half of world trade today is nal and intermediate goods, with both est increase is in food and live animals. intraindustry trade, with industries clas- having increased considerably over the Consumers like variety for farm produce, si ed in 177 (3-digit) categories, up from last 50 years. This rise in intraindustry and that means pro t in trade between about a quarter in 1962. So countries trade is not just for manufacturing. Intra- two countries that raise similar food and are becoming more similar in their eco- industry trade in machines and transport animals (see gure at left). nomic structures. This trade consists of equipment is the highest, but the larg- But the largest rise is for intermediate inputs--the produced means of produc- Intraindustry trade has risen for primary, intermediate, and final goods tion. Marginal intraindustry trade--a reliable measure of change--is highest Grubel-Lloyd Index, 3-digit in intermediate inputs. This is not just for 0.6 manufacturing. Agriculture needs inputs, Intermediate goods too. And falling communications costs 0.5 have resulted in greater fragmentation of services into "components," supplied to nal consumers from di erent parts of 0.4 Final goods the world. Trade in intermediate goods is more sensitive to transport costs than is trade 0.3 Primary goods in nal goods. Consider the following illustration: if intermediate inputs are 0.2 two-thirds of the value added for a good, a 5-percent increase in transport costs can mean the equivalent of a 50-percent 0.1 tax. Little wonder that intermediate goods trade has increased fastest in parts 0 of the world that have reduced trade and 1962 1966 1970 1974 1978 1982 1986 1990 1994 1998 2002 2006 transport costs the most. Year Source: Brülhart 2008, for this Report. Source: WDR 2009 team. equally well and much cheaper. Anchored transport costs have occupied the interest by China and Japan, countries in East Asia of researchers (see box 4). have developed production networks that Their insights should change what to trade intermediate goods back and forth. expect from the markets. They should also By specializing in a small part of the pro- inform what governments can do to pro- duction chain, they have broken into this mote the geographic transformations nec- most lucrative and fastest-growing compo- essary for development. nent of trade in manufactures. Countries in other regions can also Putting development in place benefit from the growing trade in inter- Prosperity will not come to every place at mediate goods. The key for most is mak- once, but no place should remain mired in ing a concerted effort to lower the costs of poverty. With good policies, the concen- transport. This means more concentra- tration of economic activity and the con- tion within developing countries, but--by vergence of living standards can happen allowing them to specialize at earlier stages together. The challenge for governments is of development and exploit economies of to allow--even encourage--"unbalanced" scale--it will help them converge to the economic growth, and yet ensure inclusive incomes and living standards in the devel- development. They can do this through oped world. Over the last two decades, economic integration--by bringing lag- such interactions between scale econo- ging and leading places closer in economic mies, mobility of capital and labor, and terms. Overview 21 BOX 4 New insights from a generation of analysis Researchers have been taking a fresh thick input markets, but not for "urban- Falling transport costs increase trade look at industrial organization, eco- ization" economies--especially those more with neighboring, not distant, nomic growth, international trade, and involving knowledge spillovers--gener- countries. With a decline in transport costs, economic geography, having incorpo- ated mainly by large cities (see chapter 4). countries should trade more with countries rated the e ects of scale economies in The implication: policy makers should that are farther away. But trade has become production. The results can be surpris- focus on the functions of cities. more localized than globalized. Countries ing for those schooled in conventional Human capital moves to where it is trade more with countries that are similar, economic analysis. Here are some of the abundant, not scant. Conventional eco- because increasingly the basis of trade is new insights: nomic analysis implies that people should the exploitation of economies of scale, not Plants have to be big to exploit econ- move to where their skills are scant. But the di erences in natural endowments. omies of scale, but places do not have the opposite seems to happen: educated The reason: falling transport costs make to be big to generate them. Increasing migrants seek places where many others specialization possible (see chapter 6). returns to scale arise because of xed have similar skills. Among the 100 largest The implication: falling transport costs costs of production (internal to a rm) metropolitan areas in the United States, change the composition of international and proximity to workers, customers, the 25 cities with the highest share of trade and make it even more sensitive to and people with new ideas (external to a college graduates in 1990 had, by 2000, such costs. Policies to reduce trade and rm, even an industry). The size of settle- attracted graduates at twice the rate of the transport costs should be a big part of ments matters less than their function. other 75. growth strategies for late developers. The reason: with reasonable transport The reason: educated workers gain Recognizing scale economies and their costs, towns can be large enough to from proximity to others (see chapter 5). interaction with the mobility of people and facilitate internal scale economies. Medi- The implication: policies should not products implies changing long-held views um-size cities are often large enough for ght the market force that pulls skilled about what is needed for economic growth. "localization" economies that come from people together. Source: WDR 2009 team. This integration can best be done by · Turkey is trying to change neighbor- unleashing the market forces of agglom- hoods too, in a different way. The eration, migration, and specialization, not country of 70 million has been looking by fighting or opposing them. How well toward integration with the EU. Because markets and governments work together of higher agglomeration economies and determines the speed and sustainability of lower transport costs, areas near Istanbul geographic transformations. Look at what and Izmir may be better suited for inte- is happening in Bogotá, Turkey, and West grating with Europe. The more distant Africa: areas of eastern and southeastern Anato- lia and the Black Sea have 40 percent of · Bogotá has almost 7 million citizens, but the land but less than 20 percent of the migration from rural Colombia contin- national product, with a GDP per capita ues. A third of its population growth is about half that of the western areas. The due to rural migrants, who mostly settle disparities persist despite government in poor, crowded neighborhoods as the efforts to spread economic mass toward city grows denser.Since 2000 a new public the east. Meanwhile, public investments transportation system, the TransMilenio, in social services help lagging areas, has eased congestion, now carrying a while fiscal incentives for firms to locate million passengers a month. For the poor in those areas seem ineffective.10 neighborhoods especially, it has reduced · The Economic Community of West Afri- the distance to economic opportunities. can States (ECOWAS) has a protocol that But many people still live in slums, and allows free movement of its 250 million crime and violence are getting worse. people between member states. This A municipal initiative has addressed has helped the neighborhood maintain these social divisions since 2003, helping regional labor mobility at preindepen- almost a million people integrate into the dence levels, even as it fell in East and city and change their neighborhoods. Southern Africa. But trade is another 22 WORLD DEVELOPMENT REPORT 2009 story. In the most dynamic parts of the trade similar goods and services, moti- world, the exchange of similar goods and vated more by the benefits of special- services--intraindustry trade--has been ization and scale than by differences in rising rapidly. But in West Africa, inter- natural endowments. Trade can only national borders are thickened by red tape partially offset the immobility of land and illicit checkpoints, which divide the and labor, but it will help convergence region and thwart the efforts of ECOWAS when developing countries can tap into members to specialize and trade. the most rapidly growing component: trade in intermediate goods. As the lens of economic geography is widened, different movements, stresses, Private motives are the main shapers and strains come into view. of the economic landscape, but it can be reshaped by collective action, most potently · Locally, in places like Bogotá, land must by governments. Seen through the lens of accommodate more and more people. economic geography, land use, labor mobil- If land markets work well, land will be ity, and intermediate goods trade come into mobile between uses and allocated pro- focus (see table 2). Governments should pay ductively. The cities that do this best special attention to land, labor, and product will grow, and even more people will be markets. When they do not work well, the attracted to their economic density. forces of agglomeration, migration, and · People and products move much faster specialization weaken, and the economy in and around Bogotá than they do in stagnates. When they do, land, labor, and Turkey. But even in Turkey, the western input markets bring the economic efficiency areas will become more prosperous and that comes with geographic concentration, dense, if at a slower pace. Spatial dis- and the equity associated with converging parities in incomes and poverty rates living standards. between the west and the east will likely rise and then diminish as people move A rule of thumb for economic to take advantage of economic density. If integration labor markets in Turkey are fluid, peo- The concern of policy makers is that pro- ple will reduce their economic distance duction will concentrate in some places, to these agglomerations. people in others. Cities will have economic · Internationally, these movements are density, and the countryside most of the likely to be fewer and even slower. If poor. Leading areas will have the economic regional and global markets were inte- mass, while the poor are massed in lagging grated, countries in West Africa would areas. Some countries will have much of the specialize in a few tasks and become world's wealth, others most of the world's competitive in world markets. As divi- poor. Even if this were temporary, it seems sions diminish, neighboring countries unfair. But the disparities may be long last- ing, destabilizing parts of a country, entire Table 2 Agglomeration, migration, and specialization are the most important forces-- nations, and even some world regions. and land, labor, and intermediate inputs the most sensitive factor markets Governments have many reasons to Geographic scales worry about disparities in welfare in and Local National International among countries. They also have many pol- Economic Agglomeration Migration Specialization icy instruments for promoting economic force Speeded by migration, Influenced by Aided by agglomeration integration to reduce those disparities. capital mobility, and agglomeration and and factor mobility trade specialization · Institutions--shorthand in the Report Key factor of Land Labor Intermediate inputs for policies that are spatially blind in production Immobile Mobile within countries Mobile within and their design and should be universal in between countries their coverage. Some of the main exam- Source: WDR 2009 team. ples are regulations affecting land, labor, Note: Throughout the Report, "areas" are within-country economic neighborhoods or administrative units such as states or provinces, and regions are groups of countries based on geographic proximity. and international trade and such social Overview 23 services as education, health, and water the least developed countries can end up and sanitation financed through tax and dominating policy discussions. transfer mechanisms. This Report calls for a rebalancing of · Infrastructure--shorthand for policies these debates to include all the elements and investments that are spatially connec- of a successful approach to spatial inte- tive. Examples include roads, railways, gration--institutions, infrastructure, and airports, harbors, and communication incentives. Using the findings in part one systems that facilitate the movement of and the analysis of market forces in part goods, services, people, and ideas locally, two, part three reframes these debates, nationally, and internationally. calling for a shift from spatial targeting to · Interventions--shorthand for the spa- spatial integration. tially targeted programs that often dom- The world is complicated, and the prob- inate the policy discussion. Examples lems of economic integration defy simple include slum clearance programs, fis- solutions. But the principles need not be cal incentives for manufacturing firms complex. The bedrock of integration poli- offered by state governments, and pref- cies should be spatially blind institutions. erential trade access for poor countries Where the integration challenge spans more in developed country markets. than one geographic dimension, institutions must be augmented by public investments Today, policy debates often begin and in spatially connective infrastructure. Spa- end with discussions of spatially targeted tially targeted interventions are not always incentives. The debate on how to pro- necessary. But where the problem is low mote healthy urbanization is polarized economic density, long distances, and high between those who emphasize villages, divisions, the response must be comprehen- where a majority of the world's poor still sive, involving spatially blind, connective, live, and those who believe the way out of and targeted policies. poverty lies in cities, where much of the For each spatial dimension, an instru- world's wealth is generated. As urban pov- ment of integration (see table 3). The rule erty increases, the focus is shifting from of thumb: "an I for a D." villages to slums. Motivated by within- country geographic disparities in living · For a one-dimensional problem, the standards, the debate on territorial devel- mainstay of the policy response should opment is similarly fixated on economic be (spatially blind) institutions. growth in lagging areas. At the interna- · For a two-dimensional challenge, both tional level, preferential market access for institutions and (spatially connective) infrastructure are needed. Table 3 "An I for a D?" A rule of thumb for calibrating the policy response Policy priorities for economic integration Institutions Infrastructure Interventions Place type--local (L), national (N), and international (I) Spatially Spatially Spatially Complexity of challenge geographic scales blind connective targeted One-dimensional problem L. Areas of incipient urbanization N. Nations with sparse lagging areas · I. Regions close to world markets Two-dimensional L. Areas of intermediate urbanization challenge N. Nations with dense lagging areas · · I. Regions distant from world markets Three-dimensional L. Areas of advanced urbanization that have within-city divisions predicament N. Nations with dense lagging areas and domestic divisions · · · I. Regions distant from markets with small economies Source: WDR 2009 team. Note: Throughout the Report, areas are within-country economic neighborhoods or administrative units such as states or provinces, and regions are groupings of countries based on geographic proximity. 24 WORLD DEVELOPMENT REPORT 2009 · For a three-dimensional predicament, behind a misplaced density of populations all three instruments are needed--in- in lagging areas, so that in some countries stitutions, infrastructure, and (spatially (such as Brazil) lagging areas have higher targeted) interventions. poverty rates and high population densi- ties. Internationally, developing regions The primary dimension at the local are all deeply divided, but some also may geographic scale is density; nationally, it be distant from world markets. Even if is distance; internationally, division. At regional institutions take hold and make each of these geographic scales, policies South Asia a more integrated region, some designed without explicit consideration to countries (such as Nepal) may need con- space should be seen as the primary instru- certed policy action to improve the infra- ment. In some places, these can be a large structure to reach growing regional and part of integration policies. The task of international markets. For places that integration is relatively straightforward in face two-dimensioned integration chal- areas of incipient urbanization (as in lag- lenges, investments in infrastructure that ging states in many low-income countries), connects lagging to leading places and in countries with mobile labor and capital aid market access should supplement the (such as Chile), or in regions that are close institutions that bring people together. to world markets (such as North Africa). The integration challenge is greatest where In such places, the integration challenge adverse density, distance, and division com- can be seen as one dimensional. Explicitly bine to pose a "three-dimensional challenge." spatial policies are not generally necessary. In highly urbanized areas (such as Bogotá), Universal or spatially blind institutions-- the fear is that economic density and popu- made available to everyone regardless of lation density may not coincide. Within-city location--form both the bedrock and the divisionsmaypreventtheintegrationofslums mainstay of an effective integration policy. and spawn problems of crime and grime. In As the task becomes more complicated, some countries (such as India), ethnic, reli- these institutions must be assisted by gious, or linguistic divisions discourage the infrastructure. Locally, rapid urbanization poor in densely populated lagging areas from can congest the area, increasing economic seeking their fortunes elsewhere. And in the distance and choking off agglomeration most fragmented and remote regions (such as economies. In places such as Mumbai, Central Africa or Central Asia), a clustering whose population has doubled since the of small and poor nations can lead to spill- 1970s, rising congestion has to be met by overs of the wrong kind--disease, conflict, investments in transport infrastructure, or corruption. so that the benefits of density are shared Slums in large cities, densely populated more widely. Nationally, changing eco- poor areas in divided nations, and the "bot- nomic and political fortunes can leave tom billion" countries--approximating the three billions discussed at the beginning-- Figure 5 In charted waters: the pace of urbanization today has precedents are the most difficult challenges for inte- Change in urban shares since 1800 gration. The policy responses should not be timid. But they should also be deliberate. Developing economies (median), 1985­2005 High-income economies (mean), 1880­1900 Efficient and inclusive urbanization Developing economies (mean), 1985­2005 No country has grown to middle income United States, 1800­1900 without industrializing and urbanizing. Denmark, 1800­1900 None has grown to high income without United Kingdom, 1830­50 vibrant cities. The rush to cities in develop- Germany, 1830­50 ing countries seems chaotic, but it is nec- Canada, 1880­1900 essary. It seems unprecedented, but it has 0 5 10 15 20 25 happened before (see figure 5). It had to Percentage point difference in urban shares have, because the move to density that is Source: WDR 2009 team calculations based on data from various sources (see figure 1.13). manifest in urbanization is closely related Overview 25 to the transformation of an economy from establish the institutional foundation agrarian to industrial to postindustrial. of possible urbanization in some places. Governments can facilitate the spatial Good land policies are central, and so transformations that lie behind these sec- are policies to provide basic services toral changes. Depending on the stage of to everyone. For example, the univer- urbanization, sequencing and priority- salization of land rights in Denmark setting require paying attention to different at the turn of the eighteenth century aspects of the geographic transformation. contributed greatly to the nation's take- What does not change is that a foundation off into industrialization a few decades of institutions must be universal and come later. Indeed, policies to strengthen rural first, investments in connective infrastruc- property rights are seen as instrumental ture should be both timed and located well for higher agricultural productivity in and come second, and spatially targeted sixteenth-century England, which freed interventions should be used least and last. workers to migrate to towns to work The approach requires the discipline of in manufacturing and services. A close following the integration principle set out complement to the institutions for bet- earlier. The payoff is a spatial transforma- ter land markets is the universal provi- tion that is both efficient and inclusive (see sion of basic social services--security, chapter 7). education, health services, and sanita- The principles outlined in the Report tion. In 1960, the Republic of Korea had help to prioritize policies for different stages a per capita income level that Benin has of urbanization, providing the elements today. Seventy-five percent of its people of an urbanization strategy. Map 6 shows lived in rural areas, more than a third three areas in Colombia, each with a spe- of Korean adults had no schooling, and cific geography. But the principles are quite fewer than 5 percent of children had universal. been immunized against preventable diseases such as measles. By 2000, more · Incipient urbanization. In places that than 80 percent had urbanized, almost are mostly rural, governments should everyone was literate and immunized, be as neutral as possible and should and the Republic of Korea's income had Map 6 As urbanization advances, policies must evolve a. Incipient urbanization b. Intermediate urbanization c. Advanced urbanization in Cauca, Colombia in Santander, Colombia in Colombia's Capital Area VALLE DEL CAUCA BOLÍVAR CUNDINA MARCA CAUCA Bucaramanga ANTIOQUIA Bogotá Popayán SANTANDER HULA TOLIMA META COLOMBIA Source: WDR 2009 team, using data from Schneider, Friedl, and Potere 2008. 26 WORLD DEVELOPMENT REPORT 2009 of rising economic density are more widely shared. Industrialization involves BOX 5 Concentration without congestion in western changing land use patterns as activities China: Chongqing and Chengdu concentrate, and requires moving goods An experiment in China might change Better institutions. The emphasis and services around quickly. Land use the future of urbanization policy in is on coordination across government regulations can affect location decisions, the developing world. Policy makers levels to manage land use and con- and they continue to be the institutional should take notice. version. In the countryside, the plan priority. Spatially blind social services China is taking inland the urbaniza- concentrates rural land by transfer- should continue as part of rural-urban tion strategy that was successful in ring use rights to rms and farmers. integration, so that people are pulled to the leading coastal areas in the 1980s In towns and cities, the creation of and 1990s. The "area approach" is industrial zones is a key part of the cities by agglomeration economies, not being implemented in two places-- wider framework. Large and medium pushed out by the lack of schools, health Chongqing and Chengdu, both cities are developing high value- services, and public security in rural located in the near west. At about 40 added manufacturing, while smaller areas. But even if these services are pro- percent, they have the same urbaniza- cities and towns are specializing in vided, transport costs can rise quickly tion ratio as the average for China. The labor-intensive industries, pulling in because of growing congestion, affecting aim is to increase that to 70 percent by labor from nearby villages, and facili- the location choices of entrepreneurs. 2020, promoting both concentration tating localization economies. and rural-urban convergence. More infrastructure. Massive trunk Connective infrastructure is needed to Chongqing has a population of infrastructure is planned. Chongqing keep such areas integrated. State and cen- about 40 million, with a portfolio of will spend billions on infrastructure, tral governments that work well together a capital city, six large cities, 25 small from the central government and can provide the trunk infrastructure and mid-size cities, 95 central town- through increased private invest- necessary to ensure that prosperity is ships, and 400 townships. Chongqing ment from Hong Kong, China, and widely shared. Making the administra- has been accorded the status of from Singapore. In Chengdu, about tive jurisdiction wider can help in coor- a special municipality, as Beijing, 117 billion yuan will be invested in Shanghai, and Tianjin have had for 71 infrastructure projects, including dinating infrastructure investments. A some years. Like them, it will enjoy rural-urban transport networks, and good example is Chongqing in western greater nancial autonomy. Chengdu water and sanitation projects in both China (see box 5). is smaller, a sprawling metropolitan rural and urban areas. Another 16.5 · Advancedurbanization.Inhighlyurban- area with 11 million people. Along billion yuan will be invested in 34 ized areas, besides institutions and infra- with the 2,000-year-old capital city of social projects to improve the living structure, targeted interventions may be Sichuan province, it has eight medi- standards of lagging rural residents. um-size cities, 30 central townships, If markets favor the two places as necessary to deal with the problem of 60 townships, and 600 villages. much as the government has, they slums. Services and learning require peo- The urbanization strategy involves will improve the lives of millions in ple to be in proximity to livable surround- "three concentrations" of land, the Chinese hinterland. The integra- ings. This is the stage in which slums can industry, and farmers. The idea is to tion already has had a local impact. compromise a city's ability to deliver the reap the bene ts of scale economies, In Chongqing, rural incomes in 2007 economies that come from proximity. promote the mobility of goods and increased faster than those of urban Slum-improvement programs may not workers, and improve the well-being residents. In Chengdu, farmer con- of new migrants to cities. Consistent centrations are believed to have led be a priority at earlier stages of urbaniza- with the policy priorities outlined for to a productivity increase of 80 per- tion, but at this stage they become nec- areas with intermediate urban shares cent, as industry has been absorbing essary. The lesson from assessments of of about 40­50 percent, the emphasis about 100,000 farmers a year. slum-improvement initiatives is that tar- in both places is on universal institu- geted interventions will not be enough tions and connective infrastructure, by themselves. These interventions will not spatially targeted interventions. Source: WDR 2009 team. not work unless institutions related to land and basic services are reasonably effective, and transport infrastructure is reached that of modern-day Portugal. in place. A three-dimensional challenge Another good example is Costa Rica. must be met by a three-pronged policy · Intermediate urbanization. In places response, requiring coordinated policies where urbanization has picked up at the central, state, and city levels of gov- speed, in addition to these institutions, ernment. Singapore's success shows the governments must put in place connec- advantages of such coordination in a city- tive infrastructure so that the benefits state. More recent examples are Shanghai Overview 27 and Guangzhou in China. An even more Figure 6 Quicker geographic convergence in basic living standards in Malaysia recent (and perhaps more generally appli- % difference between min. and max. poverty rates among Malaysian states cable) example is Bogotá in Colombia. 70 The experience of successful urban- 60 1976 izers indicates that the basis of successful rural-urban transformations is a set of spa- 50 tially blind policies--"institutions" in the 1970 40 shorthand of this Report. Investments in 1984 infrastructure that connects places form 30 1990 the second tier. Geographically targeted 1995 interventions should be used only when the 20 2002 challenge is especially difficult, but should 10 always be used together with an effort to improve institutions and infrastructure. 0 0 500 1,000 1,500 2,000 2,500 3,000 3,500 4,000 4,500 GDP per capita (2000 US$) Area (territorial) development policies that integrate nations Source: Malaysia Economic Planning Unit 2008. Some parts of a country are better suited in the slower-growing states, implying that for agriculture, others for industry, and tax and transfer mechanisms worked well. still others for services. And as industry Such impatience with spatial inequality and services flourish, the spatial distribu- in living standards is paying off in other tion of economic activities must change.11 countries such as China, Egypt, Indonesia, No country has grown to riches without Mexico, Thailand, and Vietnam. changing the geographic distribution of its But not all countries have experienced people and production. geographic convergence in the Millennium A rising concentration of people and Development Indicators, such as child mor- production in some parts of a country has tality, maternal health, basic education, safe marked economic growth over the last water, and sanitation. What should they do? two centuries. To fight this concentration The answer lies in integrating lagging and is to fight growth itself, and policy makers leading areas, using policies that are tailored must show patience in dealing with these to the level of difficulty of integration. While imbalances. But aided by government poli- economic motives are important, social and cies, successful development also has been political conditions influence the speed of marked by falling disparities in living stan- these spatial changes. The location choices dards between places favored by markets that people make reflect the strengths and and those less fortunate. Policies can speed inclinations of societies and political struc- up the convergence in basic living stan- tures. Poverty maps provide a snapshot of dards, so that people in the least-fortunate where the poor are concentrated (high pov- places do not have to wait for basic public erty mass--that is, the "poor people"), and amenities until their nations reach high which places are the poorest (high poverty income levels. The experience of success- rate--that is, the "poor places"). These maps ful developers also justifies impatience in can tell us a lot about the social and politi- equalizing basic living standards. cal conditions in a country: the movement of Consider Malaysia. Economic growth poor people may best reflect the constraints to and government policies have reduced mobility, because they have the most reason poverty and improved living standards, to move and the fewest resources to do so. speeding progress toward meeting the Mil- Using information on where poor people lennium Development Goals. But in the are located and which places are poor, the early years of growth (between 1970 and policy response can be calibrated to coun- 1976), poverty rates between different states try conditions. diverged briefly, to later converge as they declined for all states (see figure 6). Health · Countrieswithsparselypopulatedlagging indicators (infant mortality) declined more areas. In China the highest poverty rates 28 WORLD DEVELOPMENT REPORT 2009 Map 7 Three types of countries, differing challenges for area development a. China: Poverty rates are high in the west, but most poor people are in the east Proportion of poor (%) < 17.5 17.6­35.9 Poverty density 36.0­51.6 Each dot represents 51.7­70.3 70.4­81.1 50,000 poor persons b. Brazil: Poverty rates are high in the north and northeast, but most poor people live along the coast Proportion of poor (%) < 12 12­25 Poverty density 25­35 Each dot represents 35­45 >45 5,000 poor persons c. India: Poverty rates are high in the central states, and many poor people live there Proportion of poor (%) 6.4­9.7 9.8­16.7 16.8­24.8 Poverty density 24.9­35.4 Each dot represents 35.5­46.6 no data 50,000 poor persons Source: WDR 2009 team (see chapter 8 for details). Overview 29 are in the western provinces, but the poor ging states, home to more than 60 percent are concentrated in the southeast and cen- of the nation's poor (see map 7, panel c). tral areas (see map 7, panel a). Economic People live there for a reason: it is a fer- density and population density overlap. tile plain and was the cradle of Indian The country has few divisions--linguistic civilization. But their location is less for- and other barriers are not high--and peo- tunate now, as the world has changed. ple,including the poor,can move to reduce Labor mobility is limited because of lin- their distance to density. Spatially blind guistic and class divisions. Mobility has institutions that ensure well-functioning not been helped by policies that sought to land markets, enforce property rights, and revive growth in these lagging provinces deliver basic social services such as school- through subsidized finance and preferen- ing and health care can be the mainstay of tial industrial licensing. The debate is now an economic integration strategy to reduce shifting toward economic integration-- the economic distance between lagging policies more consistent with mobility of and leading areas. Chile, Egypt, Honduras, labor such as interregional infrastructure Indonesia, Russia, Uganda, and Vietnam and better health and education services. are other examples of countries where the These policies and the interstate migra- area development challenge is unidimen- tion they encourage will, if given time, sional--the main problem is distance. reduce the divisions that have made the ·Undivided countries with densely popu- distances long between leading areas lated lagging areas. In Brazil the poverty and densely populated lagging areas. rates are highest in the north and north- In the meantime, these areas may need east: eight of the ten poorest states are in a helping hand--from geographically the northeast, the other two are in the targeted incentives that encourage local north (see map 7, panel b). But the eco- production. Another country with a nomic mass and the concentration of three-dimensional integration agenda of poverty are highest in the urban agglom- distance, densely populated poor areas, erations near the coast, from the poor and domestic divisions may be Nigeria. northeast to the thriving southeast. Eco- In such places, the policy response has to nomic and population densities coincide be a blend of spatially blind, connective, only partially. The poverty-related symp- and targeted policies. toms are those of a country where within- Governments should not be faulted country divisions such as ethnolinguistic for being impatient with markets, and differences and political fragmentation for trying to help lagging areas. But tar- are low, but where population densities geted interventions should be designed to are--for historical and policy-related work with the institutional reforms and reasons--in the "wrong places." Bangla- the investments in infrastructure. Expe- desh, Colombia, Ghana, and Turkey have rience suggests that incentives should not similar conditions. In such places the be provided for activities that depend on pull of agglomeration economies in lead- agglomeration economies or international ing areas and the mobility of labor may market access. Incentives for agriculture not be strong enough to induce concen- are prime candidates in these largely rural tration and convergence. The problems and agrarian areas. Relying mainly on tar- of "long distance and wrong density" geted incentives for industry--as India must be met by a two-pronged policy did for decades--will not help the lagging of economic integration: spatially blind states improve living standards to levels in institutions should be augmented by spa- the leading states. tially connective infrastructure, such as interregional highways and railroads and Regional integration to increase access improved telecommunications. to global markets ·Divided countries with densely popu- The merits of global versus regional trade lated lagging areas. In India more than agreements have been debated for years. 400 million people live in the central lag- The debate is now largely concluded. Where 30 WORLD DEVELOPMENT REPORT 2009 Figure 7 Northeast, Southeast, and South Asia have been catching up to developed nations if the scale of production is big, and that Average annual growth rates of GDP per capita, 1960­2006 requires reaching the big markets of the Northeast Asia Northern Hemisphere. Other high-income countries What do late developers have to do to accelerate development? The common con- Southeast Asia & Pacific dition is division--that is, thick borders. South Asia What differs is their distance from large OECD countries world markets and whether or not there is Northern Africa a large country in their neighborhood (see Central Asia, Caucasus & Turkey map 8, panel b). Central America & Caribbean Eastern Europe & Russian Federation · Countries in regions near large world Western Asia markets. For countries near large mar- South America kets, regional and global integration does Southern Africa not require geographic differentiation. Eastern Africa Spatially blind measures such as improv- Western Africa ing economic policies and the invest- Central Africa ment climate will attract capital and ­1 0 1 2 3 4 5 6 technology from the more sophisticated Percent markets nearby. Their underused talent Source: WDR 2009 team. and cheaper labor are powerful draws. Whether they lag or lead within the regional or bilateral pacts do not discourage region is hardly relevant; the presence of trade with countries in other regions, and a sun nearby makes them all small plan- where they are accompanied by measures ets. Mexican exports to the United States to facilitate the flows of goods, people, and are about 1.7 percent of the U.S. econ- finance--such as infrastructure and com- omy. Mexico should build even stron- pensatory mechanisms--they can help. ger links with the United States. But for Otherwise, they are not worth the trouble. other countries in Central America, the This Report does not reopen that payoffs to infrastructure connections to debate. Instead, it takes up the question Mexico are small--the market in North of how developing countries can best gain America dwarfs all of Central America's. access to markets within their neighbor- And market access likely depends most hoods and across the world. Geography on economic stability. Spatially blind matters greatly in deciding what is needed, institutions should be able to integrate what is unnecessary, and what will fail. But Central America with world markets. with the right mix of policy actions, even The same is true for Eastern Europe and countries in parts of the world that have North Africa. Countries in these regions been left far behind can overcome their have better-than-average market access, geographic disadvantage. The way to tell if though depending on their economic the actions are paying dividends is whether policies and regulations, this access is market access improves noticeably. not uniform even within these regions Some regions of the world have done (see map 8, panel c). better than others (see figure 7). Countries · Countries in regions distant from in these regions now have thinner eco- large world markets that have a large nomic borders (see map 8, panel a). They economy. To integrate regions more can afford to have thin borders, because distant from large world markets but their neighbors are prospering too. For with a sizable economy--East Asia, them, regional markets are world markets. Latin America, Southern Africa, and Others, like the East Asians, have allowed South Asia--such spatially blind mea- production relationships to grow strong sures are just as necessary, but they may and cut paths even through thick borders. not be sufficient. For lagging countries But specializing can increase efficiency only in these regions, such as Mongolia, Overview 31 Map 8 Market access distinguishes world regions a. Borders are thicker in developing regions b. The size and access to markets differs greatly by region Real market access relative to the United States, 2003 < 0.040 0.040­0.090 0.091­0.240 0.241­0.910 > 0.910 No data c. The three D's suggest a simple taxonomy of the world's neighborhoods Regions with High-income countries Countries close to world markets Large countries far from world markets Small countries far from world markets Sources: Panel a: WDR 2009 team (see chapter 3 for details); panel b: Mayer 2008 (see chapter 9 for details); panel c: WDR 2009 team (see chapter 9 for details). 32 WORLD DEVELOPMENT REPORT 2009 Nepal, Paraguay, and Zimbabwe, some can hardly be blamed for worrying of the paths to world markets may go most about their own poor, and not through their larger neighbors. Brazil, their less fortunate neighbors such as China, and India are attractive to inves- Burkina Faso or Burundi. Indeed, see- tors because of their potential market ing the benefits of regional cooperation, size, and these "home market effects" they have made repeated efforts to fos- can generate the impetus for special- ter integration in their neighborhoods. ization and help their enterprises com- The ECOWAS even includes a clause pete in world markets. A qualification: that allows workers to cross borders, a for market access, the relevant measure stage of integration rivaled only (and of distance is economic, not Euclidean. only recently) by the EU. It also has With a combination of bilateral accords, tried to share regional infrastructure. inspired transport policies, and aggres- Other such regions are Central Africa, sive specialization in primary products, Central Asia, and the Pacific Islands. Chile reduced distance to North Amer- Countries in such regions face a three- ica and built global rather than regional dimensional challenge (see "Geogra- links. But such cases are exceptions. For phy in Motion: Density, Distance, and the smaller countries in these regions, Division in Sub-Saharan Africa"). A both institutional reforms and regional combination of efforts to improve insti- connectivity will be necessary for eco- tutional cooperation and regional infra- nomic integration. structure investments is needed--but it · Countries in regions distant from world is not enough. Targeted incentives also markets without large economies. The will be necessary, through preferential most difficult challenges are for the access to developed country markets, countries in parts of the world divided perhaps made conditional on regional by thick borders, distant from world collaboration to improve institutions markets, and without a large country and infrastructure. that can serve as a regional conduit to world markets, as Brazil and India Everyone should support the efforts of might. For these regions, economic these "bottom billion" countries to inte- geography poses a three-dimensional grate their economies, within and across challenge. Côte d'Ivoire or Tanzania borders. A billion lives depend on it. We are familiar with the sectoral transformations needed for economic growth--the changes in work and organization as agrarian economies become industrialized and service oriented. This Report discusses the spatial transformations that also must happen for coun- tries to develop. Higher densities, shorter distances, and lower divisions will remain essential for economic success in the foreseeable future. They should be encouraged. With them will come unbalanced growth. When accompanied by policies for integration calibrated to the economic geography of nations, these changes also will bring inclusive development--sooner, not much later. Navigating This Report I n 1971 Simon Kuznets, a Russian émigré may even be one of the main lessons of the who had built his career in the United twentieth century. After Kuznets left Russia States, was awarded the Nobel Prize in in 1922, Soviet planners implemented one Economics "for his empirically founded approach to economic geography, and the interpretation of economic growth, which United States implemented another. The has led to new and deepened insight into the Soviet strategy forced people to move to economic and social structure and process the north and east and to spread out eco- of development."1 In his prize lecture, Kuz- nomic production. Meanwhile, Americans nets summarized the structural changes that moved voluntarily toward the south and the accompany economic growth, emphasizing west, but production became more concen- "the shift away from agriculture to nonag- trated. Within five years of Kuznets' death ricultural pursuits and, recently, away from in 1985, the Soviet Union would collapse. industry to services."2 These are the sectoral At the time, Russia's per capita income was changes in production needed for nations to a quarter that of the United States. Spatial prosper. Nations do not develop by merely inefficiency was not the only reason why doing more of the same thing. They must do the Soviet Union fell. But it could not have different things, and do them better. helped. Over the years, this has been confirmed As Russia has moved from plan to mar- so often that it now seems almost obvious. ket, spatial efficiency increased. Between Less obvious but no less important are the 1989 and 2004, almost all new firms chose spatial transformations needed for these locations with the best access to Moscow, structural shifts. Some places are suited St. Petersburg, and international markets.3 for farming, others for industry, yet others Over the past three decades, researchers for services. As economies become indus- have been documenting the changes in trialized and more people are employed economic geography needed to stay spa- in services, their shapes must change, too. tially efficient as technology advances and These changes, involving social adjustment production structures change. They have as much as the economic, can take time. studied the effects of larger populations, The economic world is not frictionless. The globalizing markets, and international "what" and "how" of economic production borders on the location of people and pro- cannot be decided without deciding the duction. They are starting to assess how "where." governments can help or hurt these trans- For policy makers, especially, it is formations. This Report draws on this work important to understand these changes and and its implications for public policy. to appreciate the market forces that shape Government policies are important. them. This understanding can be the differ- With development, people and production ence between prosperity and stagnation. It become more concentrated--in towns and 33 34 WORLD DEVELOPMENT REPORT 2009 cities, and in areas of countries closer to of this chapter summarizes the Report's domestic and international markets. While scope, clarifies its terms, and outlines its economic activity concentrates in some structure. parts of a nation or the globe, many people may be spread out over the countryside or Scope in places distant from prosperity, perhaps Governments intervene (usually incorrectly) opening sizable geographic disparities in to spread the benefits of economic growth living standards. This Report discusses why more evenly across space. Even when the this happens, and assesses what has been imperatives are political, they have economic most effective in altering the economic consequences. And even if the objectives are geography of developing countries. Eco- economic, they have social and environmen- nomic activity will concentrate in any case. tal effects. Policy makers thus face sharp But managed one way, as the United States tradeoffs and must compromise. The eco- did, it can foster growth and integration. nomic costs of mistakes can be large and last- Managed another way, it can result in dis- ing: recognizing the importance of economic integration and despair, and even conflict. geography means realizing that once produc- The Report covers a broad and seemingly ers and people make decisions on where to disparate set of phenomena that span the locate, they can be difficult to reverse. spectrum from local to national to inter- Governments can do better by promoting national scales, from human to physical to the market forces that deliver both the con- political geography, and from national and centration of economic production and the global institutions to targeted interventions. convergence of living standards, and aug- To keep the inquiry disciplined requires ment them with policies to ensure afford- emphasizing some aspects of spatial trans- able basic services everywhere. They can do formations and leaving others out. The rest this by helping people and entrepreneurs take advantage of economic opportunities, wherever they arise. The market forces that BOX 0.1 What this Report is not about help most are agglomeration, migration, and specialization. Their economic benefits are To keep the Report focused, several are found in a much greater degree in a the subject of this Report. Their social and important aspects of the spatial crowded city than in a country village, where one's individual concerns are the environmental implications are not consid- transformations do not get the atten- concern of everyone.a ered in detail (see box 0.1). The unintended tion they would in a fuller study. The Migration also can have vastly dif- main aspects not considered--ex- social and environmental effects of market ferent e ects across societies, both in cept when emphasizing or qualifying forces are important policy matters. But they the places people leave and to those the most important messages--are deserve more space than can be covered in a places they go. It almost always brings the social and environmental e ects economic rewards, but as the anti- report that shows how economic geography of a changing economic geography. immigrant sentiments in many coun- is reshaped during development. Agglomeration--the growth of tries show, it also means more risk. The Report describes the geographic cities--can have social and environ- Specialization of production made transformations needed for development. mental e ects that are bene cial and possible by falling transport costs can some that are detrimental. Cities help It analyzes these changes using the insights come at an environmental price. Cod to break down societal stereotypes from economic history and recent research. is caught o Norway, transported by and increase cohesion. Most progres- It then revisits the policy debates on urban- plane to China to be cleaned, and sive movements throughout history then own back to Norway to be sold. ization, regional development, and inter- have had urban origins. But so have Such specialization based on natural national integration. This is the 31st World the most violent. The propensity of endowments ( sh in Northern Europe, Development Report, and the issues it covers people to commit crimes is believed people in China) helps both Norwe- have been visited by earlier Reports. But here to be greater in cities. And while cit- gian consumers and Chinese workers, ies allow individualism and creativity the facts, analysis, and policies related to spa- but the cod now has a longer carbon and break down social barriers, they tial transformations are the major focus, and trail. The environmental e ects of also break societal ties: the Report is structured accordingly. urbanization and transport are con- The cities have always been the cradles sidered in this Report, but only when of liberty, just as they are today the they qualify the Report's messages. Terms centres of radicalism. Every man of the world knows that isolation and solitude a. Weber 1899, p. 432. To formulate simple messages that are useful to policy makers requires an uncomplicated Navigating This Report 35 terminology. The Report uses some terms density of about 3,000 persons per square that may not be familiar to readers, intro- kilometer. The population density in the duces others, and uses yet others as short- city is about 13,000 persons per square hand. This section clarifies the terms that kilometer. the rest of the Report uses consistently. · Country. The national scale encom- passes the 23 provinces, five autono- Spatial scales--area, country, and region mous regions, and four municipalities Throughout the Report, the analysis is pro- (Shanghai is one of them) that make vided at three geographic scales--local, up China, covering about 9.6 million national, and international. The policy con- square kilometers. The distance between cerns that correspond to these spatial scales the western province of Xinjiang and the are, respectively, the speed and sustainability dynamic coastal areas in the east is more of the rural-urban transformation, the ter- than 4,000 kilometers. Restrictions on ritorial disparities in production and welfare internal migration can make the eco- within countries, and the same disparities nomic distance seem much longer. across countries and world regions. The units · Region. The international scale con- thatcorrespondtothesespatialscalesarearea, sists of China and its East Asian neigh- country, and region. These terms are used bors including Japan, Mongolia, and the consistentlythroughouttheReport.An"area" Republic of Korea. The region is divided is the same as a "territory," the target of ter- by borders, some thick, some thin. ritorial development policies. In Anglophone countries, it is the same as a "region" within a This Report uses the notion of "natu- country, as in the debates on "regional devel- ral" neighborhoods, defined by elements opment." Area is used here to avoid confusion of human, physical, and political geogra- with another spatial scale, the international, phy. The World Bank commonly classifies because "region" also describes a group of all low- and middle-income countries into countries, such as South Asia, which includes six regions, and groups all high-income India and its neighbors. countries together, regardless of their loca- To fix the terms, consider the three geo- tion. This Report classifies the world into graphic scales of the Shanghai metropoli- 16 regions that include both developed and tan area, the country of China, and the East developing countries, using geographic Asia region (see map 0.1): proximity as the most important crite- · Area. The local scale is the municipality rion (see box 0.2). It is also more detailed. of Shanghai--which includes the city of Sub-Saharan Africa, for example, has four Shanghai and neighboring cities, towns, regions--West, Central, East, and Southern. and villages in an area of about 7,000 East Asia and the Pacific has three--North- square kilometers, with a population east, Southeast, and the Pacific Islands. The Map 0.1 Three geographic scales--area, country, and region Shanghai, China, and East Asia exemplify the local, national, and international scales The first geographical scale The second geographical scale The third geographical scale The area around Shanghai Province The country of China The East Asian region Shanghai SHANGHAI PROVINCE SHANGHAI PROVINCE Source: WDR 2009 team. 36 WORLD DEVELOPMENT REPORT 2009 BOX 0.2 This Report's regions are more detailed than the World Bank's This Report is about geography and eco- sovereign states, the Report uses an aggre- the context, the analysis in this Report nomic development, focusing more on spa- gation of countries that is more detailed ignores the income of countries within tial variability of conditions and outcomes than the six standard World Bank regions, a region--say, where regional growth than economic analysis usually does. Where which can hide signi cant variation. spillovers from industrial to developing appropriate, it uses countries or areas Adapting the United Nations geo- countries are of interest--or treats the within countries as the units of analysis. But graphic regions but remaining consistent Organisation of Economic Co-operation where the emphasis is on regional integra- with World Bank regions yields the 16 and Development (OECD) and other high- tion and interactions between neighboring regions displayed here. Depending on income economies separately. Regions used in this Report EASTERN EUROPE & RUSSIAN FEDERATION WESTERN NORTH EUROPE AMERICA CENTRAL ASIA, CAUCASUS & TURKEY NORTHEAST ASIA WESTERN NORTH AFRICA ASIA SOUTH CENTRAL AMERICA ASIA & CARIBBEAN WESTERN AFRICA SOUTHEAST ASIA EASTERN CENTRAL & PACIFIC AFRICA AFRICA SOUTH AMERICA SOUTHERN AUSTRALIA & AFRICA NEW ZEALAND Source: WDR 2009 team. term "region" is used throughout the Report interpretation. Density generally signifies to refer to these 16 groups of countries. the intensity of economic activity on a unit While the choice of area or region can of land, say, a square kilometer. Data limi- be arbitrary, these spatial scales conform tations can force compromise: since pro- well to the levels of policy making. This duction and population densities are closely Report aims to inform policy making at related, and production data are less easily these three levels--subnational, national, available, population density is sometimes and international. used as a proxy for economic density. It can get a bit confusing. London is probably the Spatial dimensions--density, distance, city with the highest economic density in and division the world, but Mumbai, with 30,000 people To describe the geographic transforma- per square kilometer, is the most densely tions that accompany development, the populated. Distance signifies the costs of Report introduces the use of three spatial getting to places with economic density. dimensions--density, distance, and divi- While density and distance relate closely sion. These dimensions help the reader see to human and physical geography, division development in real space--in three dimen- refers more to sociopolitical geography. sions, in other words. The terms are easy Religion, ethnicity, and language are among metaphors, but they also have a technical the main attributes that lead to divisions Navigating This Report 37 between places. While divisions are great- density--such as the population per square est across nations, they can be considerable kilometer (as in chapters 1 and 7) or the within countries as well. places where more of a nation's poor people These dimensions are measurable. But live (as in chapters 2 and 8), it is qualified unlike height, length, and breadth, for accordingly. example, the geographic dimensions are Distance can be measured with some not orthogonal. Better analogs for the three precision, but where infrastructure is sparse, dimensions are a person's height, weight, straight-line distance is different from road and age, which are related. Likewise, as dis- or rail distance. Many other factors, such as tances increase, it is likely that divisions also the availability and affordability of trans- get stronger. Density, distance, and division port services, determine actual accessibil- are best illustrated by market access, an ity. Where such information is available, indicator of economic opportunity for a it is used. Chapter 1, for example, reports a location that tells the size of the potential uniform measure of urbanization based on markets in its vicinity, and the ease of reach- places that both have minimum levels of ing them. Market access across geographic population density and are within an hour's scales determines where economic activity travel time to sizable settlements. In com- can thrive and thus where firms will locate puting this "agglomeration index," the qual- and populations will grow. ity of transport infrastructure is taken into Using this concept of market access, the account. Division is associated with interna- three dimensions are defined as follows: tional borders, because they usually impede the ease of exchange or travel. But not all bor- · Density indicates the size of economic ders imply divisions. Those in the European output or total purchasing power per Union (EU), for instance, have increasingly unit of surface area--say, a square kilo- ceased to reflect divisions between countries. meter. It is highest in large cities where And not all divisions imply international bor- economic activity is concentrated and ders. Where religious, ethnic, and linguistic much lower in rural neighborhoods. differences are manifest spatially, there can · Distance measures the ease of reaching be divisions within countries. markets. It determines access to oppor- There is a correspondence between the tunity. Areas far from economically geographic scales and dimensions. Locally, dense centers in a country are more within an area, the most important dimen- likely to lag. sion is density, because generally distances · Division arises from barriers to eco- are short and divisions few. Nationally, the nomic interactions created by differences most important dimension is distance to in currencies, customs, and languages, density; divisions within countries tend which restrict market access. It is most to be fewer, though they can be serious in relevant in an international context. some countries. Internationally, across a regional or global spatial scale, distances The concept of distance is also relevant and divisions are usually more serious. internationally. The difference between Using these three dimensions, the Report distance and division is that distance mod- summarizes the geographic transformations ulates access to economic opportunity in a needed for development (part one). It shows more continuous way--a distance decay. how market forces drive these transforma- Division, by contrast, presents discrete tions (part two). And it assesses how govern- barriers to access and economic integra- ments can augment these forces to sustain tion. It can be seen as increasing economic growth and reduce poverty (part three). distance or travel time for a unit of physical (or Euclidian) distance. These definitions are not scientifically Instruments for integration-- exact. But the terms are used consistently institutions, infrastructure, and in the Report. When "density" is used, it interventions means economic density: production per Through good policies, governments can area of land. When any other measure of promote economic integration between 38 WORLD DEVELOPMENT REPORT 2009 places where economic production is education and health, and electricity, concentrated and places that are lagging. water, and sanitation. Systems for col- Some of these policy instruments are spa- lecting taxes and financing the spending tially explicit, like a slum-upgrading pro- associated with these services are also gram in a city, a Brazilian state's fiscal best designed without specific places in incentives to a U.S. automobile company, mind. or the EU's structural and cohesion funds. · Infrastructure is the summary term for Others are intended to be universal in their all spatially connective investments and coverage, including compulsory and free the associated rules and regulations. It basic education for all children, such labor includes roads and railways, airports market regulations as minimum wage laws, and air transport systems, telecommu- and the enforcement of property rights. nications, and the Internet. Between these spatially targeted programs · Interventions is shorthand for all spa- and "spatially blind" policies are invest- tially focused incentives. These include ments and regulations that connect places, regulations and investments that favor such as roads, airports, and communica- some places, such as export processing tions systems. zones. They also include place-based pro- In their current form, the debates on grams--such as slum-upgrading schemes how governments can foster rural-urban like Rio de Janeiro's Favela Bairro, or transformation, help lagging areas reduce Superintendency for the Development of poverty, and--in the poorest nations in the the Northeast (SUDENE), Brazil's devel- world--improve access to world markets all opment agency for the lagging Northeast, emphasize geographic targeting. The debate or the Everything But Arms initiative of on how to promote healthy urbanization is the EU, which gives the least developed polarized between an emphasis on villages, countries preferential trade access to where a majority of the world's poor still European markets. live, and a belief that the way out of poverty lies in cities; if urban poverty increases, the Because these definitions do not con- focus shifts from villages to slums. Moti- form strictly to common usage, additional vated by within-country spatial disparities clarification is necessary: in living standards, the debate on territorial · First, spatial blindness does not mean development tends to be similarly fixated spatial neutrality. A progressive tax sys- on promoting economic growth in lagging tem, for example, may not be neutral in areas. At the international level, preferential its effects or outcomes. Cities may end market access for the least developed coun- up contributing more in taxes than the tries can end up dominating policy discus- countryside, and richer states may con- sions. Part three of the Report reframes tribute more than those that are poorer. these debates, calling for a shift from spatial But the guiding principle is that tax rates targeting to integration. differ not by place alone, but by the attri- The policy instruments for economic butes of firms and families that happen integration can be classified in three cat- to be located there. egories, based on how explicitly place is · Second, in the common use of the term, considered in their scope and design: infrastructure includes nonconnective · Institutions is shorthand for all the pol- investments such as water supply and icy instruments that are spatially blind. energy. In this Report, infrastructure is These are the amenities that governments reserved for the spatially connective com- should provide to everyone, regardless ponents. Nonconnective public utilities of place. The word "institutions" con- are included in institutions, as for such notes universality, and includes mecha- basic services as sanitation. nisms for financing and delivering such · Third, each of these categories includes basic amenities as the administration of all three tools of government policy-- justice, public security, the regulation of taxes, transfers and public expenditures, land, labor, and capital markets, primary and regulations. Navigating This Report 39 · Finally,governmentinitiativescaninclude more than one instrument. Slum devel- BOX 0.3 This Report's message is not anti-equity opment can include steps to make urban land markets work better by formalizing Policies for spatially balanced growth The Gini coe cient for the spatial property rights, improving streets, and are often justi ed by equity. The EU inequality of GDP is 0.53 for the United States and 0.41 for the EU. For offering monetary incentives for some of describes its territorial policy as gov- erned by the principle of solidarity population, the coe cients are 0.54 the slum-dwellers to relocate. because it "aims to bene t citizens and 0.32, respectively. For subna- and regions that are economically tional areas in the EU and states in the Structure and socially deprived relative to United States, the numbers change, The main finding of this Report--at all EU averages."a The policy seems to but the conclusion is the same.b Spatial disparities in living stan- three spatial scales--is that economic equate social and spatial equity-- equality across individuals, and the dards. EU-15 countries have greater development is not smooth, linear, or neat. equality of living standards across spatial inequality in per capita income The processes of economic growth leave states and countries. This Report, by and unemployment rates, two com- behind a bumpy landscape, with economic contrast, argues in favor of the ben- mon indicators of individual living mass concentrated in some places. Liv- e ts from geographic concentration standards in high-income countries. ing standards in such places--especially of economic production. But it shows GDP per capita, for example, exhib- rising prosperity, good access to educa- that in the earlier stages of develop- ited greater variation across EU areas than it did across U.S. states in 2005. tion and health facilities, and safe shelter, ment, increased concentration is associated with spatial divergence in Although production is more concen- water, and sanitation, some of the most living standards such as income. So is trated geographically in the United urgent among the Millennium Develop- this Report's message anti-equity? States, people are also more likely to ment Goals--improve faster than where No. It is important to distinguish live where production is, so GDP per there is less economic activity, widening between three types of disparities: capita varies less. The same is true of the spatial disparities in welfare. But where spatial disparities in economic pro- unemployment rates. In the United there is sustained economic growth, the duction, spatial disparities in living States, the state with the highest unemployment in 2007 (Michigan) convergence in living standards begins to standards, and social inequality. Spatial disparities in economic had an unemployment rate of 7.2 supplant divergence. Nations become both activity. In both the United States percent, 2.8 times the lowest unem- spatially efficient and equitable (see box and the EU-15 countries, gross ployment state (Hawaii). But in the EU 0.3). The challenge of development is to domestic product (GDP) and popula- in 2006, the ratio was 8.1. There is less institute policies that allow--even encour- tion have lumpy spatial distributions. spatial inequality in living standards age--"unbalanced" economic growth and In the United States, three states in the United States. yet ensure geographically balanced devel- (California, New York, and Texas) Social inequality. While spatial inequality in living standards is greater opment outcomes. generated 21 percent of national GDP in 2005. The same three states have in the EU than in the United States, the opposite is true for social inequality The facts 19.8 percent of the U.S. population, but only 12.8 percent of the country's between individuals. During the past Part one of the Report presents the facts land. Meanwhile, 10 EU subnational few decades, the Gini coe cient for about the spatial transformations--the areas were responsible for 20.5 per- the United States has been about 0.40, changes in economic density, distance, cent of the EU's GDP in 2005. These compared with 0.33, 0.28, and 0.23 for and division. Chapter 1 shows that devel- areas have 16.9 percent of the EU-15's the United Kingdom, Germany, and Austria, respectively.c opment is accompanied by the rising den- population, but only 8 percent of its land. So, in both cases, economic sity of human settlements: no country has Contributed by Mark Roberts. activity and population are con- reached high income without this rise in a. http://europa.eu/pol/reg/overview_ centrated. But spatial inequality of en.htm. density. Chapter 2 expands the scale and production and population is higher b. Puga 2002. shows that development is also accompa- in the United States than in the EU. c. Burkey 2006. nied by the greater concentration of eco- nomic activity in areas of countries closer to economic density. Chapter 3 incorpo- The long experience of countries shows rates international divisions that slow, but that income differences between leading do not prevent, the concentration of eco- places and following places first diverge nomic activities in some countries. At the and then converge, but only in the more local, national, and international scales the dynamic areas, countries, and regions. At pattern is similar: rapidly rising concentra- each of the three spatial scales, it pays to tions at the early stage and then a slowing be in dynamic neighborhoods. Economic down. growth leads to congestion in cities--and to 40 WORLD DEVELOPMENT REPORT 2009 the growth of towns and cities that are well scale associated with places, not plants--in connected to fast-growing agglomerations. producing goods, services, and ideas. Places This pattern is repeated at the national and of different sizes provide varying agglom- international levels. Expanding economic eration benefits, and congestion associated activity spills over to areas and countries with spatial concentration leads to a portfo- that are--in economic terms--near places lio of places that facilitate economic growth, doing well. with different parts in the lead depending on the stage of development. The insights Chapter 5 explains the interaction The second part of the Report is the "engine between scale economies and factor mobil- room." It exploits the main insights from a ity, focusing on the migration of workers. quarter century of work spanning several Chapter 6 explains the nonlinear rela- subdisciplines in economics, such as indus- tionship between transport costs and the trial organization, urban economics, inter- geographic concentrations of production, national trade, and economic geography. focusing on intraindustry trade, which is Distilled to its essence, the engine works especially sensitive to transport costs. These through a three-way interaction between chapters summarize the new insights pro- scale economies, the mobility of workers vided by the three-way interaction between and entrepreneurs, and the costs of trans- scale economies, factor mobility, and trans- porting and communicating between places port costs--and their implications for (see figure 0.1). development policy (see box 0.4). Firms are generally more productive when they locate in large places and when The policy framework they operate at a relatively large size. If it Circular causation, unevenness, and spill- is relatively easy to transport produce, the overs make for a world in which poli- scale can be even higher, since the poten- cies can promote economic growth and tial market is bigger. Workers move to these improve social welfare beyond what mar- places, bringing with them both a supply of kets yield, because well-executed policies labor and a demand for goods and services. can set these transformations in motion or As people become more mobile and as speed them up. transport and communications costs fall, These features of economic develop- these economies of scale create a circular ment also make policy making a diffi- and cumulative causation, where economic cult enterprise. Part three of the Report activities become even more concentrated reframes three important policy debates, spatially. Rising concentration inevitably using a principle derived from its first two leads to congestion, which slows the pro- parts: for developing countries to realize cess and eventually reverses it. Declines the benefits of both spatial concentration in transport costs first make concentra- of production and convergence in con- tion possible, and then, when they fall low sumption, development is best facilitated enough, they make it unnecessary. by economic integration. Using the three Part two discusses these interactions in dimensions--density, distance, and divi- some detail, summarizing more than a cen- sion--described in part one, and the (mal) tury of experience and the novel insights functioning of pivotal markets at each spa- that come from a generation of research tial scale--land, labor, and intermediate recognizing how factor mobility and fall- inputs--analyzed in part two, the chapters ing transport costs feed economies of scale in part three provide a simple framework (see box 0.2). They should change what we and illustrate its workings through real- can expect from the markets, and what gov- world policy experience. At each of the ernments can and should do to facilitate the geographic scales, the response rule is the concentration of production and promote same--an instrument per dimension. Here the convergence in living standards. is a somewhat oversimplified summary, Chapter 4 provides evidence of agglom- using examples from only the local scale eration economies--increasing returns to (chapter 7): Navigating This Report 41 BOX 0.4 Fresh insights from economic geography: concentration, convergence, and integration Over the past two decades, new analysis transport costs, but the relationship is not pulled rapidly through the development has changed the way we think about the linear. When these costs are high, rms process. In the neoclassical world, being location of production, trade, and develop- avoid shipping their output long distances behind can be an advantage--places lag- ment. The analysis builds on two elements. by spreading out their production. Firm ging farther can catch up faster. But with First, large markets are disproportionately location is then mostly determined by agglomeration economies, the farther attractive for rms producing with scale local access to immobile demand, such as behind an area, country, or region, the economies. Firms with a larger home from farmers and miners. For intermediate tougher it is to catch up. What should lag- market have more sales that, with scale values of trade costs, it becomes feasible ging places do? economies, imply lower unit costs and to supply markets from a distance, and Integration is the answer. Because more pro ts, which encourage existing places that get an advantage in market size both high and low trade costs can rms to expand and attract new rms. Sec- build on it and take o relative to other encourage production to spread out, ond, large markets are big partly because places. When trade costs fall to low levels, lagging areas, countries, or regions could many rms and consumers locate there. it matters little whether one sells and buys in principle turn to either import substi- Market access and mobility creates a circu- locally. Firm location is then determined tution or export-oriented industrializa- lar and cumulative causation. A large mar- mostly by the local cost of immobile fea- tion. But import substitution becomes ket attracts rms and workers--and the tures, including the cost of land and hous- less feasible as a development strategy demand for intermediate inputs by rms ing, but also by the ability to have face-to- over time. Why? Because it limits for- and the demand for nal goods by workers face interactions or to nd a good match in eign access to local immobile demand, make the market even larger, attracting a specialized labor market. So once trade whereas export-oriented industrialization more rms and workers, and so on. costs decline su ciently, some activities reduces the cost of purchasing foreign This is both good and bad news for will spread out in response to cost di er- intermediates for processing and export. places with poor initial conditions. It is ences, and others will remain concentrated. The falling share of agriculture and the good because it means that rm location Convergence is the objective. The tendency of manufacturing and services is not as constrained by nature as theories forces of market access and mobility to agglomerate have reduced the share of based on comparative advantage would have implications for the way we think demand in lagging places. And the frag- have us believe. Places with poor endow- about convergence. The view of develop- mentation of production has made access ments can sustain concentration of activity. ment as smooth and linear gives way to a to intermediate inputs more important. It is bad news because the circle of market lumpier nonlinear process. As a country Both make development strategies based access and mobility produces persistence. grows, new producers locate close to on fencing o local immobile demand Once a place gets far ahead, it is di cult for existing production, widening the pro- hopeless. The observation that some lagging areas to catch up. While agglom- duction di erences between lagging and developed countries or provinces indus- eration raises the cost of labor, rms do leading places. When wage gaps become trialized while being closed to trade is of not move to low-wage areas, because this wide, industry starts to spread to places little help to lagging areas, countries, or would mean forgoing the bene ts of prox- that have low wages. But this does not regions today. The ones left behind are so imity to suppliers and customers. lead to steady development of all places. small relative to the world economy that Concentration is the rule. The strength Instead, development takes place in isolation is no longer a feasible option. of the agglomeration forces created by waves, where some areas or countries are market access and mobility depends on drawn in sequence out of poverty and are Contributed by Diego Puga. ·For one-dimensional problems, a cali- are not only to facilitate the increase in brated response would be spatially density, but also to alleviate the problem blind policies. In areas experiencing of distance caused by growing conges- incipient urbanization, for example, the tion. The response includes improve- policy objective should be to facilitate ments in institutions to facilitate rising rising density, and policy makers should density as just outlined--and invest- pay special attention to institutions to ments in infrastructure to address the improve the functioning of (rural and growing problem of economic distance. urban) land markets. · For three-dimensional problems, the ·For two-dimensional problems, the response should include spatially blind, response should include both spatially connective, and targeted policies. In blind and connective policies. For exam- highly urbanized areas of a country, ple, in areas of a country undergoing for example, the problems of density rapid urbanization, the policy problems and distance are compounded by divi- 42 WORLD DEVELOPMENT REPORT 2009 sions within urban areas, most notice- economic concentration, chapter 8 pays ably between formally settled parts of a special attention to labor mobility, for metropolis and slums, where land mar- which the potential for market malfunc- kets use informal conventions. An effec- tioning is greatest. tive policy response includes institutions, · At the international spatial scale, the infrastructure, and interventions. policy objective should be to promote convergence in living standards in a At the national level, a similarly gradu- world in which divisions hamper the ated policy response can help to integrate movements of labor and capital. Dis- lagging and leading areas (chapter 8), and cussing how developing countries can at the international level, it can help to inte- gain access to world markets, chapter grate poor countries with world markets 9 emphasizes specialization and intra- (chapter 9). industry trade, in addition to exploiting At all three geographic scales, policy comparative advantage based on natural debates have one thing in common: cur- endowments. It pays attention to trade rently, they begin and end with discus- in intermediate goods, which is especially sions of spatially targeted interventions. sensitive to transport costs. This Report calls for a rebalancing of these debates to include all the elements of a suc- The Report draws on both experience cessful approach to spatial integration--in- and analysis to discipline the inquiry in a stitutions, infrastructure, and incentives. policy area as broad and difficult as devel- This Report takes a long-term per- opment itself, and it should be useful for a spective, chronicling spatial disparities in wide readership. But the Report is struc- today's developed economies when they tured to be friendly to readers interested were at incomes comparable to those of only in specific aspects of this inquiry: today's low- and middle-income countries. The Report has descriptive, analytical, It also systematically documents the rela- · and prescriptive parts and progresses tionship between spatial disparities and gradually from the positive to normative. development for a large set of countries. In Each part is a section of an integrated its conclusions, it makes a sharper distinc- inquiry, but each can be read separately. tion between spatial disparities in economic Policy makers pressed for time can read production and those in welfare. And it just the overview and the three policy recommends using agglomeration rents in chapters in part three. Students interested leading areas to push up social welfare in in the world's spatial transformation can lagging areas--and not, except in special read just the three chapters of part one, circumstances, to push economic produc- which provides a three-dimensional tour tion out to those places. of economic development. · At the local spatial scale, the policy objec- · The Report progressively widens the spa- tive should be to improve the quality of tial scale for addressing the policy ques- urbanization to maximize its growth tions posed by economic geography, from effects. Chapter 7 discusses how the pri- local to national to international, with orities of policy makers should change the specialized reader in mind. Readers as urbanization advances. It pays special interested in just the policy debate on attention to land use, where the potential urbanization in developing countries can for market malfunctioning is greatest. read just the three density cluster chap- · At the national spatial scale, the policy ters--1, 4, and 7. Those who are mostly objective should be to improve the mar- interested in the policy discussion on ket access of workers and entrepreneurs, territorial development and geographic especially in a world in which dimin- disparities within countries can read ished distance has changed the notion chapters 2, 5, and 8--the distance clus- of markets from local to global. Discuss- ter. Readers interested in regional inte- ing how policy makers can reconcile the gration can read just chapters 3, 6, and 9 political objective of national unity with in the division cluster. Navigating This Report 43 Figure 0.1 A navigational aid for the reader 1 DENSITY 2 DISTANCE GIM 1: 4 3 NORTHAMERICA AGGLOMERATION DIVISION 5 FACTS MIGRATION 7 6 GIM 2:WESTERN URBANIZATION SPECIALIZATION EUROPE 8 TERRITORIAL ANALYSIS DEVELOPMENT 9 GIM REGIONAL 3: LOCAL EAST INTEGRATION ASIA POLICIES TIONAL NA GIM 4: SUB-SAHARAN TIONAL AFRICA INTERNA Source: WDR 2009 team. · Chapters 1 through 9 slice the problem lenges posed by geography for develop- of economic development into digestible ment--and some clues to how geography bites, each serving a pedagogical func- was reshaped--can read these notes on tion. The arguments in the Report are different parts of the world. punctuated with four notes on "Geogra- phy in Motion," which connect the dif- Figure 0.1 shows how the Report can be ferent components by spotlighting the read horizontally (facts, analysis, and poli- experiences of North America, West- cies, respectively) or vertically, according to ern Europe, East Asia, and Sub-Saharan the policy interest of the reader. Africa. Readers interested in the chal- Geography in motion Overcoming Division in Western Europe Overcoming istance North America When Europeans began to colonize beyond their shores, the prospects for economic growth in North America seemed remote. Dur- ing the Seven Years' War (1756­63), as the French and British battled over Canada, Voltaire wondered why they should fight over "a few acres of snow." They should have been more interested in the economic potential of the Caribbean, where climate and soil were good for growing sugarcane, and they were. Manhattan was famously traded away by the Dutch in exchange for land around Suriname. But over time, it has been the few acres of snow and the rocky landscape of Plymouth (Massachusetts) that gave birth to the "reversal of fortune" between frigid northeastern America and the warmer south.1 T o understand how this reversal southwest, ending up in Missouri in were paramount. The Constitution happened,onehastounderstand 2000. By this time, America's popula- and the Northwest Ordinance (1787) how North Americans managed tion had settled mostly on its two coasts. provided the procedural mechanisms the growing density, the vast distances Americans are as physically distant as for transforming unsettled areas into in the continent, and the sharp divi- they have ever been. states. Public land was disposed of sions between slaves and their own- How did America overcome these through sales to private individuals ers, between natives and colonialists, vast physical distances? Initially, and outright grants. Eminent domain between French and English--in short, institutional mechanisms to allo- was used to put land to its best use, how North America's economic geog- cate land and secure property rights especially when required for railroads. raphy has been reshaped. Size and American economic Map G1.1 The U.S. geographic center of population gravity moved 1,371 kilometers between 1790 and 2000 ascendancy Size is the most obvious feature of the United States' economic geography.2 In 1800 5.3 million individuals lived on the 865,000 squares miles of land given to the fledgling nation under the Treaty of Versailles (1783). By 1900 a little more than 2 million square miles had been added through outright pur- 1790 chase, spoils of war, or treaty. Today 1900 1850 1950 the United States has more than 300 million people and a territory of 3.5 2000 million square miles. Since 1790 the population density of the country has multiplied nearly 18 times. The challenges of distributing pop- ulation and production over such a vast space are enormous. Both people and productive land have moved west and south. In 1800 the population was UNITED STATES centered in Maryland, on the eastern seaboard (see map G1.1). By 1900 the center had moved to Indiana. Over the twentieth century, the center veered Source: Geography Division, U.S. Census Bureau. Geography in Motion 45 The first transcontinental railroad was Convergence in living standards the dominant pattern of movement completed in 1864. Indigenous popu- The American Civil War had long- was from poorer to richer states. Prob- lations were removed forcibly, where lasting economic effects that divided the ably the most important example is the necessary, by the U.S. Army. States and country. Per capita incomes fell sharply migration of African Americans from local governments encouraged Ameri- in the South after the Civil War, both the rural South to the urban North cans to move by offering land, build- absolutely and relative to the rest of the (and West), beginning in earnest dur- ing canals, and supplying schools, country. In 1900 per capita income in ing World War I and becoming a tidal roads, and other public goods. These Alabama was still half of the national wave during and just after World War local governments competed with each average. In 1938 Franklin Roosevelt II. States such as Mississippi and Loui- other to attract people and firms, offer- famously remarked that the South was siana now rank lowest in disposable ing tax and other incentives. the nation's "number one economic income, but it is easy to imagine that People and firms were also encour- problem."America had its lagging areas. they would have been much worse off aged to move by the commerce clause But the twentieth century experience without this migration. of the U.S. Constitution, which explic- was one of steady convergence of living Convergence has been aided by itly prohibits state governments from standards. reductions in transport costs. Many of engaging in restraint of trade across In the United States, a clear negative the most important inventions in trans- state boundaries. The institutional relationship exists between the level of port and communications happened in structure thus permitted the free move- per capita income in a state in 1900 and the United States. In the twentieth cen- ment of people (except slaves), capital, the income growth in that state over tury, the network expanded with the and goods, with attendant property the next century. That is, poorer states diffusion of the airplane, the automo- rights so that movement could occur grew faster than richer states between bile, and electronic communications. without economic loss. 1900 and 2000, a phenomenon known Today, 16 of the 30 busiest airports in In this policy environment, the as"beta-convergence."The main expla- the world are in the United States, and "transport revolution"of the nineteenth nation for this phenomenon is migra- there are more than 75 automobiles for century and growing density permit- tion of people. In the twentieth century, every 100 Americans. ted a fundamental change in American economic structures. The combination Map G1.2 America's large cities are in the Northeast and on the two coasts of rail, canals, and steamboats vastly reduced the costs of medium- and long- haul transport compared with wagon transport alone.3 The country became Seattle more urban and dense, while regional economic structures diverged. New Boston Minneapolis England, which had been 80 percent Detroit agricultural in 1800 despite its poor soils New York Chicago and climate,started to develop manufac- Philadelphia Washington, D.C. tures, while the Midwest specialized in San Francisco food. By the beginning of the twentieth Los Angeles century, the United States had become Riverside the largest manufacturer in the world. Phoenix Atlanta Dallas The growing density and the migra- San Diego tion of people and firms were driven largely by market forces. Most settle- Tampa Houston ment was cautious. Railroads were built when (and where) investors thought Miami they could make a profit and moved incrementally across the country. Occasionally settlement did "leapfrog," jumping over large expanses of land to Population, 2007 (millions) get someplace else, as in California after 18 12 the discovery of gold in 1849. But that 6 4 simply accelerated the pace of realloca- 2.5 tion of labor in America. Source: Population of Metropolitan Statistical Areas; U.S. Census Bureau. 46 WORLD DEVELOPMENT REPORT 2009 The Report at a glance The invention and diffusion of the ies, in the Northeast and on the coasts, across states; over a decade, more than automobile led to the enlargement of producing what is known as "sigma- a quarter of the population changes cities through a pronounced "flatten- convergence," a reduction in the income its state of residence. By overcoming ing" of urban density as one moves inequality across states (see map G1.2). distance and division, and by permit- from the center city to the suburbs. This By one measure, the dispersion across ting population and production to helped magnify agglomeration econo- Density, Distance and Division states in per capita income had fallen to be uneven across space through free mies, but it also produced social divi- one-third its 1880 level by 2000. mobility, per capita incomes in the sions. The U.S. system of local public United States today are both high and finance, relying on local property taxes Rising density,falling remarkably similar across the different to fund services, is poorly designed to disparities,persisting divisions states. effect income redistribution. Rich and The long-run economic performance A remaining challenge for the United middle-class households can avoid sub- of the United States is exemplary. Per States is the removal of divisions. The sidizing others by moving to new sub- capita income growth has averaged North American Free Trade Agree- urbs. Race also plays a role--the central 1.8 percent per year for the last 180 ment (NAFTA) is a step in this direc- city is predominantly "black" whereas years, leading to a cumulative 26-fold tion. But it is a modest step. Consider the suburbs are "white." improvement in living standards. Canadian-U.S. market integration. One For better or for worse, growth in Alongside this growth, income inequal- study found that trade among Cana- automobiles benefited from the Federal ity across states has fallen. America has dian provinces was much larger than Highway Act of 1956, which authorized realized economies of scale--first at between Canada and the United States, building of the Eisenhower Interstate the plant level, then at the local level controlling for distance and the eco- System of highways.In a famous speech, as towns specialized in manufacturing, nomic size (gross domestic products) President Eisenhower recounted how and later at the metropolis level in the of the trading partners, in this case, as a young officer he participated in major urban agglomerations in places states and provinces.4 Given California's the first transcontinental motor convoy like Los Angeles and New York. size, for example, its trade with Ontario from Washington, D.C., to San Fran- The United States today is composed should have been 10 times Ontario's cisco in 1919. The trip took 62 days, of a highly effective set of national mar- trade with British Columbia, Califor- encountering every type of delay imag- kets in goods and factors of produc- nia's closest Canadian neighbor. In fact, inable along the way. Today, courtesy tion. Place still matters in determining Ontario's trade with British Columbia of the system, a driver can cover the income, but it matters in the short was three times its trade with Califor- 2,819-mile journey in two days. Recent run, not the long, and the short run nia. Even one of the thinnest borders in research shows that the 47,000-mile is much shorter than it was a century the world has a large negative influence network of highways has integrated ago. Major local shocks like Hurricane on trade. formerly isolated rural areas into the Katrina have far less impact on local Along its northern boundary, the national economy and fostered metro- growth prospects than before. After the United States and Canada share 3,987 politan growth. Mariel boatlift brought 125,000 Cuban miles, the longest unguarded interna- What have these connections done to refugees to Miami in the early 1980s, tional border in the world.The situation the distribution of population and eco- regional wages did not experience a is markedly different along the south- nomic activity? Paradoxically,as the cen- perceptible impact. ern border with Mexico. The border is ter of gravity moved toward the interior The result is a seeming paradox: guarded--not closely enough for many of North America, the interior--except wages in America (corrected for human U.S. citizens--to keep potential illegal for its metropolises--has hollowed out. capital) are similar in different loca- immigrants from entering. There are Missouri has just 5.5 million people, tions, while economic activity is highly even proposals to build a fence stretch- more than half of them in the greater St. unequal across space. Europe is lauded ing across the 1,933 mile border. Such Louis area. Spreading out the transport for having lower social inequality, but barriers are an obstacle to convergence infrastructure has not spread people out, North America is more spatially equal. between countries in the North Ameri- but it has allowed growth from agglom- And it has a more spatially efficient can continent. eration economies to occur in more cit- distribution of economic production. ies across the country. The distribution The reason: a mobile labor force. Every of population in 2000 is clustered in cit- year about 8 million Americans move Contributed by Robert A. Margo. Part One Seeing Development in 3-D As the world's economy grows, people and production are concentrating, pulled as if by gravity to prosperous places--growing cities, leading areas, and connected countries. As it did decades ago in today's high-income countries, the drive to density in low- and middle- income countries can increase the sense of deprivation as the economic distance between prosperous areas and those left behind widens. And although rapid advances in transport and communication increasingly bind together geographically distant communities around the world and open new opportunities for exchange, political divisions that obstruct the ow of people, capital, and goods remain. Part one of this Report de nes the spatial dimensions--density, distance, and division--and describes their evolution with economic development. Chapters 1, 2, and 3 show how the economic geography at the local, national, and international scales is changing, and how the scope and pace of these changes compare with transformations in the economic geography of North America, Europe, and Japan when they were at similar stages of development. This broad sweep of stylized facts informs the analysis in part two and the policy discussions in part three of the Report. CHAPTER 1 Density M ostly off the world's radar, on Despite its industriousness, Bamako is a dusty plain in West Africa, one of the sleepier cities in West Africa. is a city of 1.6 million people. Many of the manufactured staples come Bisected by the River Niger, its two halves-- 1,184 kilometers by road from one of the with about 800,000 people each--are linked region's metropolises, Abidjan, which has by only two bridges. The pressure of move- more than twice Bamako's population. ment is so strong that every morning one of Abidjan seems small beside Lagos, where these bridges is dedicated to incoming traf- activity is so concentrated that its residents fic: minibuses, bicycles, motorbikes, pedes- speak of living in a pressure-cooker. Some trians, and occasionally private cars. In the families rent rooms to sleep for six hours evenings, to leave the center means joining and then turn them over to another fam- an exodus of people toward the minibus ily that takes their place. Shopping does depots. Green vans loaded with passengers not necessarily require travel: goods are file out to residential neighborhoods as far brought on foot and cart to drivers stuck in as 20 kilometers away. This is Bamako, Mali. Lagos's interminable traffic jams. To some, It contracts into its center every morning like the authors of Lagos's 1980 master plan and breathes out again in the evening. written when the city had just 2.5 million With each breath Bamako grows bigger. residents, the continuing growth of the city It happens to be one of the fastest-growing is "undisciplined."2 What can possibly be so cities in the world. Natural demographic attractive about living in Lagos that, despite growth is supplemented by migration from its congestion and crime, it continues to the countryside and other Malian cities. Its draw migrants? population in 2008 is 50 percent larger than The short answer: economic density. 10 years ago, making it the same size as Lagos is not the most economically dense Budapest, Dubai, or Warsaw. It has 10 times city in the world, nor even the most densely more inhabitants than the next biggest populated. Those distinctions belong to Malian city and accommodates 70 percent Central London and Mumbai, respec- of the country's industrial establishments.1 tively. Even so, Nigeria's economic future New neighborhoods--quartiers--formerly and Lagos's growth are as inextricably villages, become consolidated with the rest tied as Britain's economy is with London's of the city, toward the south, east, and west. growth. No country has developed with- Some of Bamako's people are now moving out the growth of its cities. As countries out into surrounding neighborhoods in become richer, economic activity becomes search of cheaper land and some tranquil- more densely packed into towns, cities, ity, but they remain within reach of the city and metropolises. This geographic trans- because it provides their livelihoods. formation of economies seems so natural 48 Density 49 that--at an impersonal aggregate level--it nomic density continues to increase in is taken for granted. But moving to eco- a postindustrial economy because ser- nomic density is a pathway out of poverty vices are even more densely packed than both for those who travel on it and, ulti- industry. mately, for those left behind. Jane Jacobs, · Rural-urban and within-urban dispar- the noted urbanist, did not have Bamako ities in welfare narrow with develop- and Lagos in mind when she wrote, "A met- ment. In the early stages of development, ropolitan economy, if it's working well, is geographic disparities in welfare are constantly transforming many poor people large. With development, these gaps into middle-class people, many illiterates may increase initially. Rural-urban gaps into skilled people, many greenhorns into in income, poverty, and living standards competent citizens. Cities don't lure the begin to converge as economies grow, middle class. They create it."3 She might faster for access to social services, and as well have written: as Lagos and Bamako faster in areas of more vibrant growth. grow, they will fill in West Africa's missing Within-city gaps in welfare and hous- middle. ing--most obvious in informal settle- This chapter introduces density, the first ments or slums--persist for much of the geographic dimensions of develop- longer, and narrow only at later stages of ment, defined as the economic mass or out- development. put generated on a unit of land. Surveying · Neither the pace of urbanization nor the evolution of density with development, its association with economic growth the chapter presents stylized facts about is unprecedented. Today's developing how density in a country rises with urban- countries are sailing in waters charted by ization, rapidly at first, and then more developed nations, which experienced a slowly. These changes are associated ini- similar rush to towns and cities. The tially with a divergence of living standards speed is similar, and the routes are the between places with economic density and same. What is different today is the size those without, later with a convergence. of the ship: the absolute numbers of peo- Living standards thus eventually converge ple being added every year to the urban between areas of different density, such as populations of today's developing coun- urban and rural. Even within cities, densely tries are much larger than for even the populated slums amid formal settlements, most recent industrializers such as the the differences slowly disappear with devel- Republic of Korea and Taiwan, China. opment. But this convergence does not hap- Later chapters of this report investigate pen by itself. It requires the institutions to the policy implications of these similari- manage land markets, investments in infra- ties and differences. structure, and well-timed and executed interventions. Defining density The main findings: Density refers to the economic mass per unit · The concentration of economic activ- of land area, or the geographic compactness ity rises with development. The world's of economic activity. It is shorthand for the densest areas or settlements are in devel- level of output produced--and thus the oped countries. But the path to these lev- income generated--per unit of land area. It els, "urbanization" in this Report, is not can, for example, be measured as the value linear. The share of a country's popula- added or gross domestic product (GDP) tion settled in towns and cities rises rap- generated per square kilometer of land. idly during its transformation from an Given that high density requires the geo- agrarian to an industrial economy, which graphic concentration of labor and capital, generally coincides with its development it is highly correlated with both employment from low to middle income. The pace of and population density. Density is the defin- urbanization slows after that, but eco- ing characteristic of urban settlements. 50 WORLD DEVELOPMENT REPORT 2009 The economic world is not flat This density contrasts markedly with The geographic distribution of economic ion the agricultural areas of Belgium. In the activity, at any resolution, is uneven. No Flemish Flanders (Vlaams Gewest) area, matter the geographic scale examined, be ed, 6,323 square kilometers of land are used for it the country or a subnational area such as agriculture. Its area is almost 40 times that a province or district, there is a hierarchy of Brussels, but its employment is just 13 ju of density. At the top is the primary city, percent of Brussels and its GDP a mere 4.5 and at the bottom are agricultural lands or at percent, translating into employment and rural areas. Between them is a continuum GDP densities of only seven workers and wo of settlements of varying density. 330,000 per square kilometer. The ratio of The geographic unevenness of economic output density between Brussels and Flan- mass, or bumpiness, tends to increase ders is 1,000 to 1. In between metropolitan with a country's land area. But even the Brussels and rural Flanders is a range of set- economic geography of small countries is aph tlements, each with a different density (see bumpy. The Belgian city of Brussels has map 1.1). The cities of Antwerp, Brugge, a land area of 161 square kilometers, of Gent, and Leuven have an average output which 159 square kilometers are used for of 22 million and employment density of nonagricultural purposes. On this small 342 workers per square kilometer.5 area, a GDP of 55 billion is generated by In both developed and developing lop lop about 350,000 workers--that is, the aver- wo countries, then, the economic landscape is age square kilometer of land has more than bumpy. But the topography does not corre- top aph 2,000 workers annually producing almost spond to a simple urban-rural dichotomy. 350 million of services and goods. Brus- go A continuum of density gives rise to a port- sels not only has high densities of GDP and folio of places. At the head is a country's io employment; it also has the highest popu- leading, primary, or largest city. Below the lation density of any European (EU27) of rop primary city is a spectrum of settlements-- area classified as NUTS1 (Nomenclature secondary cities, small urban centers, of Territorial Units for Statistics)--more towns, and villages (see figure 1.1). In some es see than 6,000 people per square kilometer, countries, such as France and Mexico, the 18 times the average for Belgium.4 For the size difference between the top two cities is sake of comparison, the population den- phenomenal.l. With a population of 10 mil- sity of London and Madrid is about 5,000 lion, Paris dwarfs second-ranked Marseilles is dwarfs second-ranked Marsei le people per square kilometer. with just 1.5 million. And with a population Map 1.1 The landscape of economic mass is bumpy, even in a small country like Belgium Brussels BELGIUM Antwerp FLANDERS Gent WALLONIA Leuven Brugge Source: WDR 2009 team and World Bank Development Research Group, based on subnational GDP estimates for 2005. See also Nordhaus 2006. Density 51 of 22 million, Mexico City is more than four Figure 1.1 From dichotomy to continuum: a portfolio of places times as populous as Guadalajara, Mexico's The simplified area economy and a more realistic representation second city. Conversely, in India and the United States, the size difference between Urban the two biggest cities is relatively small. With Metropolis populations of more than 22 million people, Rural Urban Rural Towns Large city Mumbai and New Delhi stand shoulder to Secondary cities Villages shoulder. New York has a population of 22 million, Los Angeles 18 million.6, 7 An evolving portfolio of places Source: WDR 2009 team. Although the growth of cities appears chaotic, the underlying patterns have a primacy" notwithstanding, the "portfolio of remarkable order (see figure 1.2). A coun- places" is an enduring feature of economic try's urban hierarchy is characterized by development. two robust regularities: Settlements of different sizes complement · The"rank-size rule"--the rank of a city in one another. Metropolises, secondary cit- the hierarchy and its population are lin- ies, market towns, and villages are all linked early related. through their complementary functions (see box 1.2). The primary city is often but not · Gibrat's law--a city's rate of population always the national administrative center and growth tends to be independent of its theseatofpoliticalpower:Cambodia'sPhnom size. Penh, Cameroon's Yaounde, and Colombia's According to a special case of the rank- Bogotá. A country's leading city also tends to size rule, known as Zipf's law, the popula- beitsmostdiversified,bothintheprovisionof tion of any city is equal to the population of goods and services and in cultural and other the largest city, divided by the rank of the amenities. For the cultural amenities, think of city in question within the country's urban Broadway in New York City, the Opera House hierarchy (see box 1.1).8 As early as 1682, in Sydney, and the Louvre in Paris. But think Alexandre Le Maître observed a systematic also of Trinidad and Tobago's Port of Spain, pattern in the size of cities in France.9 For famous for the annual carnival that attracts all classes of country, the relative size dis- large numbers of visitors. tribution has remained stable over time, Just as a primary city forms the core of even as incomes and populations grew a country's metropolitan area with other (see figure 1.2). Concerns about "urban adjacent cities, other large urban centers or Figure 1.2 Almost a law: relative size distributions of settlements remain stable over time Low-income countries Middle-income countries High-income countries Log of rank Log of rank Log of rank 6 6 6 5 5 5 4 4 4 3 3 3 2 2 2 1 1 1 1950 1980 2005 1950 1980 2005 1950 1980 2005 0 0 0 2 4 6 8 10 2 4 6 8 10 2 4 6 8 10 Log of population Log of population Log of population Source: United Nations 2006c. Note: Each data point represents an agglomeration area of population size of 750,000 or more. 52 WORLD DEVELOPMENT REPORT 2009 BOX 1.1 Two laws and a rule: the empirical regularities of a country's city-size distribution The rank-size rule, discovered in 1913, can U.S. city sizes is more even--and that the path can have important long-term be expressed as the rank r associated with rule fails to hold at the extremes of the U.S. repercussions for the welfare of a city's a city of size S is proportional to S to some city-size distribution, a common nding inhabitants. On whether the power in the negative power. The special case in which for many countries.b Moreover, the rank- rank-size rule equals ­1, so that Zipf's law the estimated power equals ­1 is known as size rule also holds for countries as diverse holds, many researchers seem to agree Zipf's law, named after a linguist, George as Kazakhstan and Morocco, providing that, in general, it does not. Zipf. Evidence on the pervasiveness of the further evidence of its universality (see the The robust message from the rank-size rank-size rule comes not only from large gure below). rule is that, for a given country or area, a cities belonging to countries of di erent Whether the rank-size rule is really a wide range of city sizes coexists. Even the income classes, but also from the experi- rule with underlying theoretical structure most developed countries have a portfolio ence of individual countries. The remark- is still under debate. It can be shown to of settlements of di erent sizes, ranging able westward and southward expansion follow from Gibrat's law, which implies from the small to the large, as opposed to of the U.S. urban hierarchy notwithstand- that cities grow in parallel.c This is consis- a single megacity or a collection of cities, ing, the rule provides a good description tent with the absence of any systematic all of similar size. Agglomeration is a bal- of the size distribution of U.S. cities for growth di erences between cities. But ancing act between centripetal and cen- every decade between 1790 and 1950.a this does not imply that policy is inca- trifugal forces. The balancing point di ers Indeed, even today, the rank-size rule con- pable of in uencing a city's size and depending on the sector, the economic tinues to describe well the size distribution economic performance. Cities can and activities, and the type of industries. of U.S. cities (see gure below). This is so do move up and down their national Contributed by Mark Roberts. despite evidence that the shape of the rule urban hierarchies as a result of good a. Madden 1956, cited in Kim and Margo 2004. has changed over time, becoming slightly and bad policy choices. And even transi- b. Gabaix and Ioannides 2004, p. 14. atter so that the overall distribution of tory departures from a parallel growth c. Gabaix and Ioannides 2004, pp. 16­17. The rank-size rule, for nations as diverse as the United States, Morocco, and Kazakhstan United States 2000 Morocco 1993 Kazakhstan 1993 Log of rank Log of rank Log of rank 6 3.0 4.0 4 3.0 2.0 2.0 2 1.0 1.0 0 0.0 0.0 13 15 17 11.5 12.5 13.5 14.5 15.5 11.5 12.5 13.5 14.5 Log of population Log of population Log of population Sources: The graph for the United States is from Rose (2005); the graphs for Kazakhstan and Morocco are based on data for cities and urban agglomerations from Brak- man, Garretson, and Marrewijk (2001). secondary cities act as regional foci for both private medical colleges, is a seat of learning the economy and society. For example, they in southern India. are the local centers for the financial sector, These large regional cities are connected which serve the areas around them. Düs- to smaller cities or major towns. The Ruhr seldorf, Hamburg, Hanover, and Munich area of Germany, the Randstadt area of the are all home to regional stock exchanges, as Netherlands, and the Padang-Medan hub in well as local concentrations of venture capi- Indonesia's Sumatra represent alliances of cit- tal firms.10 Dallas and Atlanta emerged as ies. Smaller cities within these areas consti- regional centers of commerce and finance tute more specialized urban centers, typically in the lower South of the United States, focusing on manufacturing and the produc- and both host regional offices of the Fed- tion of traditional and standardized items. eral Reserve Bank.11 Large urban centers Symbiosis is the ruling order: just as the larger and secondary cities also act as local politi- cities help to serve the smaller cities, so the cal centers, and provide advanced public reverse is true. For instance, the larger cities health, education, and cultural facilities. depend on the smaller ones for the daily pro- Hyderabad, the state capital of Andhra vision of workers through commuting.12 Pradesh, with numerous universities, lead- Just as there are mutually beneficial links ing institutes for technical education, and between larger and smaller cities, the same is Density 53 BOX 1.2 The Republic of Korea's portfolio of places Illustrating a well-developed portfolio manufacturing, especially standardized At the bottom of the hierarchy, of places are seven settlements in the manufacturing, than cities farther up the Jeongeup and Sunchang, both in the Republic of Korea's urban hierarchy: hierarchy. Although both cities serve as Jeonbuk province, are close to the inter- Seoul, Pusan, Daegu, Ansan, Gumi, Jeon- manufacturing centers, they di er in their face between rural and urban. So while geup, and Sunchang. specializations. Gumi is heavily specialized Jeongeup has a relatively large popula- Seoul is at the pinnacle of the hierarchy. in the radio, television, and communica- tion (129,050), one in four of its inhabit- Located 50 kilometers from the Republic tion equipment industry, which by itself ants is a farmer. Likewise, Sunchang is a of Korea's border with the Democratic accounts for more than 50 percent of local rural town: half of the 32,012 residents are Republic of Korea in the Han River basin, manufacturing employment. Ansan is farmers. To the extent that they exhibit it is the country's capital and home to specialized in such high-tech industries as any specialization in manufacturing, it a quarter of its population (that is, 9.76 electrical machinery and computers and is either in traditional resource-related million people). It serves as the nation's o ce machinery. It also has agglomera- industries, as in Jeongeup, or in the man- political center and cultural heart. Also tions in several heavy industries: almost ufacture of food and beverage products, typical is its specialization in business 14,000 workers, or 14.7 percent of the local as in Sunchang. services, nance, insurance, real estate, manufacturing workforce, are employed in and wholesaling and retailing. Overall, the fabricated metal products industry. Contributed by Park Sam Ock. services account for 60 percent of the local economy. Seoul is also highly specialized Seoul heads the hierarchary of settlements in the Republic of Korea in publishing and printing and in fashion design and high-end apparel, with the two industries employing more than half the city's 465,000 manufacturing workforce. Next in the urban hierarchy are Pusan and Daegu. With a population of 3.7 Seoul million, Pusan is the Republic of Korea's REPUBLIC OF KOREA second largest city. In the southeastern Ansan corner of the Korean Peninsula, its sea- port, one of the world's largest, handles more than 6.5 million container ships a year. Daegu is a metropolitan area of 2.5 million, dominated by textile and cloth- Gumi ing manufacturing and automotive parts manufacturing and assembly. Since 1970, Daegu the Gyeongbu Expressway has connected Jeongeup Pusan to Seoul through Daegu. About 20 Sunchang ights operate daily between Seoul and Pusan Daegu, and since 2001, the two cities have been linked by a high-speed train. Population, 2007 Much farther down the hierarchy, Ansan (thousands) and Gumi are secondary cities, with popu- > 4,000 1,000­4,000 lations of around 679,000 and 375,000, 500­1,000 150­500 respectively. In Gyunngi province, Ansan < 150 belongs to the Seoul National Capital Area, as part of Seoul's suburban area. Gumi is in Gyungbok province, in the southeast. As tends to be the case with secondary cities, Ansan and Gumi are more specialized in Sources: WDR 2009 team, using data from the National Statistical Office of the Republic of Korea. true for smaller cities and towns, and towns scale in postsecondary education and health and rural areas. Towns are the connective tis- care services. Symbiosis is again the rule. sue between rural and urban areas. They act Towns draw sustenance from the agricultural as market centers for agricultural and rural activity of rural areas, but their prosperity output, as stimulators of rural nonfarm activ- also spills over to villages by providing non- ity, as places for seasonal job opportunities for farm employment opportunities. Farmers in farmers, and as facilitators of economies of Vietnam migrate seasonally to work in urban 54 WORLD DEVELOPMENT REPORT 2009 construction, returning to invest the money interchangeably, agglomeration, density, earned in their farms.13 Farmers in Makueni, or geographic concentration of economic Kenya, use nonfarm income to invest in ter- activity--across countries. racing, planting trees, clearing bush, building The index identifies an area of 1 square houses, and educating their children. Farm- kilometer as urban, agglomerated, or dense ers in the semiarid Diourbel region of Senegal if it satisfies the following three conditions: have responded to growing urban demand for · Its population density exceeds a thresh- meat by diversifying away from groundnut production into animal husbandry.14 old (150 persons per square kilometer). · It has access to a sizable settlement within some reasonable travel time (60 minutes Measuring density by road). Measures of gross product at a refined spa- · The settlement it has access to is large tial scale, such as a district or a city, are in that it meets a population threshold difficult to come by. Even for developed (more than 50,000 inhabitants). countries, output estimates tend to be available only for rather broadly defined Box 1.3 summarizes the rationale and subnational areas (first level and adminis- methodology underpinning the index. trative units, such as provinces or states). At One advantage of the agglomeration this level, important variations in economic index is that it incorporates both density density are likely to average out. Fortunately and the local distance to density. Based on though, as illustrated earlier for Belgium, the criteria of population density and acces- output and population density are closely sibility to a sizable market, the index also correlated. Reliable population estimates comes closer to providing an economic defi- are more easily available, even for villages nition of an area that can both benefit from or townships, because in most countries, a and contribute to agglomeration economies. population census is taken every decade. Although economic density is both a cause The strong correlation between popula- and a consequence of agglomeration econo- tion density and economic mass is consistent mies, accessibility to this economic mass with urban areas being a conglomeration of from the outer parts of the city facilitates the consumers and producers, of buyers and sell- exploitation of such benefits to proximity. ers, and of firms and workers. For a typical This is especially true in the service sector metropolitan area, the gradient of popula- in which face-to-face interactions are often tion density for distance from the city center necessary. By reducing the need to allocate is similar to the corresponding gradient for valuable land area to residential uses in and employment density.15 As implied above, the near urban centers, transport infrastructure extent to which a country's population lives facilitates economic density. in urban areas bears a strong relationship to Going to work by car or by high-speed how "bumpy" its economic geography is. public transportation is a luxury that devel- Density goes from smoothly spread out to oped country commuters do not always quite uneven as a country develops. Urban- share with their counterparts in developing ization is thus synonymous with a tendency countries. For any given geographic dis- toward greater agglomeration within a coun- tance, therefore, accessibility to a city tends try. A country's urban share is a good proxy to be lower in developing countries because for the proportion of its population living in of the need to rely on alternative, more time- areas of high density and, therefore, for the intensive modes of transportation, such as "bumpiness" in its economic geography. walking, cycling, or inefficient public trans- This Report proposes the use of an portation operating on poor-quality roads. agglomeration index computed using geo- In Mumbai, India, 44 percent of people walk graphic information systems as a measure to work,16 and in Hefei City, China, more of density. Measures of urbanization are than 70 percent either walk or cycle.17 nonuniform across countries, which makes Such variations in accessibility deter- comparability and aggregation a challenge. mine both the shape and form of a city. The index allows for a more consistent com- When most people walk to work, a city is parison of the level of urbanization--or, more likely to be monocentric and densely Density 55 BOX 1.3 Computing the agglomeration index The United Nations maintains the World adjusted for purchasing power di er- settlement center is calculated based Urbanization Prospects database, a trea- ences between countries--is that it on the maximum travel time to the sure trove of information. It provides allows international comparisons and center. urban shares and population data for calculations that aggregate poverty for · Create population density grids. These 229 countries stretching back to 1950. regions and the world. The agglomera- are created at a 1-kilometer spatial But these data are based on country tion index allows the same comparisons resolution using two global grid-based de nitions, which can be quite di erent. and aggregation. population data sources, GRUMP and This Report proposes a new measure of The methodology underlying the cal- LandScan.b agglomeration, based on a uniform de - culation of the agglomeration index can · Identifytheareas. Identify the grid cells nition of what constitutes an "urban" or be summarized as follows: that satisfy thresholds for all three criteria. agglomerated area, using the technique · Specify thresholds. To be classi ed as · Aggregate grid cell populations. The outlined in Chomitz and others (2007) and "urban" using the agglomeration index, result is analogous to urban popula- elaborated in Uchida and Nelson (2008). an area must satisfy three criteria based tion. The proportion of this number to This should not be read as implying on (1) minimum population size used that country's total population is the that World Urbanization Prospects data to de ne a sizable settlement, (2) mini- agglomeration index, a summary mea- are awed. A better interpretation is to mum population density, and (3) maxi- sure of the proportion of the popula- see the challenge of measuring urbaniza- mum travel time, by road, to the sizable tion living in areas of high density. tion as analogous to the measurement settlement. of poverty. Each country has its own In calculating the index, this Report uses poverty line and criteria to track changes · Locate the centers of sizable settlements. a base case set of thresholds of 50,000 for in national poverty rates. But these mea- This mapping is done for cities that minimum population size of a settlement, sures do not allow reliable comparisons meet the minimum population size 150 people per square kilometer for pop- of poverty between countries, and they criterion using data from the Global ulation density, and 60 minutes for travel cannot be used to aggregate poverty Rural-Urban Mapping Project (GRUMP) time to the nearest large city. for groups of countries. The merit of human settlements database.a The density and travel time thresh- a uniform poverty measure--such as · Determine the sizable settlement's bor- olds are those employed in Chomitz, those living below US$1 or US$2 a day, der. The border surrounding a sizable Buys, and Thomas (2005). The density threshold is the same as the one used by the Organisation for Economic Co- The internationally comparable agglomeration index can yield different urban shares than those operation and Development (OECD). The from country-specific definitions threshold of 50,000 for a sizable settle- Urban share (%) ment is reasonable for developing and 100 developed countries. Many developing Country definition 90 nations have more than 10 percent of their total population in urban centers 80 Agglomeration Index of between 50,000 and 200,000. Some 70 examples include Chile in 2002, Brazil 60 in 2000, and Malaysia in 2000, all with 50 around 17 percent of their national 40 population living in urban centers of 30 50,000­200,000 inhabitants. Of India's 20 urban population in 2001, 20 percent 10 lived in settlements of this size. According to the World Urbanization 0 of Prospects database, the worldwide urban Brazil rkey Angola Lanka India China Uganda Vietnam Rep. Tunisia ation Tu share in 2000 was 47 percent. Using the Argentina Sri Bangladesh Feder Arab base case criteria, this ratio is 52 percent, ian but using 100,000 as the minimal settle- Egypt, Russ ment size, it is 44 percent, according to Sources: Chomitz, Buys, and Thomas 2005; Nelson 2008; Satterthwaite 2007; United Nations 2006c. the agglomeration index. But country a. The GRUMP human settlements database was developed by the Center for International Earth Science level estimates can be further apart (see Information Network (CIESIN) at Columbia University (http://sedac.ciesin. columbia.edu/gpw/index.jsp). b. LandScan was developed by Oak Ridge National Laboratory (http://www.ornl.gov/sci/landscan/). gure at left). populated at its core. In Mumbai, half of all agglomeration in industrial districts, work- workers commute less than 2 kilometers, ers in nineteenth-century Britain had to live implying that they live close to their places of nearby. The centers of industrial towns were work. Similarly, to obtain the advantages of densely populated, and overcrowded housing 56 WORLD DEVELOPMENT REPORT 2009 was common. Not until the electric tram was redefined as 15,000 (Nigeria and Syria, introduced did this change. for example, have cutoffs of 20,000), that In determining accessibility, and thus the share would drop to 67 percent. shape and form of cities, features of physical · Mauritius. In 2000 about a quarter of geography can also be important. Manhattan Mauritius's population lived in settle- Island in New York City is difficult to get to, ments with between 5,000 and 20,000 simply because of geography, so it has sky- inhabitants. Some of these settlements scrapers and a classic monocentric structure, are district capitals, but none of them with half its employment within a three-mile are classified as urban. If they were, the radius of Wall Street. By contrast, in Los urban share would have been more than Angeles, one has to widen the area to a radius two-thirds rather than less than half. of 11 miles from the center to find as large a share of employment.18 The implication: At a regional level, according to World economic density in New York City is $1.44 Urbanization Prospects data, South Asia billion of gross product per square kilometer, poses the paradox of being the least urban- in Los Angeles it is $0.49 billion.19 ized region (27 percent urban) in the world In the United Kingdom, Stevenage, Basil- while also the most densely populated. Using don, and Crawley are commuter towns that the agglomeration index, South Asia's urban serve London. About 11 percent of Lon- share in 2000 was 42 percent, making it more don's GDP is generated by commuters from urbanized than both Sub-Saharan Africa suburban areas.20 Similarly, in the United and East Asia and the Pacific (figure 1.3). States, a daily tide of workers commute into The World Urbanization Prospects also pose Washington, D.C., from the neighboring a puzzle for Latin America and the Carib- states of Maryland and Virginia. In 2005 the bean. The urban share in this region in 2000 net contribution of commuters from these was greater than that in Eastern Europe and two states to Washington, D.C.'s output Central Asia and almost on par with the was $36.4 billion. Maryland's Montgomery OECD's. The OECD has an average GDP per County--within easy commutable distance capita more than six times that of the aver- of the district--alone contributed $6.4 bil- age Latin American country. More reason- lion to Washington's gross product.21 ably, the agglomeration index indicates that The biggest advantage of the agglomera- Latin America and the Caribbean's urban tion index is its comparability across coun- share in 2000 was similar to that of Eastern tries. Here the index has an advantage over Europe and Central Asia, and 15 percentage the United Nations' World Urbanization points lower than that of the OECD. Prospects database, which contains the "de Despite these drawbacks, the World facto population living in areas classified Urbanization Prospects data are the only avail- as urban according to the criteria used by able information for comparisons over time. each area or country."22 The heterogeneity The agglomeration index is available only for across countries can makes cross-country 2000, because time-series data on road net- comparisons misleading. A few examples: works, necessary to estimate travel time, are not readily available. So, the agglomeration · India. With the criterion for an urban index and World Urbanization Prospects data- area used by Zambia or Saudi Arabia, base should be considered as complementary defined as settlements with populations data sources for examining urbanization of 5,000 or more, the share of India's and density, and this Report uses both the population in urban areas in 1991 would agglomeration index and the World Urbaniza- be 39 percent instead of the official figure tion Prospects data.23 Calculating comparable of 26 percent. This is because 113 mil- urban share measures for at least some coun- lion inhabitants of 13,376 villages would tries in the past is possible; going forward, it be reclassified as urban. should be a priority for all countries. · Mexico. Based on Mexico's official cri- terion of settlements of 2,500 or more Economic concentration-- as urban, the country's urban share in the richer, the denser 2000 was 74.4 percent. But if the settle- In the early stages of development, when an ment population threshold were to be economy is primarily agrarian, people live Density 57 spread out on farmland. Even the largest Figure 1.3 The agglomeration index helps to compare urbanization across regions towns and cities are small. Urban settlements Urban share (%) are likely to be small port cities and market 100 towns, serving the rural needs and trading 90 surpluses of agriculture. Industrialization 80 brings with it a rapid process of urbaniza- 70 tion--new cities are born, and existing cities Agglomeration Index 60 expand. As people crowd into these cities at 50 a faster rate than their boundaries expand, United Nations population and economic density increase. 40 Quite early in a country's development, this 30 leads to a hierarchy of places. 20 So, two transitions characterize eco- 10 nomic development. The first involves 0 Sub-Saharan South East Asia Middle East Latin Europe & Other OECD the movement from a primarily agrarian Africa Asia & Pacific & North America & Central high-income countries economy to a much more manufacturing- Africa Caribbean Asia economies oriented economy. The second transition, Sources: Chomitz, Buys, and Thomas 2005; Nelson 2008; Satterthwaite 2007; United Nations 2006c. taking place at a much higher level of devel- opment, involves the transformation to a for disproportionate shares of their national service-oriented economy. The first phase GDP. In 2005, Mexico City contributed 30 of urbanization, which occurs at a faster percent of Mexico's GDP despite occupying rate, coincides with the transition from only 0.1 percent of its land. Luanda contrib- a rural to an urban economy. The second uted a similar share of Angola's GDP, while phase of urbanization, at a slower rate and a occupying 0.2 percent of its land. Like- much higher level of development, is linked wise, the largest cities in Hungary, Kenya, to a within-urban evolution. In most coun- Morocco, Nigeria, and Saudi Arabia-- tries, these transformations happen at the Budapest, Nairobi, Casablanca, Lagos, and same time but in different areas. Riyadh--contributed about 20 percent of To measure concentration, we have to their country's total GDP while taking up define an area. The policy debate often less than 1 percent of land.25 involves a discussion of urban primacy, Density, defined as GDP in purchasing such as whether developing country cities power parities per square kilometer, rises are too big or too small. More academic with the level of development, and the dens- discussions use a purer geographic notion est places in the world are in the richest coun- of space. This chapter uses both spatial tries. Dublin, London, Paris, Singapore, and units--primary cities and the densest grid Vienna ranked at the top, in 2005, with more cell of 1° longitude by 1° latitude of a coun- than $200 million in gross product per square try--to measure concentration. kilometer. Likewise, Tokyo-Kanagawa, New York­New Jersey, Oslo­Akershus-Vestfold, Historically, rapidly rising and Vienna-Mödling were the densest grid concentration, then a leveling off cells of 1° longitude by 1° latitude, generating By one definition, a city is a geographic area more than $30 million of gross product per characterized by a concentration of eco- square kilometer (figure 1.4). nomic actors.24 Globally, the top 30 cities, A century of data on aggregate urban ranked by GDP, generated around 16 per- shares, and two centuries of population cent of the world's output in 2005, while the estimates for primary cities, suggest that top 100 generated almost 25 percent. The urbanization is initially rapid before slowing. urban agglomerations of Tokyo and New Developing countries--especially those in York have estimated GDPs (in purchasing Africa and Asia--are at phases during which power parity) broadly similar to those of urban shares increase sharply. People in Canada and Spain, respectively, whereas Western Europe and North America, which London has a higher estimated GDP than went through the same phase a century ago, either Sweden or Switzerland. Similarly, pri- have understandably forgotten. Emerging mary cities in developing countries account economies such as the Republic of Korea that 58 WORLD DEVELOPMENT REPORT 2009 Figure 1.4 The richer a country, the more concentrated its economic mass a. Primary city b. Area of 1° longitude by 1° latitude Gross product (US$ millions) per km2 Gross product (US$ millions) per km2 300 London Singapore 90 250 Seoul Madrid Dublin 70 200 Vienna Tokyo-Kanagawa Paris 150 50 New York­New Jersey Ontario Oslo­Akershus-Vestfold 100 30 Vienna-Mödling 50 10 0 0 5 10 15 20 25 30 ­10 GDP per capita (constant US$, thousands) 0 10 20 30 40 50 GDP per capita (constant PPP US$, thousands) Sources: WDR team estimates based on World Bank (2007j), and databases from www.citymayor.com and www.gecon.yale.edu. Figure 1.5 Developing countries have a pace of urbanization similar to that of early developed rapidly provide the best case stud- developers ies for understanding the pace and pattern of a. At magnified scale: GDP per capita < $10,000 geographic concentration. Their experience Urban share (%) traces the initially rapid and the more grad- 100 ual growth of today's wealthiest nations. 90 At the aggregate level, using the popula- 80 tion shares in urban areas, the urbanization 70 pattern of developing countries in Asia, 60 Africa, Middle East, and Latin America 50 over the last 50 years closely tracks the first 40 part of the historic path earlier traversed by 30 OECD countries between 1900 and 2000 20 (figure 1.5). The urbanization in Asia mir- 10 rors the rapid phase of urbanization that 0 0 1 2 3 4 5 6 7 8 9 10 OECD countries experienced in the nine- GDP per capita (1990 int'l Geary-Khamis $, thousands) teenth century. Likewise, the geographic transformations in Latin America and the b. Full range of GDP per capita Urban share (%) Caribbean, in Eastern Europe and Central 100 Asia, and in the Middle East and North 90 Africa are qualitatively similar to those 80 experienced by the OECD in the first phase 70 of urbanization. Quantitatively, the urban 60 shares for Latin America and the Carib- 50 bean and for Eastern Europe and Central 40 Asia regions are higher than those for the OECD at comparable incomes. 30 OECD Middle East & North Africa This may, however, be an artifact of the 20 East Asia & Pacific Latin America & Caribbean South Asia Eastern Europe & Central Asia data. Data from the World Urbanization 10 Sub-Saharan Africa Prospects database systematically overstate-- 0 0 5 10 15 20 25 purely as a definitional matter--the urban GDP per capita (constant int'l Geary-Khamis $, thousands) shares of Latin America and the Caribbean, Eastern Europe and Central Asia, and Sub- Sources: Maddison 2006; United Nations 1969; United Nations 1949; United Nations 1952; Historical Database of the Global Environment; United Nations 2006c. Saharan Africa. The safest conclusion may be that the pattern of urbanization--the Density 59 BOX 1.4 Africa's urbanization reflects industrialization Between 1970 and 1995, the urban popula- rate population estimates, a population of at in total GDP--a doubling of their tions in Sub-Saharan Africa were growing least 1 million in 1995, and data on sectoral economies--also witnessed the fastest at 5.2 percent a year while their GDP per value added for 1970 and 1995. growth in urban population--a four- capita was shrinking at 0.66 percent a year. This whittles the sample down to just 10 fold increase. The leaders in the sample Since the work by Fay and Opal (2000), countries: Benin, Botswana, Central Afri- were Benin and Zimbabwe. many have argued that urbanization does can Republic, Ghana, Mauritania, Niger, · The pace of urbanization was positively not necessarily accompany development, Rwanda, Senegal, Zambia, and Zimbabwe. correlated with growth in industries with Sub-Saharan Africa in mind (Com- Of these 10 countries, ve experienced and services, activities predominant in mission for Africa 2005). But Satterthwaite con ict at least once, and the other ve urban areas. (2007) questions the validity of the urban were peaceful throughout the period. The population numbers in most studies. Since results do not appear to di er systemati- These patterns do not support the claim many were based on projections, some cally between these two sets of countries. of African urbanization without growth. may have been grossly overestimated. The main ndings follow: In contrast, countries with higher GDP The problem is the lack of regular popula- growth experienced faster urbanization, tion censuses. For Chad and Eritrea the pop- · Except for Botswana, the countries and rapid urbanization came hand-in- ulation projections spanning 1950 through experienced on average a doubling hand with higher growth in industries 2030 were based on one population census. of population, but only 60 percent and services. A counterfactual of an Africa Those for the Democratic Republic of Congo cumulative growth in GDP. Population without urbanization is one with even were derived from two observations, the growth outpaced increases in gross slower economic growth, greater GDP per most recent for 1984. It is thus reasonable value added, and GDP per capita fell. capita losses, and increases in poverty. to consider only countries with at least · Urban population growth and total two censuses during the period examined GDP growth are positively correlated. Sources: Fay and Opal 2000; Satterthwaite (1970­95), a census post-2000 for more accu- Countries with the fastest growth 2007; United Nations 2006c. relationship between economic growth and At a disaggregated level, the primary urbanization--is not unprecedented. Even city's population share of a country dis- in Sub-Saharan Africa, faster urbanization plays a similar, nonlinear pattern of initially between 1970 and 1995, albeit with negative rapidly rising concentration, followed by a GDP per capita growth, was associated with subsequent leveling (figure 1.6). This inten- higher total GDP growth. Urbanization also sification of economic mass within a coun- came hand-in-hand with rapid growth in try's largest cities is seen for a wide range industries and services (see box 1.4). of incomes, from Budapest, Cairo, Kuala Figure 1.6 Density intensifies rapidly in the early phase of urbanization before leveling off % city population to % city population to national population national population Santiago, 1800­2000 35 35 Athens, 1800­2000 30 30 Vienna, 1800­2000 Lisbon, 1800­2000 25 25 Dublin, 1800­2000 Seoul, 1800­2000 20 20 Sydney, 1800­2000 Budapest, 1850­2000 Toronto, 1800­2000 15 15 Cairo, 1800­2000 Zurich, 1800­2000 10 São Paolo, 1850­2000 10 Brussels, 1800­2000 Kuala Lumpur, 1900­2000 5 5 Warsaw, 1850­2000 0 0 0 5 10 15 0 5 10 15 20 25 GDP per capita (constant int'l $, thousands) GDP per capita (constant int'l $, thousands) Sources: WDR 2009 team estimates, based on the Staff City Population Database, Human Settlements Group, International Institute for Environment and Development (IIED). Data from 1950 to the present, primarily from United Nations (2006c); data before 1950, primarily from Chandler and Fox (1974), Chandler (1987), and Showers (1979). Latin America drew on a review of 194 published censuses. 60 WORLD DEVELOPMENT REPORT 2009 Lumpur, and Warsaw to Athens, Lisbon, urban share and development holds until Santiago, and Seoul. These evolutions have a GDP per capita of around $10,000. This also been observed in Brussels, Dublin, Syd- incipient urbanization is associated with a ney, Toronto, Vienna, and Zurich over the rapid shift in the number of people moving two centuries since 1800. from rural to urban areas. Subsequently, the pace of urbanization slows and density Again today, rapidly rising levels off as the urban share surpasses 60 concentration, then a leveling off percent, and the level of GDP per capita A similarly shaped pattern reappears in con- surpasses $10,000. With only a handful of temporary comparisons between a country's exceptions, countries with GDPs per capita level of development and the concentration of above $25,000 have an agglomeration index density. During 2000­05, the average urban above 70 percent. population growth for low-income countries Administratively defined areas. Tak- was 3 percent a year--faster than upper- ing individual cities as the geographic middle-income countries at 1.3 percent and unit, a positive concave relationship exists high-income countries at 0.9 percent. The between a country's level of development relationship is robust. It holds for a variety and its primacy--the share of urban popu- of concentration measures, ranging from the lation living in the country's primary city, a agglomeration index, to population, gross widely used concentration measure. Similar product, and household consumption den- to the relationship between agglomerations sity. It is robust to geographic scale: an area and the level of development, primacy also of 1 square kilometer, a city, a grid cell of 1° rises rapidly before stabilizing during the longitude by 1° latitude, and an aggregated latter stages of urbanization (see figure 1.8, urban sector. panel a). Population and output density are Local 1-square kilometer areas. Esti- highly correlated, but population density mated agglomeration indexes produce a understates the geographic concentration pattern similar to the historical time series: of economic mass. Agglomeration econo- rapidly rising density for countries during mies, the benefits that firms and workers the early phase of urbanization (figure 1.7). enjoy as a result of proximity, make it likely This strong positive relationship between that output density will increase more than proportionately with employment or popu- lation density. Figure 1.7 Shares of population living in urban agglomerations rise with the level of 1° longitude by 1° latitude. Using the development terrestrial grid cells to estimate concentra- Agglomeration index tion as the share of the densest cell's gross 1.0 Egypt, Arab Rep. of Japan product in the country's GDP, concentra- tion of economic mass rises rapidly among countries with a GDP per capita of less than 0.8 Korea, Rep. of United States $15,000, and then stabilizes and tapers off among higher-income countries (see figure 0.6 1.8, panel b). Brazil India Urban areas of countries. Concentra- South Africa tion measured by consumption, rather 0.4 Norway than by population or GDP, suggests the China same concave relationship with the level of 0.2 development. For instance, the urban shares of household consumption in Malawi and Ethiopia Cameroon at GDPs per capita of $150 and 0 0 5 10 15 20 25 30 35 40 $700, respectively, are 36 percent and 48 GDP per capita (PPP, constant 2000 US$, thousands) percent. At about 63 percent, the shares are Sources: Calculated by WDR 2009 team using Nelson (2008) and World Bank (2006g). higher for Jordan and the Arab Republic of Note: The size of each circle indicates the population size of that country. PPP = purchasing power parity. The Egypt with GDP per capita of around $1,600, agglomeration index uses the following criteria: density of 150 persons per kilometer or more, access time of 60 minutes or less to a sizable settlement, defined as one that has a population of more than 50,000. and rise to 80 percent in Panama and Poland Density 61 Figure 1.8 Geographic concentration of population, gross product, and household consumption rises sharply with development, then levels off Cross-country evidence, late 1990s and 2000s a. Population b. Economic mass c. Consumption Spatial unit: city Spatial unit: grid cell of 1° longitude by 1° latitude Spatial unit: aggregated urban areas Gross product in densest area Urban share (%) of % urban population in largest city as % of country's total GDP household consumption 50 50 100 40 40 80 30 30 60 20 20 40 10 10 20 0 0 0 0 10 20 30 40 0 10 20 30 40 0 1 2 3 4 5 6 7 GDP per capita (constant US$, thousands) GDP per capita (constant US$, thousands) GDP per capita (constant US$, thousands) Sources: WDR 2009 team estimates, based on World Bank (2007j), Nordhaus (2006), and more than 120 household surveys for more than 75 countries. with GDPs per capita of $3,500 and $5,000, will approximate a 50/50 urban-rural split. respectively (see figure 1.8, panel c).26 During more advanced urbanization--now a within-urban transformation in a postin- A portfolio of bigger and denser places dustrial area--the distribution of popula- It follows from these stylized facts of geo- tion can be approximated as 75 percent graphic transformation that high-income urban and 25 percent rural. countries have a portfolio of places with a This generalization corresponds well to higher proportion of large settlements and the experience of the United States. In 1690, a lower proportion of small settlements when the average GDP per capita was a mere than do middle-income countries. And the $500 (1990 international dollars),27 the pri- middle-income countries have a signifi- mary city in colonial British America was cantly higher proportion of medium-size Boston. With a population of 7,000, how- settlements than do low-income countries. ever, Boston was by modern-day standards In low-income countries, about three- little bigger than a small town. In the urban quarters of the population live in small hierarchy, only three other cities had popu- settlements of less than 20,000 people, lations greater than 2,500, two of them New and only 10 percent live in urban agglom- York and Philadelphia. The early phase of erations of more than 1 million people. In American industrialization brought with it high-income countries, the opposite is true. an increase in the urban share from 7 per- Less than a quarter of the population live in cent in 1820 to 20 percent in 1860, as GDPs small settlements of less than 20,000 peo- per capita rose from $1,257 to $2,170 (1990 ple, and about half of the population live in international dollars). During this time, the settlements of more than 1 million people population of the primary city, now New (see table 1.1). York, expanded from 123,706 to 805,651. Its At an incipient stage of urbanization, rapid growth allowed the urban hierarchy the portfolio of places in a small country to expand and stretch out. or part of a larger country, such as a prov- ince or even a large district, can be approxi- Table 1.1 The size of urban settlements grows with development mated as 75 percent rural and 25 percent Low-income Middle-income High-income urban, all settlements of relatively low den- Population size countries (%) countries (%) countries(%) sity. As urbanization accelerates--still pre- Small settlements: less than 20,000 73 55 22 dominantly a rural-urban transformation Medium settlements: 20,000 to 1 million 16 25 26 driven by industrialization--and the area or province grows toward a GDP per capita Large settlements: more than 1 million 11 20 52 of $10,000, its distribution of settlements Source: World Bank 2007j. 62 WORLD DEVELOPMENT REPORT 2009 The number of cities with a population and public health facilities in urban areas. greater than 1 million increased from just Along with diverging wages, this promotes one, New York, in 1820 to nine in 1860. All divergence in more basic measures of wel- these cities were in the Northeast, where fare between urban and rural areas.30 But industrialization began. As the geographic rural-urban disparities begin to narrow as transformation wore on, and the United the urbanization process slows, and gov- States completed its transition to a mature ernments become more capable. The exo- industrial economy, population density in a dus of people and workers from rural areas consistent sample of U.S. cities with popu- to towns and cities reduces surplus labor lations greater than 25,000 increased from from the land in agriculture--and reduces 7,230 persons per square mile to 8,876 per competition between workers in rural labor square mile. The average land area of a city markets. And labor-saving technological increased from about 19 square miles to 40 progress releases labor for migration to square miles.28 Cities became more packed urban areas and improves productivity. In and more sprawling at the same time. time, investments and fiscal redistributions give rural residents better local access to Convergence--rural-urban and basic amenities, such as a clean daily source within cities of running water, sanitation, and electricity, A"bumpy"economicgeographydistributing as well as schooling and health care. Indeed, production and people unevenly across the with development and the passage of time, space in a country is a natural feature of the a country's economic geography approxi- working of a market economy. This bumpi- mates a "natural" balance that equalizes ness tends to become more pronounced as a welfare between urban and rural residents. country develops. The question often asked In this situation, people choose to live where is: what does this do to the geographic distri- they expect to be best off in material and bution of poverty, consumption, and other nonmaterial well-being. The Islamic Repub- living standards? The answer can determine lic of Iran illustrates this rural-urban con- the political and social sustainability of the vergence (see box 1.5). process of concentration. Evidence from today's industrial coun- tries suggests that development has largely Rural-urban disparities in well-being-- eliminated rural-urban disparities. High first wide, then narrow urban shares and concentrated economic Rural-urban disparities in productivity, density go hand in hand with small differ- wages, and well-being can be expected to be ences in rural-urban well-being on a range large and increasing in the earlier stages of of indicators. The 15 countries that joined development. With the rapidly increasing the European Union (EU) before 2004, all concentration of economic mass in a coun- with GDPs per capita in excess of $13,000 try's towns and cities in the earlier stages of (1990 international dollars), consider the development, significant disparities in pro- unemployment rate an important policy ductivity, wages, and basic welfare occur target.31 But rural-urban unemployment between urban and rural areas. The agglom- differences should not be a concern. The eration of capital, consumers, and workers unemployment rates are 10.1 percent for quickly brings production advantages, and urban areas, and 9.9 percent for rural areas. transport costs restrict the benefits to the This is also evident for youth: 19.4 percent locality. These larger local markets enable in urban areas compared with 18.7 per- firms to spread the fixed costs of production cent in rural areas. The rates of labor force across a wider number of consumers, pro- participation in urban and rural areas are ducing cost and productivity advantages.29 68.3 and 69.4 percent, respectively.32 For This means higher wages in towns and cities, England, the high degree of rural-urban and greater availability of a more diversified equality in well-being is reflected in similar range of goods and services. disposable incomes: indeed, at £522, weekly The concentration of mass also helps to disposable income in villages is 10 percent ensure a better supply of basic infrastructure higher than the £476 in cities.33 Density 63 BOX 1.5 Urbanization and narrowing rural-urban disparities in the Islamic Republic of Iran Rural-urban disparities have narrowed · First, the share of the urban population the lagging provinces. Between 1976 in the Islamic Republic of Iran. In 1976, has increased from 49 to 67 percent and 1996, the female literacy rate rose on the eve of the Iranian revolution, the between 1979 and 2005. This is a con- from 17 to 62 percent, while for urban mean per capita household income in tinuation of a longer-term trend: the women it rose from 56 to 82 percent. rural areas was 44 percent of that in urban urban population had grown by 5.4 During 1994­2000, infant mortality areas. By 2005, it had increased to 63 percent per year (and in Tehran by 6 and under-5 mortality fell fastest in the percent. percent) between 1966 and 1976. poorest provinces. The Shah's government favored cities · Second, the rural-urban gap in house- · Finally, overall poverty has fallen. The over the countryside. Price controls for hold incomes has narrowed. Between national poverty rate was at 8.1 percent essential foods depressed agricultural 1976 and 1984, agricultural value added in 2005, with relatively modest di er- incomes. High tari s, import bans, and grew by 31 percent, twice the rate of ences in rural and urban poverty of 10 licensing for industrial goods propped the nonoil economy. One reason for this and 7.1 percent, respectively. But pov- up prices of manufactured goods and growth was that farmgate prices rose 55 erty rates still vary a lot between prov- depressed farmers' purchasing power. percent. Another reason was that more inces, ranging from 1.4 to 23.3 percent. An inward-looking development strategy was spent on projects to increase the The political commitment to spatial oriented toward nal domestic demand productivity of small and medium-size equity has produced mixed outcomes ampli ed internal migration to Tehran farms. Growth could also be attributed during the last 30 years: overall poverty and a few other large cities. For every to the fact that agricultural production in declines and a convergence in rural- indicator of development, the center per- the Islamic Republic of Iran is dominated urban standards of living, but persistent formed far better than the periphery. In by the private sector, whereas large di erences in interprovincial living stan- 1973, the poverty rate was 23 percent in industrial enterprises and service provid- dards. the central region and 42 percent for the ers were nationalized after the revolu- country. This spatial inequality matched tion, which hindered their e ciency. the nation's ethnic map, fueling tensions. What has happened since the commit- · Third, rural and urban human devel- Based on a contribution by Anton Dobro- ment in 1979 to address spatial disparities? opment indicators improved, even in nogov, Alexander Kremer, and others. For 21 of the 30 OECD countries, the three times higher. But for OECD countries higher the GDP per capita in 2003,34 the with average GDPs per capita above $10,000, lower the ratio of GDP per capita in predom- the ratio is between one and two (except for inantly urban areas to that in rural areas Norway). Given the well-developed fiscal (see figure 1.9).35 For the Czech Republic, redistribution mechanisms in OECD coun- Hungary, Poland, the Slovak Republic, and tries, and differences in age-demographic Turkey, with an average GDP per capita profiles between urban and rural areas, below $10,000 (1990 international dollars), these disparities in GDP per capita will GDP per capita in urban areas is two to overstate rural-urban differences in, say, Figure 1.9 Rural-urban disparities in GDP per capita tend to be smaller in richer OECD countries Ratio of urban to rural GDP per capita Ratio of urban to rural GDP per capita 3.50 3.50 3.00 3.00 2.50 2.50 2.00 2.00 1.50 1.50 1.00 1.00 0.50 0.50 0.00 0.00 5 10 15 20 25 30 40 50 60 70 80 90 100 GDP per capita (constant 1990 int'l $, thousands) Agglomeration Index, 2000 (%) Source: WDR 2009 team, based on data from OECD (2007), pp. 1­256. 64 WORLD DEVELOPMENT REPORT 2009 Table 1.2 Rural-urban disparities in earnings, wealth, and consumption characterize development over the last two centuries Rural-urban Country (year) disparity (%) Description and country sample Sweden (1805) 221.0 Wealth per male adult in urban and rural areas. Finland (1805) 146.0 Wealth per male adult in urban and rural areas. England (1830s) 73.2 Urban wages are wages per laborer in the building trades, and rural wages are for agricultural laborers. France (1882) 29.0 Urban wages are for unskilled wages in the regional capital city (department chef lieu), and rural wages are France (1911) 51.0 based on average farm wages . United States (1925) 28.0 Urban earnings are manufacturing earnings, and rural earnings are agricultural earnings. United States (1935) 75.0 Developing countries 51.2 Urban wages are for unskilled general laborers, and rural wages are agricultural wages, including payments (nineteenth century) in kind. The countries included are Argentina 1872; Australia 1887; Denmark 1872; France 1892, 1801; Hungary 1865; Japan 1887; and the United States 1820­29, 1890. Developing countries 41.4 Urban wages are based on wages for unskilled construction workers, and rural wages are agricultural cash (twentieth century) wages. There are 19 countries (1960­70) underlying this average: Argentina, Cameroon, Chile, Costa Rica, Côte d'Ivoire, Guatemala, Kenya, Pakistan, Malawi, Malaysia, Mexico, Morocco, Panama, Sri Lanka, Tanzania, Trinidad and Tobago, Tunisia, Uruguay, and R. B. de Venezuela. Developing countries 42.0 Based on per capita household consumption, after controlling for household characteristics. There are 72 (twenty-first century) countries (2000­05) underlying this average disparity: Armenia, Angola, Bangladesh, Belize, Benin, Bhutan, Bolivia, Brazil, Burkina Faso, Burundi, Bulgaria, Cambodia, Cameroon, Chad, Chile, Colombia, Dem. Rep. of Congo, Costa Rica, Côte d'Ivoire, Croatia, Djibouti, Ecuador, Arab Rep. of Egypt, El Salvador, Ethiopia, The Gambia, Georgia, Ghana, Guatemala, Guinea, Guyana, Honduras, Hungary, India, Indonesia, Jamaica, Jordan, Kyrgyz Republic, Madagascar, Malawi, Maldives, Mali, Mauritania, Mexico, Moldova, Mongolia, Morocco, Mozambique, Nepal, Nicaragua, Nigeria, Pakistan, Panama, Paraguay, Peru, Philippines, Poland, Romania, Russian Federation, Rwanda, Senegal, South Africa, Sri Lanka, Swaziland, Tajikistan, Tanzania, Thailand, Timor-Leste, Uganda, Ukraine, Vietnam, and Zambia. Sources: Sweden and Finland 1805: Soltow 1989, table 1, p. 48; England 1830s: Williamson 1987, table 3, p. 652; France 1882, 1911: Sicsic 1992, table 2, p. 685; United States 1925, 1935: Alston and Hatton 1991, table 3, p. 93; Developing countries (nineteenth century): Clark 1957, table II pp. 526­31; Developing countries (twentieth century): Squire 1981, table 30, p. 102; Developing countries (twenty-first century): WDR 2009 team estimates based on individual country's household survey for 72 countries; the data set is described in detail in Montenegro and Hirn (2008). Note: Rural-urban disparity (in nominal terms) is computed as the difference in wages, earnings, wealth, or consumption between urban and rural areas relative to the rural averages. average levels of personal disposable income disparities in productivity and income. and consumption. The agglomeration index For a sample of developing countries in the produces the same qualitative pattern. 1960s--among them Malaysia, Mexico, Rural-urban disparities in these countries and Trinidad and Tobago, which have since were wide throughout the nineteenth and reached upper-middle-income or high-in- early twentieth centuries. Wealth per male come status--urban wages exceeded rural adult in nineteenth century Sweden was more wages by more than 40 percent. Similar gaps than 200 percent higher in urban areas than can be observed in per capita consumption in rural areas, and 150 percent higher in Fin- between urban and rural areas for a recent land (see table 1.2). Meanwhile, for rapidly sample of 72 developing countries. urbanizing England, urban wages were more The rural-urban discrepancy between than 70 percent higher than rural wages in economic mass and population distributions the 1830s. France and the United States saw diminishes with urbanization. Another way big increases in the urban wage premium to examine consumption disparities between from 1882 to 1911 and from 1925 to 1935. urban and rural areas is to look at the popu- Indeed, in the United States, the premium lation share of a country's urban areas and increased almost threefold in a decade.36 For compare it with the share of consumption developing countries in the nineteenth cen- in these areas. If this ratio is greater than tury, including Australia, Denmark, France, one, consumption per capita is, on average, Japan, and the United States, urban nominal higher in urban areas than in rural areas, wages were 50 percent higher. while the converse is true if the ratio is less Today's developing countries are still than one. in the first phase of urbanization and, Rural-urban disparities in consump- not surprisingly, have large rural-urban tion fall with density in today's developing Density 65 Figure 1.10 Rural-urban gaps in per capita points. For countries where urbanization is consumption become smaller with urbanization advanced and the urban share is approach- Ratio of urban consumption share ing its natural maximum, almost no differ- to urban population share ence exists between urban and rural areas 4.0 in access to basic services. Equalization of 3.5 access to basic services can be expected to 3.0 promote a corresponding convergence in 2.5 nonmaterial indicators of welfare and liv- 2.0 ing standards (see table 1.3). 1.5 Narrowing rural-urban disparities is 1.0 important, but the progress in absolute 0.5 measures of basic welfare in the rural areas 0 of the world's poorest countries is even more 0 20 40 60 80 100 Urban population share (%) important.Risingrural-urbandisparitiesare consistent with an absolute improvement in Source: WDR 2009 team estimates from more than 120 house- hold surveys for more than 75 countries. basic welfare in both rural and urban areas. The overall evidence is encouraging. Over countries (see figure 1.10).37 In Malawi and the past decade, most low- and middle-in- Sri Lanka the ratio is around two: urban come countries have experienced absolute areas account for about 10 percent of the improvements on a range of basic welfare population but 20 percent of consumption. indicators, including infant and under-5 For countries with higher levels of urbaniza- mortality rates, malnutrition, immuniza- tion, the spatial distribution of population tion, and school participation in rural and more closely resembles that of production. urban areas. Of 32 low-income countries, Madagascar and Tanzania have urban popu- three-quarters reduced infant and under-5 lation shares of around 20 to 25 percent and mortality rates and the incidence of severe urban consumption shares of about 30 to stunting and severe underweight, especially 35 percent. By the time a country enters an in rural areas.40 And since 1990, school advanced stage of urbanization, population attendance rose in four-fifths of these is more or less proportionately distributed countries, especially in rural areas.41 Both with economic mass, so that the ratio is close to one. In Chile 85 percent of the popula- Table 1.3 Rural-urban disparity in basic services narrows with development tion reside in urban areas, and these urban Disparity in Disparity in access residents account for 92 percent of national Urban population access to clean to sanitation consumption. In Brazil 80 percent of people share (mean GDP water (percentage (percentage Examples of countries in live in urban settlements, and these 80 per- per capita) points) points) the sample cent are responsible for 85 percent of con- 75% or higher 8 8 United States, Norway, (mean GDP per Switzerland, Spain, sumption. As development progresses and capita: $21,602) Germany, Canada, Mexico, the concentration of economic activity in Chile, Brazil, Argentina, areas of high density increases, rural-urban Gabon, R. B. de Venezuela, Djibouti, Lebanon, Jordan, disparities narrow. A downward sloping line United Kingdom at all levels of urbanization is a good omen: 50%­70% 15 20 Estonia, Panama, Turkey, most developing countries may have passed (mean GDP per Hungary, Ecuador, the peak in their rural-urban disparities.38 capita: $9,672) Colombia, Malaysia, Syria, What is true for private consumption is Azerbaijan, South Africa, Rep. of Congo, Algeria, true for basic amenities. Among low-income Tunisia, Bolivia countries with urban population shares of 25% or lower 24 26 India, Rep. of Yemen, less than 25 percent, access to water and (mean GDP per Madagascar, Chad, sanitation in towns and cities is around capita: $2,585) Tajikistan, Bangladesh, 25 percentage points higher than in rural Tanzania, Kenya, Nepal, areas.39 But for more urbanized countries, Cambodia, Malawi, Uganda, Sri Lanka, Bhutan such as Algeria, Colombia, and South Africa, Source: World Bank 2007j. the disparity in access is 15 to 20 percentage Note: Disparity refers to the percentage point difference between urban and rural areas. 66 WORLD DEVELOPMENT REPORT 2009 Figure 1.11 Even at the subnational level, rural-urban disparities fall as density increases Philippines, 2000 China, 1999 and 2006 India, 1983 and 1994 Ratio of urban to Ratio of urban disposable income Disparity in life expectancy rural incomes to rural net income urban-rural ratio (by state) 4.0 6 1.25 3.5 5 1.20 3.0 4 2.5 1.15 2.0 3 1983 1.5 2006 1.10 2 1.0 1999 1994 1 1.05 0.5 0.0 0 0 0 20 40 60 80 100 0 20 40 60 80 100 0 0.1 0.2 0.3 0.4 0.5 Urban share (%) Urban population share (%) State-specific urban share (%) Sources: Balisacan, Hill, and Piza forthcoming; Yao forthcoming; Cali 2008. urban and rural areas in these nations have and deprivation. Disparities within cities achieved progress toward the Millennium can be large. In Nairobi poverty is high in Development Goals. the inner city but much lower in the rest of Rural-urban convergence takes place the city and the suburbs (see figure 1.12). In sooner in more urbanized subnational Mombasa, Kenya's second-most-populous areas. In both China and the Philippines, city, marked geographic divisions in the urbanized provinces exhibit lower internal poverty rate are evident (see map 1.2). South urban-rural disparities in incomes (see fig- African cities also show internal disparities ure 1.11). In China the entire relationship in the poverty rate. Cape Town has a low has shifted upward over the past decade poverty rate in the coastal areas, but a higher so that, in general, rural-urban disparities poverty rate in the interior of the city. Simi- have increased over time, consistent with larly, both Johannesburg-Pretoria-Tshwane China's early stage of development, which and Durban have visible divisions. But the is marked by rapid urbanization. In India geography of poverty in Durban is different rural-urban gaps in life expectancy were from that in Cape Town and Johannesburg: smaller in the more urbanized states in the poverty rate is, in general, higher outside both 1983 and 1994. But the entire relation- the city boundaries than inside. ship has shifted downward over time. The most obvious sign of divisions within cities is slums. Slums have chronically over- Slums--divergence and convergence crowded dwellings of poor quality in under- within cities served areas. The reason for the lack of basic In poor countries, higher average living public services and infrastructure is the standards in cities do not rule out poverty inability or unwillingness of many urban Figure 1.12 Slums grow with the pace of urbanization, and fall with its level Annual growth of slum population (%) Percentage of slums 20 120 15 100 10 80 5 60 0 40 ­5 20 ­10 0 ­2 0 2 4 6 8 10 ­20 0 20 40 60 80 100 120 Annual growth of urban population (%) Urban share of a country (%) Source: Kilroy 2008. Density 67 Map 1.2 Local divisions--spatial disparities within urban settlements can be large Poverty rates in African cities KENYA SOUTH AFRICA Nairobi Johannesburg and Pretoria Poverty rate: Poverty rate: Mombasa Proportion of Proportion of poor (%) poor (%) 0­11 17­42 11­25 42­48 25­40 48­58 40­57 Durban 58­63 57­92 63­84 Cape Town No data No data Source: The Poverty Mapping Project, Columbia University, using data from Alderman and others (2002); Statistics South Africa; the Central Bureau of Statistics, Kenya; and the Ministry of Planning and National Development, Kenya. governments, utilities, and service provid- Slums are part of rapid urbanization, and ers to operate in slums, generally because of it is not uncommon for a fifth to a third of the informality and illegality of such settle- a city's population in a contemporary devel- ments.42 So living standards, especially oping country to reside in slums (see figure health, security, and sanitation, are lower in 1.12).44 Goiâna, the capital of the Brazilian slums than in formal settlements close by. state of Goiàs, a medium-size city of 40,000 Mumbai's Dharavi, believed to be Asia's big- in 1950, is today a city of more than 1 mil- gest slum, has "maybe a million residents . . . lion, with much of the population increase crammed into a square mile of low rise wood, accommodated in slums.45 Since 1950, concrete and rusted iron . . . a family of 12 liv- Delhi's population has risen more than ing in a 90-square-foot room." In Shiva Shakti tenfold, from 1.4 million to 15.6 million,46 Nagar, again in Mumbai, each community accompanied by an increase in the number tap is shared by roughly 100 people.43 of slum clusters from 200 to 1,160. The growth of slums in major cities "A dirtier or more wretched place he is characteristic of rapid urbanization. had never seen. The street was narrow and Because rapid population growth cannot muddy, and the air was impregnated with be satisfactorily accommodated, slums and filthy odors. . . . Covered ways and yards, shantytowns grow bigger and more visible. which here and there diverged from the This contributes to wide and increasing main street, disclosed little knots of houses, geographic divisions in well-being within where drunken men and women were posi- urban areas. Development--both economic tively wallowing in filth." A contemporary and institutional--and better infrastruc- description of a developing country slum ture, combined with focused interventions, such as Nairobi's Kibera or Huruma, Abi- eventually bring about a convergence in liv- djan's Washington, Delhi's Majboor Nagar or ing standards in urban areas. Kanchan Puri, Buenos Aires's San Fernando, 68 WORLD DEVELOPMENT REPORT 2009 in multistory tenements arranged along nar- row, unlit foot passages. This "housing was BOX 1.6 Slums, then and now hopelessly inadequate in all respects--in The term "slum," probably originating reached out beyond the working quantity, in quality and environmental from an old English or German word class, nally motivating strong politi- amenities, if needs as basic as clean water meaning a poorly drained or muddy cal action. But rather than attempting and safe sewage disposal can be described as place, was applied to housing in the to stop more workers from coming, amenities."47 Apart from the obvious mis- early Industrial Revolution in the or clearing out these areas of disease United Kingdom before the railways and poverty, the government in the ery, slums were prone to deadly outbreaks of were in place, when canals trans- 1870s passed legislation for strict measles and scarlet fever and high rates of ported heavy goods along the length building regulations, prescribing the mortality attributable to diarrheal diseases, and breadth of the country. During dimensions of streets and houses, and typhus, and respiratory diseases.48 Britain's rapid industrialization, most making it mandatory that all dwell- Yesterday's slums are today's world-class factories were built beside canals, the ings be connected to newly built cities. Britain is not the only industrial coun- main channel for transporting coal sewerage systems. Major municipal try to suffer from slums and wide intracity for their steam engines and other investments in water works, sewage inputs of production. facilities, and public health dramati- divisions in welfare during the earlier phases Poor workers, migrating to cities cally reduced mortality in Britain's of development and rapid urbanization (see for factory jobs, could ill a ord to cities between 1874 and 1907. box 1.7). The stylized pattern of divergence walk long distances to and from their Despite atrocious and lthy con- followed by convergence is a hallmark of places of work. Before electric trams, ditions, millions of migrants keep other modern-day developed countries as other forms of transport were expen- leaving rural areas for the teeming well. Slums for these cities are now much a sive. So workers settled close to fac- economic opportunity o ered in the thing of the past. Aided by improving land tories. Cheap housing grew around cities of poor and middle-income these factories in low-lying, poorly countries. Even though health hazards markets, investments in infrastructure, and drained areas. Housing was over- and mortality rates are far worse in the targeted incentives, within-city welfare dis- crowded. Sanitation was inadequate shanties around many cities in Africa, parities tend to narrow, but only in the more and in most cases nonexistent. And people there are trading, working, and advanced stages of urbanization. Indeed, for air quality was poor, with soot and sending large sums of money home. "world" cities such as London, New York, other pollutants. Sickness was com- The challenge facing policy makers Paris, Singapore, and Tokyo, slums can, with monplace. Diarrhea, typhus, respira- today is similar to that faced by the the benefit of hindsight, be viewed as part of tory diseases, measles, and scarlet Victorians in London: how to nurture fever cut the life expectancy of those these agglomerations with functional their "growing pains." Britain cleaned up its born in cities by 12 years compared land markets, better transport, and Dark Satanic Mills over a century, and if it with those born in rural areas. public health infrastructure to capture had started the cleanup sooner, the working The growing public health hazards the bene ts of economic growth. class would have suffered from slower wage in Britain's urban slums exacted a Sources: Satterthwaite and others 2007; growth and lower consumption.49 terrible health toll that eventually Crafts 2008; The Economist 2007a. The emergence and growth of slums in the early and intermediate stages of a coun- try's development can be explained by the or Rio de Janeiro's Rocinha? No, this is an interaction of functioning labor markets excerpt from Charles Dickens's Parish Boy's with dysfunctional land markets. In the Progress, published in 1838, describing the rapid phase of urbanization, the labor mar- rapidly expanding city of London in the ket signals higher labor demand in urban nineteenth century (see box 1.6). areas, the higher demand that arises from London was by no means the only city growth in industries and services. Labor or urban area in nineteenth century Britain responds by moving to towns and cities. with large slum settlements. Chronically As a reflection of this, slum dwellers in overcrowded and inadequately serviced developing countries are often productively housing was a common feature of British engaged, taking advantage of the economic cities and industrial towns of the time. In opportunities the city offers. Mumbai's Edinburgh rapid population growth and a Dharavi has 15,000 "hutment" factories, first wave of suburbanization by the then- and "the clothes, pots, toys and recycled rising middle classes meant that by the materials its residents produce earn the fac- 1860s, the core of the city had a large slum tories millions of dollars a year." Many slum area with population densities as high as 600 residents started businesses after the state persons per acre. Residents in this area lived government provided them with limited Density 69 BOX 1.7 Many of today's world-class cities were littered with slums "In Antwerp and in most Belgian towns the "Here the background embraces the tal soldier turned milkman, Vergniaud. basic problem in matters of working class pauper burial-ground, the station of the There the Colonel lives in a single room housing was . . . no individual sanitation or Liverpool and Leeds railway, and, in the with a dirt oor and a straw bed." individual water supply. . . . The three heavy rear of this, the Workhouse, the "Poor-Law "Between 1815 and 1851 France's popula- cholera epidemics of the 19th century had Bastille" of Manchester, which, . . . looks tion grew from 29 to 36 million . . . it was terri c e ects in these slums . . . " threateningly upon the working-people's the cities that absorbed the thousands of quarter below. . . . Passing along a rough "The rst encampments of Baltimore's migrants unable to nd work in the country- bank, among stakes and washing-lines, poor were at the water's edge. Time and side. . . . But there were simply not enough one penetrates into this chaos of small again, outbreaks of yellow fever, malaria, jobs. Unemployment and overcrowding cre- one-storied, one-roomed huts, in most of cholera, typhoid fever swept the town. ated appalling living conditions. Only one in which there is no arti cial oor; kitchen, These epidemics seemed peculiarly asso- ve houses had running water. In 1832 chol- living and sleeping-room all in one. In ciated with the low-lying encampments era wiped out some 20,000 Parisians." such a hole, scarcely ve feet long by six of the poor. The yellow fever epidemic of broad, I found two beds--and such bed- "Like so many other European cities, 1797, for example, was said to have begun steads and beds!--which, with a staircase Paris su ered from chronic post-war in the stagnant waters of the Fells Point and chimney-place, exactly lled the housing shortages. Of the 17 slum areas cove and to have spread . . . to the huts room." designed for clearance, most were still and hovels on the banks of the Jones Falls intact in the 1950s." and thence on to the shacks and shanties "Melbourne's most infamous slum, at the foot of Federal Hill." Little Bourke Street, . . . by the 1880s . . . was "One of the worst outrages of indus- crowded, bustling and growing. . . . The trialism in China against humanity is the "By the 1890s, Polish immigrants had lane is completely lled up with all kinds of herding of these workers in noisome supplanted the Irish and Germans, creat- lth comprising garbage tips, putrid liquid, slums in the factory districts, . . . so foul ing a ghetto of a new dimension. Single straw rags, and other rubbish. A most dis- and revolting . . . in Shanghai. . . . There dwellings housed from six to eight families, agreeable odor arose from this o ensive are no sanitary provisions of any kind, and one [family] to a room. . . . Fells Point was mass . . . the loathsome mass . . . exposed the passages between the rows of houses described by a health o cial as an Augean and allowed to rot and spread its contami- are practically open latrines. Overcrowd- stable . . . a mass of nuisance . . . Open nating in uences." ing exists to a distressing extent. The drains, great lots lled with high weeds, many children who are reared in these ashes and garbage accumulated in the "About 200 years ago, Lower Manhat- lthy quarters are covered with running alleyways, cellars lled with black water, tan was adorned by a pretty ve-acre sores from dirt and bodily neglect." houses that are total strangers to the lake known as the Collect. . . . By the mid- touch of whitewash or scrubbing brush, 1700s, however, the Collect was already "In the 15 years between 1930 and the human bodies that have been strangers rimmed with slaughterhouses and tan- end of the war, the population of Singa- for months to soap and water . . . that's neries. The e usions from these bloody pore doubled to a million people. The Pigtown." businesses were poured directly into the population explosion had generated a lake and more industries, more trash, housing shortage of epidemic propor- "The slums of Dublin were among the quickly followed. By 1800 the Collect was tions. Small shophouses gave shelter to worst in Europe, rivaled only by Glas- a reeking cesspool. By 1813 it had been as many as 100 people. The average living gow. Tall town houses, originally built as entirely lled in and by 1825 something space was 9 feet by 9 feet, about the size elegant homes for the rich in the eigh- entirely new stood on the site--America's of a prison cell." teenth century, fell into the Tomae hands rst real slum, the Five Points." of avaricious and pitiless landlords who "All of the ghettos of the 1920s within lled them to bursting point with the "Although this is a hugely expensive the city of Tokyo were products of Tokyo's desperate and impoverished urban poor. area in Paris to live today, in Victor Hugo's urban development and Japan's modern Conditions were often unspeakably vile, day it was a slum area, close to the Bastille economic growth. . . . The sheer size of these with massive over-crowding and utterly Prison." ghettos was astonishing. . . . Poverty pockets inadequate sanitation." re-emerged in all parts of the metropolis of "[T]he lawyer Derville ventures into the Tokyo after the Second World War, even in "Katajanokka's transformation in its slums of Saint Marceau, the poorest sec- the midst of the old city of Tokyo." entirety from a low-income housing area tion at the outskirts of Paris. Taking his to an enclave for the city's civil service coach through the lthy rutted lanes, he Sources: Belgium: Lis; Baltimore: Garrett elite and bourgeoisie represented an arrives at a broken-down building, made 2002; Dublin: Kearns 2006; Helsinki: Mäki- nen; Manchester: Engels 1987; Melbourne: urban growth pattern that emerged for entirely of second-hand materials and Mountford; Manhattan: Baker 2001; Paris: the rst time in the history of Helsinki. poorly built, where Colonel Chabert is Sanderson, Villon 2000, The Economist; A former slum had become a prestigious lodged with the cows, goats, rabbits and Shanghai: Schwenning 1927; Singapore: residential area for the privileged classes." impoverished family of a former regimen- Baker 1999; Tokyo: Koji 1969. 70 WORLD DEVELOPMENT REPORT 2009 rights over their dwellings in 1976 and began European countries lived in urban settle- to supply water and power to parts of the ments of 5,000 inhabitants or more.52 In this settlement. Because Dharavi is sandwiched respect, at least, little had changed from the between the city's two main railway lines previous five centuries. So the takeoff into and is surrounded by six stations, it also acts urbanization over the next century broke as Mumbai's transportation hub.50 In short, dramatically from the past. slums arise in many developing countries as low-income households take advantage of The pace and pattern of urbanization spatially concentrated employment oppor- is similar tunities and as businesses take advantage of It started in Great Britain. In 1800 Britain's their location in a land-constrained envi- urban share stood at 19.2 percent, about ronment. Consistent with today's industrial twice the European average. But in the first countries, the correct response is not to two decades of the century, the number of slow, stop, or reverse urbanization. It is to people living in urban areas doubled. By tackle dysfunctional land markets. 1820 the urban share was 40 percent. By the The interplay of such market forces and close of the century, seven of every 10 Brit- responses from rational market actors can ons were living in urban settlements. Britain also be seen in many Sub-Saharan African was joined in its headlong rush into urban- countries. But inefficient land markets, ization by other early European industrial- often thanks to misguided urban plan- izers. By the second half of the nineteenth ning and zoning, produce only a limited century, urbanization spread beyond the and unresponsive supply of affordable, legal Old World to the United States and Canada. land sites for building housing to keep pace By World War I, four of every 10 Americans with the demand.51 were living in urban settlements with popu- lations of 5,000 or greater; just 60 years ear- What's different for today's lier, the ratio was one in 20. developers? So if anything is different for today's At the beginning of the nineteenth century, developers, it is certainly not the pace of one person in every 10 in today's developed urbanization. Indeed, the average pace of Figure 1.13 Urbanization's speed has precedents Percentage point difference in urban population, 1985­2005 (except where specified) 35 30 Canada, 1880­1900 25 United Kingdom, 1830­1850 20 Germany, 1880­1900 Denmark and United States, respectively, 1880­1900 15 Switzerland, Mean of high-income 1880­1900 countries, 1880­1900 Mean of developing 10 countries, 1985­2005 Median of developing countries,1985­2005 5 0 All countries Sources: WDR 2009 team calculations based on data from the United Nations (2006c); historical data for Canada, the United King- dom, and industrial countries' averages are from Bairoch and Goertz (1986) and Dumke (1994). Density 71 urbanization for developing countries over Between 1985 and 2005, China added 1985­2005 is remarkably similar to the 225 million people to its towns and cities, average for European and North Ameri- almost the entire population of the United can countries53 between 1880 and 1900 (see States. Yet China for the same time period, figure 1.13).54 For the early developers the ranked only fifteenth in its absolute increase average absolute increase in the urban share in urban share. In India the number of peo- over the 20 years was 7.7 percentage points, ple in towns and cities rose by 137.8 million, and for current developers the respec- adding a Germany and an Italy to its urban tive median and mean absolute increases areas in just two decades. were 7.1 and 8.0 percentage points. The Today's developing countries had an pace of urbanization among most of the average increase in their urban popula- early developers in the last two decades of tion of 8.3 million over 1985­2005, almost the nineteenth century ranked in the top three times the increase for many of today's quartile of the contemporary distribution high-income European and North Ameri- of urbanization speeds. can countries between 1880 and 1900. But when China and India are excluded from The volume of urbanization is greater the group, the average urban population for today's developers increase in recent decades has only been What then is different? One difference is the 4.4 million, about 50 percent more than unprecedented absolute increases in urban the average for the early developers during populations in many developing countries 1880­1900 (see figure 1.14).57 in recent decades. Today's developing coun- Correspondingly, megacities in devel- tries simply have larger populations than oping countries are unprecedented in their the industrializing countries of the nine- size. Through the nineteenth century the teenth and early twentieth centuries. The world's largest city was London. But its urban population today, estimated at 3.3 1900 population of 6.6 million was only billion, is far greater than the world's total a third that of modern-day Mumbai or population as recently as 1960. It took more New Delhi, the largest cities in low-income than 10,000 years for the urban population countries. The London of 1900 and, indeed, to reach 1 billion in 1960, 25 years to add even the London of today are also smaller the second billion, and only 18 to add the than modern-day Shanghai (10 million), third.55 According to the UN projection, it the largest city in lower-middle-income will take just 15 years to add the fourth.56 In countries, and several others (Cairo, East Asia alone, 500 million people will join Jakarta, and Manila) among the more suc- today's 750 million urbanites over the next cessful developers. With more than 22 mil- 25 years, essentially adding another Paris or lion people, Mexico City, the largest city in Kuala Lumpur every month. upper-middle-income countries, is three Figure 1.14 The population increment in urban areas of today's developing countries is much larger Change in urban population (thousands), 1985­2005 Change in urban population (thousands), 1985­2005 30,000 140,000 United States, 1880­1900 25,000 120,000 100,000 20,000 80,000 United States, 1880­1900 15,000 60,000 Average for developing countries (excluding Average for developing countries, 1985­2005 10,000 China and India), 1985­2005 40,000 Average for developed countries, 1880­1900 Average for developed countries, 1880­1900 5,000 20,000 0 0 All countries including China and India All countries excluding China and India Sources: WDR team calculations based on data for 1985­2005 from the United Nations (2006c) plus historical data from Bairoch and Goertz (1986). 72 WORLD DEVELOPMENT REPORT 2009 times the size of London at the start of the workers in London earned an urban real twentieth century. wage premium of 67 percent, a large part of this premium was compensation for the evi- Urbanites today enjoy both higher dent health hazards of city living.64 private earnings and better public In Germany during the second half of the services nineteenth century, infant mortality rates Cities now do better than rural areas in both in rural areas were about 150 per 1,000 live income and nonincome indicators of well- births. But expanding Berlin had the high- being. In 2000 the infant mortality rate in est infant mortality in the Kaiserreich era, rural Malawi was 117 per 1,000 live births, hovering around 300 per 1,000 live births in in urban Malawi it was 83. Urban Benin the 1860s, and peaking at 410 per 1,000 live did much better than rural Benin in low- births in the 1870s. The rural-urban gap in ering under-5 mortality rates and reducing physical well-being remained for decades diarrhea and acute respiratory infections.58 during the nineteenth century.65 Urban Ugandan women were less likely to As the U.S. economy industrialized and suffer from anemia or malnutrition. Supe- urbanized, people living in high-density rior health indicators are repeated in urban areas at the turn of the twentieth century areas throughout the developing world-- were exposed to infectious and parasitic from Chad and Cameroon in Sub-Saharan diseases. In 1880 urban mortality for adults Africa, to Nepal in South Asia, Kazakhstan was 50 percent higher than rural mortality, in Central Asia, and Nicaragua in Latin and two decades later, the urban mortality America, and to Morocco and Egypt in rate was still 18 percent higher. The rural- North Africa and Middle East.59 urban mortality difference was even greater But the opposite was true for the devel- for infants and young children. For infants, opers of the nineteenth and early twentieth excess urban mortality was 63 percent in centuries. Migrants to cities could expect 1890 and 49 percent in 1900, and for young better material standards of living, offset by children ages one to four, the respective fig- poorer health and shorter lives for them and ures were 107 percent and 97 percent. In 1900 their children. In 1881­91 life expectancy at male life expectancy was 10 years shorter in birth was 51 years in English and Welsh vil- urban areas than in rural areas.66 lages, but only 44 years in London and 39 That the cities and towns of modern- years in large towns.60 In 1850s Britain the day developing countries do better than infant mortality rate in cities with popula- villages on indicators of health, while the tions greater than 100,000 was, at 196 per opposite was true for the developed coun- 1,000 live births, far higher than the 138 per tries at similar incomes in the nineteenth 1,000 live births in rural communities.61 century, reflects advances in public health Even as late as 1937, George Orwell saw it and medicine, and improvements in sewers fit to characterize industrial towns and cit- and water systems. It also reflects the pub- ies as places where "one always feels that the lic benefits that today's cities in developing smoke and filth must go on for ever and that countries confer. So the advantages of high no part of the earth's surface can ever escape density are not limited to income genera- them."62 It is perhaps no surprise, then, that tion and wealth creation--they also include the absence of respiratory diseases attribut- social services. able to poor air quality in the cities would With these differences in private and have resulted in life expectancies 4.7 years public sources of well-being, it should longer in the England and Wales of 1861­70. hardly be a surprise that cities and towns in In the absence of cholera, diarrhea, dysen- the developing world are growing rapidly. tery, and typhus, life expectancy might The surprise is that this move to density is have been 1.7 years longer, and the absence not faster. And the policy implication? Any of measles and scarlet fever, common in strategy for a less desperate and more delib- the cities, would have added 2.3 years to erate urbanization must include efforts to life expectancy.63 Thus in the 1830s, while improve public services in rural areas. Distance CHAPTER 2 D eng Xiaoping, generally seen as the standards can happen in parallel. Develop- architect of China's resurgence as ment in the United States was accompanied an economic superpower, insisted by a rapidly rising concentration of manu- onopennesstoworldmarkets.Healsoinsisted facturing activity in a relatively small area on concerted development of the country's of the northeast and eastern part of the coastal areas, like Shanghai and Guangzhou, Midwest at the turn of the twentieth cen- as launching grounds for connecting to these tury.3 Throughout this process, U.S. states markets. When asked about the growing witnessed a slow, if sometimes halting, wealth disparities between the coast and the convergence of per capita incomes.4 Today, interior, he reportedly countered, "If all of roughly half of the U.S. population is in China is to become prosperous, some [areas] only five states,5 but long-term unemploy- must get rich before others." ment disparities among states have been This chapter shows that all successful fairly small since World War II. developers support Deng's insight. But his The convergence of living standards in wisdom may have eluded leaders in the the United States has been assisted by the developing world, even the few lauded as willingness of workers to "pull up their visionaries, as later chapters in the Report roots" and relocate.6 But basic welfare indi- will show. For decades, "spatially balanced cators have converged even in countries growth" has been a mantra of policy mak- where such a willingness has been less evi- ers in many developing countries. It was an dent, because development has been accom- obsession of planners in the former Soviet panied by the spread of public services. Union (see box 2.5). And it has been the Take France and Germany. Even though objective of governments of various politi- Paris generates 28 percent of France's gross cal hues in the Arab Republic of Egypt, domestic product (GDP)7 using only 2 per- Brazil, India, Indonesia, Mexico, Nigeria, cent of its land, infant mortality rates in the the Russian Federation, South Africa, and country show little spatial variation. The other great developing nations. There has lagging area of Lorraine had the highest rate, even been a strong commitment to spatially 4.5 deaths per 1,000 live births in 2005, but balanced development in the economic this is not much higher than the national history of many developed countries. The average of 3.8.8 In Germany the lead- United Kingdom pursued it between the ing area of Hamburg--with an economic late 1920s and 1980s,1 and Canada did so density of 114 million of GDP per square between the late 1950s and late 1980s.2 But kilometer--enjoyed a GDP per capita more in these cases, even with the popularity than twice that of the northeastern lagging of these policies, Deng's insight remained area of Mecklenburg-Vorpommern and an valid. economic density more than one hundred Indeed, the concentration of economic times higher. Despite the phenomenal dif- activity and the convergence of living ferences in economic density between these 73 74 WORLD DEVELOPMENT REPORT 2009 areas, there is no difference in basic welfare. country industrializes, it concentrates The numbers of physicians and hospital its limited initial human and physical beds per 1,000 habitants in both Hamburg capital in leading areas, those with high and Mecklenburg-Vorpommern closely growth potential. Areas distant from track the national averages.9 the new density lag. Spatial disparities This chapter presents stylized facts about in productivity and income can persist economic concentration in parts of a country, for generations, even with mobile labor usually called "leading areas," and the con- and capital. History points to persistent vergence in living standards between house- spatial divergence in living standards in holds in these areas and those in distant or today's developed countries in their ear- disconnected parts, called "lagging areas," in lier stages of development, followed by the same country. It introduces the concept of slow convergence many years after they economic distance, which is related to but not attained high income.10 the same as physical distance. When supple- · Technological progress and globaliza- mented with the economic density discussed tion have increased market potential in chapter 1, distance helps characterize the in the leading areas of developing coun- spatial transformations that accompany tries, intensifying concentration and development and that may be necessary for amplifying spatial disparities. Although rapid economic growth. the basic forces shaping the internal eco- The main findings: nomic geography of developing countries · As countries develop and integrate inter- are the same as those that earlier shaped nally, location matters more for eco- the economic landscapes of today's devel- nomic activity but less for social welfare. oped countries, the magnitudes have Greater economic mass (which accumu- changed. Larger international markets, lates where firms carry out production) better transportation, and improved and higher living standards (reflected in communication technologies mean that household consumption, poverty, and leading areas in open developing coun- access to basic services) are not spatially tries have greater market potential than synonymous. During the early phases of industrial countries did in their early development, infrastructure and social development. So the forces for spatial services tend to be confined to areas of divergence between leading and lagging economic mass. But as countries develop areas are now stronger. and integrate internally, the distinc- tion between leading and lagging areas Defining distance becomes sharper for economic mass and Density, discussed in chapter 1, is also rele- more blurred for living standards. vant at the country level. Denser concentra- · The spatial concentration of economic tions of economic activity increase choice activity first rises and then levels off. and opportunity. They ensure greater mar- As an economy changes from agrar- ket potential for the exchange of goods, ian to industrial, the spatial distribu- services, information, and factors of pro- tion of people and economic production duction. This chapter examines the dispari- becomes more compact. Within a coun- ties in economic mass and welfare between try, agglomeration and city-periphery areas within countries, linking these dis- integration give rise to metropolitan parities to the distance from economic den- areas and leading areas of dense eco- sity. So while chapter 1 discussed changes nomic mass. This process eventually at the local scale--where the most relevant levels off, and the spatial distribution of spatial dimension is density--this chapter economic activity stabilizes. addresses the spatial transformations at the · Spatial disparities in living standards country scale, where both density and dis- follow an inverted-U path, widening tance are relevant. Chapter 3 will propose in the early stages of economic devel- that although density and distance also opment, and remaining high for a long matter for world regions, the most impor- period before slowly converging. As a tant dimension at the international scale is Distance 75 division--political barriers to the flows of well connected to markets in dense settle- goods, entrepreneurship, people, and infor- ments. But people in many parts of India mation between nations. have difficulty getting to markets because of the travel time, determined by the type As the crow flies? Distance as an and quality of roads and other transport economic, not Euclidean, concept infrastructure (see map 2.1). Distance refers to the ease or difficulty for For labor mobility, distance also captures goods, services, labor, capital, information, the "psychic costs" of separation from famil- and ideas to traverse space. It measures how iar territory. Between 1985 and 1995, the easily capital flows, labor moves, goods are share of migrants in a Chinese province orig- transported, and services are delivered inating from another province fell as distance between two locations. Distance, in this between the provinces increased. And addi- sense, is an economic concept, not just a tional costs exist for migration between non- physical one. Although economic distance neighboring provinces.11 So, as with trade, is generally related to Euclidean (straight- economic distance for migration is related to, line) distances between two locations and but not synonymous with, physical distance. the physical features of the geography sepa- In this Report, the destination of interest is a rating them, the relationship is not always location with the greatest economic density straightforward. One reason is that distance or highest market potential. Distance is thus for the exchange of goods is different from a metaphor for access to markets. that for the migration of people. Manmade barriers, including policies, For trade in goods and services, distance can also increase distance. Roadblocks and captures time and monetary costs. The local barricades--improvised "toll sta- placement and quality of transport infra- tions" for local police and others to extract structure and the availability of transport payments--are common for journeys by can dramatically affect the economic dis- road in many Sub-Saharan countries.12 tance between any two areas, even though And where local political autonomy is the Euclidean distance between them high, there may be territorial fragmenta- could be identical. Two villages may have tion as policies of protection are pursued the same straight-line distance to a city, at the local level. Map 2.2 shows the time but one could be near a national highway, to human settlements, assuming few or no the other on an unpaved rural road. Based manmade barriers. Distances can be long, on straight-line distance, most of India is even in high-income countries. Map 2.1 Access to markets is not a straight line a. Based on Euclidean distance b. Based on economic distance c. Roads and settlements Source: WDR 2009 team. Note: The lighter color represents greater access to places with economic mass. 76 WORLD DEVELOPMENT REPORT 2009 Map 2.2 Distances can be long even in the developed world Travel time to sizable settlements, by subnational administrative area Travel time in hours and days to the nearest city of 50,000 or more people 0­1 1­2 2­3 3­4 4­6 6­8 8­12 12­18 18­24 24­30 30­36 36­2d 2d­3d 3d­4d 4d­5d No data Contributed by Andrew Nelson; see Uchida and Nelson (2008) for this Report. Locations close to markets have as the spread of disease. The main determi- a natural advantage nant of the strength of these interactions is Provincial governments in 1980 in China distance. Waldo Tobler's First Law of Geog- heightened their administrative powers raphy states that "everything is related to under decentralization reforms. They used everything else, but near things are more these powers to protect local firms--raising related than distant things."15 Areas closer tariffs and imposing bans on shipments from to economic density have easier access to other provinces. Imports between provinces beneficial interactions and exchanges. fell from 50 percent of GDP to 38 percent In Indonesia better road connections between 1992 and 1997, while local absorp- shorten travel time and the distance to eco- tion of goods within provinces rose from 68 nomic centers, creating larger agglomer- percent to 72 percent. The magnitudes are ated areas. Because of good roads and easier similar to those for goods crossing the U.S.- access to markets, villages 60 kilometers Canada border and international borders in from the district center generate as much the European Union (EU).13 China's hukou manufacturing activity as the district cen- system of permanent household registra- ter itself, and the well-connected periphery tion--linking place of residence with access becomes part of the agglomerated area. But to consumer goods, employment oppor- in poorly connected peripheries, the density tunities, and social protection--similarly of economic activity falls off rapidly beyond reduced internal migration.14 25 kilometers from the center (figure 2.1). Distance to density affects spatial Spillovers from proximity to density show movements in goods, services, informa- up in both developed and developing coun- tion, knowledge, and people. Commuting, tries. In European manufacturing, an area's migration, telecommunication, informa- total factor productivity growth is positively tion flows, and shipments of goods connect and significantly related to the density of originating and receiving areas. Most spa- manufacturing production in neighboring tial interactions, such as learning and trade, areas. And faster demand growth in neigh- are beneficial. But some are harmful, such boring areas stimulates, through spillovers, Distance 77 Figure 2.1 Manufacturing activity in Indonesia flourishes in areas with shorter economic distance to density Faster travel speed, better market access Slower travel speed, poorer market access Density Density 0.06 0.06 All households Manufacturing Others 0.04 0.04 0.02 0.02 0.00 0.00 0 20 40 60 80 100 0 20 40 60 80 100 Distance to district center (km) Distance to district center (km) Source: Yamauchi and others, forthcoming. faster total factor productivity growth.16 there. Responding to these incentives, firms In Canada, North York and Waterloo are, and workers enlarge the market opportuni- thanks to proximity and local research ties available in the dense area. The result is universities, becoming an extended part of a circular and cumulative process of dense the Toronto information and communica- areas continually gaining workers and firms tion technology (ICT) hub. Firms closer to from less dense areas. In this process, migra- Toronto do better than those farther away.17 tion balances the distribution of population The phenomenon is repeated in emerg- against the spatial disparity in economic ing economies. When a network of high- density. Reducing distance-related costs or ways surrounding Jakarta was built in the spatial frictions increases movements of 1980s, many firms moved out of the center people, firms, and ideas--as well as those to save on land and congestion costs. But of goods and services--and thus brings less they stayed near the metropolitan region developed areas into the national system of to have access to the large market.18 Similar production. With trade, the mobility of peo- but less pronounced is the pattern in other ple is probably the most potent mechanism Indonesian agglomerations, where growth for integrating areas of low economic density has been strongest in peripheral areas sur- with markets of high density. But for inter- rounding megacities.19 In Brazil indus- nal migration to bring about a convergence tries moved out of greater São Paulo to the in living standards, large population move- lower-wage populated periphery. Following ments may be necessary over generations. the transport corridors, these industries Every year, approximately 40 million moved through São Paulo state and into people in the United States change resi- the neighboring state of Minas Gerais. In dences, and 8 million people change states.21 the Republic of Korea the early decentral- The reason for this mobility is that economic ization of manufacturing from Seoul was to production is concentrated in a few parts of peripheral locations within an hour's drive. the country, and accessing this economic Only in the 1990s did industries decentral- density generally means moving closer to it. ize to towns and rural areas.20 People moving to economically dense areas contribute to production and boost The natural way to reduce distance is their incomes. But they also increase com- for people to migrate petition among workers in dense areas, A leading area of dense economic activity, reducing it in less dense areas, and contrib- through its market opportunities, creates uting to the convergence of living standards incentives for firms and workers to move between low- and high-productivity areas. 78 WORLD DEVELOPMENT REPORT 2009 Among today's industrial countries, the criteria that correspond to fairly self- quickest convergence occurred between 1870 contained labor markets and zones of eco- and 1913, largely driven by the largest flows of nomic activity. But data on such functionally people from Europe to emerging markets in defined economic areas are hard to come Asia and the Americas. For Ireland between by.24 So subnational areas are more com- 1851 and 1908, mass outmigration contrib- monly defined by administrative or political uted at least a third to the catch-up in Irish boundaries. Such definitions can bias econo- real wages with those in the United States metric analysis (see box 2.1), but they have and Britain--by reducing competition in the advantage of corresponding to the areas the domestic labor market. The virtual ces- for defining and implementing subnational sation of catch-up or convergence among the policy. This chapter examines administra- industrial countries between the two world tively or politically defined areas based on wars was attributed largely to more restric- different data sources, ranging from national tive immigration policies.22,23 accounts and household surveys to terrestrial grid cells of 1° longitude by 1° latitude. Density in leading areas, distance for In this Report, leading areas have a high lagging areas economic density, and lagging areas have a Subnational areas, when compared, should long distance-to-density. An area is more ideally be defined according to economic likely to be lagging the farther it is from BOX 2.1 Defining an area: impossible or NUTS? Subnational policy analysis relies on data within-area di erences in employ- The gure below shows the density of for areas that range from small primary ment or poverty can be as large as economic activity for Germany's 16 prov- sampling units to districts, and to states between-area di erences. Any change inces (länder) and 439 districts (kreise). or provinces. Typically, these areas are in the boundaries between areas The highly aggregated data indicate that de ned administratively or politically, could change the results. The potential 30 percent of GDP is produced on 10 per- re ecting historical characteristics more implications are succinctly summa- cent of the country's area, and the more than current patterns. For instance, the rized by the title of a classic paper on disaggregated data show that almost 60 existing administrative structure of the this topic, "A Million or So Correlation percent of GDP is produced on the same EU's member states generally consists Coe cients."a 10 percent. Aggregate information can be of two levels, such as länder and kreise · Second, analytical ndings depend on useful, but be mindful of these biases. in Germany, regions and départements in the aggregation or spatial scale, the Source: WDR 2009 team. France, comunidades autónomas and pro- ecological fallacy of inferring charac- a. Openshaw and Taylor 1979. vincias in Spain, and regioni and provincie teristics of individuals from aggregate b. Using state level data for the United in Italy. The Nomenclature of Territorial data. The classic study by Robinson States, the study showed that the propor- Units for Statistics (NUTS) provides a single (1950) illustrates this problem.b A broader tion of foreign-born people is positively cor- uniform classi cation of territorial units related with the proportion literate in Eng- aggregation will yield smaller di erences lish, suggesting that native-born Americans for producing regional statistics for the between units of analysis--and lower were more likely to be illiterate. Analyzing EU. The rst two administrative levels variances. So, results can di er signi - the same relationship using individual data in most member states correspond to cantly depending on the size of the units. showed a negative correlation. NUTS 2 and NUTS 3. NUTS 1, a larger unit representing the major socioeconomic Different spatial scales yield different results because of an aggregation bias regions, often does not correspond to Provinces (la¨ nder), Germany Districts (kreise), Germany existing administrative units within mem- ber states. Cumulative GDP (%) Cumulative GDP (%) 100 100 Which spatial scale to use, or how best to de ne a subnational area, depends on 80 80 the issue and the information available. 60 60 But the choice can dramatically a ect the 40 40 conclusions drawn from studying social 20 20 and economic conditions across di erent 0 0 parts of a country--for two reasons. 0 50 100 0 50 100 · First, areas are not de ned keeping in Cumulative area (%) Cumulative area (%) mind the policy issues. For instance, Source: Estimates based on Nordhaus 2006. Distance 79 leading areas because greater distance-to- such as access to sanitation and electricity, density implies a lack of integration into the are not met. In developed countries, lagging economy of leading areas. It also implies areas are locations with poorer job pros- poorer access to the "thick" markets of capi- pects than leading areas, but no differences tal, labor, goods, services, and ideas, and the in basic welfare. So distance and market spillovers of knowledge and information they access, in this Report, capture a wide range provide. A lagging area is usually a remote of criteria that different countries use to part of the country with one or more of the define a lagging area (see box 2.2). following features: high poverty, low pro- It follows that distance-to-density is the ductivity and income, high unemployment, cause of low income per capita, labor pro- and stagnant growth, which are typically the ductivity, and real wages--and of the high criteria governments use to define lagging rates of poverty and unemployment. In the areas. United Kingdom, economic density in the In developing countries, lagging areas leading London and southeast areas pro- tend to be remote places where basic needs, duces a wage premium of 18 percent, which BOX 2.2 How developed and developing countries define lagging areas: a quick survey In this Report, a lagging area is de ned around 71 percent of funds under the Mexico is noteworthy not only because as a place distant from density. How does convergence objective. But, even in EU of the sophistication of the measure used this de nition compare with how policy regional policy, funding is available on to identify lagging areas, but also because makers in developing and developed more favorable (and complicated) terms of the sophisticated manner of de ning countries have, today and historically, for those areas whose GDP per capita areas. Rather than using crude administra- de ned lagging areas? is not only less than 75 percent of the tive boundaries to de ne areas, geograph- Usually, the criteria national govern- EU average, but which are in a country ical information system (GIS) techniques ments use to classify an area as "lagging," whose GDP per capita is less than 90 are used to consider an area's geographi- "disadvantaged," or "backward" are linked percent of the EU average. These areas cal proximity, ethnic and cultural identity, to explicit strategies or policies for spatial are considered to be "more lagging."c and geoeconomic characteristics. or regional development. The criteria · Precise and complicated. Between 1982 So the criteria that di erent countries might be vague or precise. They might and 1987 Canada's Department of use to identify lagging areas depend on relate to a single indicator of economic per- Regional Industrial Expansion used a the level of development and on domes- formance or to a weighted average of sev- development index to classify areas tic political considerations. High levels of eral. And they might re ect the de nition for allocations under its Industrial and poverty and marginalization de ne lag- of lagging areas at di erent spatial scales. Regional Development Program. The ging areas in developing countries, and a high rate of unemployment often de nes · Vague. UK regional policy in the 1980s index assigned a 50 percent weight to them in developed countries. classi ed a lagging area as being either an area's unemployment, a 40 percent India's 10th Five-Year Plan (2002­07) a "development area" or an "intermedi- weight to its personal income, and a 10 identi es the northeastern region as ate area." But the law was vague in the percent weight to the scal capacity of "backward" and "disadvantaged" and criteria it set to designate such areas. the province to identify 15 percent of thus deserving special policy attention. "In exercising his powers under the the "least developed."d EU regional policy, under its convergence preceding provisions of this section [in · Sophisticatedly de ned and measured. objective, makes special provisions for the designation of development and To identify areas considered as lag- "the outermost regions," deemed to intermediate areas] the Secretary of ging, Mexico's microregional strategy require additional assistance. State shall have regard to all the circum- uses a "marginalization index" based This Report's de nition of lagging stances actual and expected, including on indicators of access to such basic areas--as distant from density--captures the state of employment and unemploy- services as electricity and drinking this wide range of criteria. ment, population changes, migration water, and indicators of the quality of and objectives of regional policies."a dwelling conditions and the proportion · Precise and simple. EU regional or "cohe- of the local working population that Contributed by Mark Roberts. sion" policy for the period 2007­13 is poorly paid.e It is mainly targeted a. Industrial Development Act 1982, chapter de nes lagging areas as those qualify- at remote rural communities in the 52, part I, para. (3); bold emphasis added. ing for assistance under the "conver- south, because the "remoteness of b, c. http://europa.eu/pol/reg/index_ en.htm, "Activities of the European Union-- gence objective," equated with NUTS2 rural communities often translates into Regional Policy," 2008. areas with a GDP per capita of less conditions of poverty and a substantial d. Atkinson and Powers 1987. than 75 percent of the EU average.b lack of access to a wide range of basic e. Villarreal 2005; OECD 2003, p. 6. These areas are budgeted to receive public services."f f. OECD 2003. 80 WORLD DEVELOPMENT REPORT 2009 distant areas in the north and southwest of have lower wages, and improving an area's England and in Scotland and Wales do not growth prospects largely depends on reduc- enjoy.25 In Indonesia the potential profit- ing distance.28 In Brazil's leading area, eco- ability of firms in textiles and other sectors nomic density implies a wage premium of is negatively related to distance-to-density: 13 percent, comparable to that in European more distance, less profit. This is true for countries.29,30 In Mexico the southern rural distance-to-density within the country and areas--distant from the economic density for distance to an international port and in Mexico City and the United States--have thus to the density in international mar- the lowest wages and highest poverty. kets.26 Again, lagging areas unable to attract Lagging areas in many countries are investment and employment are those with home to ethnic minorities. Tribal, racial, a high distance-to-density. and religious differences in access to As in today's rich countries, distance-to- resources show up as spatial disparities. In density affects incomes in emerging mar- a vicious cycle, disparities between areas ket countries. In China good market access that coincide with different ethnic groups produces higher individual wages, even can deepen political divisions and fuel after controlling for individual, sector-, and tensions, contributing to greater diver- province-specific attributes, living cost dif- gence in living standards. They can even ferences, and human capital externalities.27 fuel civil conflict that is difficult to extin- In Brazil lagging areas economically distant guish, causing "development in reverse" from São Paulo and other large markets (see box 2.3).31 BOX 2.3 Dangerous disparities: when divisions aggravate distance The academic literature argues that inter- Sikkim, and Tripura make up the lagging Economics Research at the United Nations nal labor migration is the strongest force northeast. Except for the Assamese, University in Helsinki (UNU-WIDER) conjec- for convergence in economic and other the population is predominantly tribal, tured that "Spatial inequality is a dimen- measures of household welfare across speaks Tibeto-Burman and Austro-Asiatic sion of inequality overall, but it has added areas of a country. But di erences in lan- languages, and has a strong genetic simi- signi cance when spatial and regional guage, religion, ethnicity, and race are larity with the people of East Asia. Hin- divisions align with political and ethnic probably one of the strongest barriers to duism is the dominant religion, but the tensions to undermine social and political internal migration, a troubling dilemma for proliferation of Christianity has set the stability."a These somewhat abstract words policy makers. The ethnic, linguistic, and area apart from the rest of India. By con- chillingly foreshadowed the violence in religious barriers that may keep house- ventional measures of economic welfare Kenya in early 2008, which left 1,500 peo- holds from taking advantage of many and development, northeastern states ple dead and another 250,000 displaced. opportunities to arbitrage geographic rank among the lowest in India. Violence began over the disputed outcome di erences for employment and earnings Disparities in Africa. A study of 11 of a presidential election in late December can be the same barriers that cage poor Sub-Saharan countries found that ethnic- 2007, quickly exposing deep ethnic cleav- people in lagging areas, perpetuate their ity was on its own a strong predictor of ages that demarcate Kenya's economic and poverty, and sharpen spatial disparities. di erences in child mortality, but when political geography. Communal ghting Disparities in East Asia. In Thailand combined with geography, it continued was most pronounced around the town 17 percent of people in the northeast to predict the probability of survival of El Doret in the Rift Valley, and on the are poor, compared with 0.5 percent in among children. For instance, in Côte outskirts of Kisumu in the Western district Bangkok. About half of Thailand's ethnic d'Ivoire, mortality among two year olds of the country. The Rift Valley and Western minority groups live in the Northeast. In fell much faster from 1970 to 1994 for the districts are among Kenya's economically Indonesia poverty and welfare indicators Baoule than for other ethnic groups. Chil- lagging areas and are the traditional home are persistently worse in West Kaliman- dren of Ashanti women in Ghana were places of the minority Kalenjin, Luo, Kisi, tan--home to such ethnic minorities as about 20 percent less likely to die than and Luhya tribes, who along with other the Dayak, Bugis, and Sambas--than in other children. In Uganda, Baganda chil- ethnic minorities in these areas harbor Java, home to Indonesia's ethnic majority. dren under ve were a third less likely to resentments related to economic depriva- Disparities in South Asia. In India die than children of other ethnic groups. tion and neglect. the states of Arunachal Pradesh, Assam, A 2005 study on spatial inequalities Source: Brockerho and Hewett 2000. Manipur, Meghalaya, Mizoram, Nagaland, by the World Institute for Development a. Kanbur and Venables 2005. Distance 81 Lagging areas have higher poverty rates, Economic concentration in leading leading areas have more poor people areas The rate of poverty (the poverty head- As economies develop, economic activity gen- count) is related to distance, and the mass erally becomes more concentrated, not less. of poverty is related to density. Lagging In about a quarter of the world's nations-- areas tend to have a higher proportion of such as Botswana, Brazil, Norway, Russia, and poor residents, and the leading areas tend Thailand--morethanhalfofnationalincome to contain a higher share of the country's is generated on less than 5 percent of the land poor people, because of the dense popu- area. In half of all nations--such as Argen- lation in leading areas. Vietnam's lagging tina, Saudi Arabia, Slovenia, and Zambia--a inland areas have the highest poverty rate, third or more of national income is generated but its prosperous leading areas contain on less than 5 percent of land. Only one coun- the mass of poor people (see map 2.3). try in 10 has a dispersed economic mass, with And in Honduras the country's poverty less than a tenth of national income generated mass is concentrated in its two leading on 5 percent of its land. Among the few coun- areas of Tegucigalpa and San Pedro Sula, tries with this high spatial dispersion: Bangla- while distant eastern areas generally have desh, the Democratic Republic of Korea, the a high poverty rate (map 2.4). Netherlands, and Poland.32 Map 2.3 Vietnam's poverty rate is higher in lagging inland areas, but its poverty mass is greater in leading coastal areas Hanoi Hanoi Poverty rate: Poverty density proportion of poor (%) 2.5­68.6 3­18 68.6­145.8 18­36 145.8­245.9 36­41 245.9­410.6 41­48 410.6­2757.3 48­94 Thanh Pho Thanh Pho Ho Chi Minh Ho Chi Minh VIETNAM Source: The Poverty Mapping Project. Columbia University, using data from Minot, Baulch, and Epprecht 2003. 82 WORLD DEVELOPMENT REPORT 2009 Map 2.4 The poverty rate is high in distant eastern Honduras, but the poor are concentrated in the two largest metropolitan areas HONDURAS San Pedro Sula Poverty rate: Poverty density Tegucigalpa proportion of poor (%) Tegucigalpa 2.50­35.00 42­57 35.00­59.19 57­72 59.19­96.63 72­76 96.63­318.09 76­84 318.09­342.72 84­97 No data No data Source: The Poverty Mapping Project. Columbia University, using data from Robles 2003. This section presents the historical expe- not as dramatic, but both countries expe- rience of selected industrialized countries. rienced the same pattern of rapidly rising Spanning more than a century, this section concentrations at low levels of development, shows how these countries experienced followed by a leveling off as GDP per capita rapidly rising spatial concentrations, fol- rose past $10,000 (see figure 2.2).35 lowed by a leveling off. It then turns to a Patterns are similar in today's developing large sample of developed and developing countries. As Thailand industrialized and countries to document how the concentra- grew rapidly, the concentration in the leading tion of economic mass rises with a coun- Bangkok metropolitan area increased from try's development. 1.8 in 1975 to 3.1 in 2004, while GDP per cap- ita increased fourfold. In Brazil too, the con- Rapidly rising concentration in the early centration in the leading São Paulo area edged stages of development, then a leveling upward from 7.3 in 1960 to 8.4 in 2004, as the It is difficult to come by data that track the country's GDP per capita almost tripled. evolution of spatial concentrations of eco- For Japan during its post­World War II nomic activity.33 The information available industrialization, the concentration in its reveals that economic development, in its leading area of greater Tokyo increased from a early stages, is accompanied by a rapidly rising highof7.1in1955toabout8in1970asitsGDP spatial concentration in a country. Not only per capita more than doubled. This increasing does the volume of economic activity grow, spatial concentration eventually levels off, as but its generation becomes more compressed the spatial distribution of economic activity into a smaller land area. Leading areas benefit in a country stabilizes. After 1970, the con- most from this compression and growth. centration in greater Tokyo stabilized. Economic concentration in the Ile de In the United States as GDP per capita France--the leading area of France, with rapidly increased from $1,806 in 1850 to about 2 percent of the country's land--in- $4,091 in 1900,36 concentration came in creased rapidly from a value of around two the manufacturing belt of Green Bay­St. times the hypothetical share in 1801 to three Louis­Baltimore­Portland ME, which times in 1851 and to six times by 1910.34 It accounted for three-quarters of U.S. manu- continued to rise, but less rapidly, to nine facturing employment. Over the next 60 times that share in 1960. French GDP per years, the belt's share of manufacturing capita grew from less than $1,000 in 1801 employment remained stable at two-thirds to $7,000 in 1960. From 1960 on, however, to three-quarters.37 Despite structural its economic concentration stabilized, even changes in the U.S. economy and shifting though its GDP per capita tripled. In Can- patterns of economic concentration, that ada and the Netherlands the increases were concentration remained stable after 1960. Distance 83 Figure 2.2 Rising density of economic mass accompanies development over decades, even centuries a. From low-income to middle-income b. From low-income to high-income Concentration Concentration Index 10 10 9 9 8 8 7 7 6 6 5 5 4 4 3 3 2 2 1 1 0 0 0 2 4 6 8 10 2 10 20 30 40 GDP per capita (constant 2000 US$, thousands) GDP per capita (constant 2000 US$, thousands) Indonesia, 1989­2005 Spain, 1850­1960 France, 1801­1999 Canada, 1890­2006 Thailand, 1975­2004 Brazil, 1960­2004 Spain, 1850­2000 Japan, 1900­2000 France, 1801­1963 Netherlands, 1850­1960 Netherlands, 1850­2006 Philippines, 1980­2005 Chile, 1976­2004 Sources: WDR team estimates based on national accounts--statistical yearbooks of various years in respective countries. 1890 data for Canada come from Green (1969). Data on France are based on population numbers from Catin and Van Huffel (2003); Barro and Sala-I-Martin (2004). Data on Japan, the Netherlands, and Spain came from the Staff City Population Database, Human Settlements Group, International Institute for Environment and Development (IIED). Another corroborative piece of evidence of administrative areas, which may be of dif- rising concentration comes from the falling ferent geographic sizes. But controlling for share of land area occupied by 80 percent of these factors, a comparison of 24 develop- the U.S. population in the densest counties ing countries--ranging from Mozambique from 25 percent of the U.S. land areas in with a GDP per capita of $211 to Greece 1900 to 17 percent in 2000.38 with more than $12,000--reveals the same As countries grow beyond $10,000 GDP pattern as the historical experiences of per capita, concentration tends to stabilize, Canada and France. The share of national with the details differing. The concentra- GDP produced in the leading administra- tion in the leading area is greater in Canada, tive area tends to increase with the level of France, and Japan than it is in the Nether- development (see figure 2.3, panel a). lands and the United States. For develop- Statistical areas. Statistical areas, broad ing countries too, Brazil, Indonesia, and census regions, can differ from administra- the Philippines seem to be on paths toward tive areas. The United States has nine sta- greater spatial concentration than either tistical areas but 50 states; Canada has five Chile or Thailand. statistical areas but 10 provinces and three territories. A country's statistical office gen- International comparisons of erally uses these areas to stratify its sam- concentration today support pling frame for household surveys, with the historical trends areas corresponding to the geographic par- The relationship between a country's devel- titions of a country such as east and west.39 opment and its spatial concentration holds Despite the difference in aggregation, the for countries at different levels of develop- data for statistical areas suggest the same ment. It holds for countries based on admin- relationship between concentration, mea- istrative areas (Canadian provinces, Japanese sured by consumption rather than GDP, prefectures, Russian oblasts, and U.S. states), and development (see figure 2.3, panel b). statistical areas (the nine census regions of Land areas. Terrestrial grid cells of 1° the United States, the three regions in Ecua- longitude by 1° latitude, each correspond- dor), and land areas (terrestrial grid cells of ing to a land area of 100 square kilometers 1° longitude by 1° latitude). And it holds for can provide purer geographic resolution.40 different measures of concentration. Spatial concentration within a country can Administrative areas. Different then be measured as the share of national countries have different numbers of GDP generated on the densest 5 percent of 84 WORLD DEVELOPMENT REPORT 2009 Figure 2.3 Measures based on national accounts, household surveys, and geoscaled economic data confirm the historical pattern of a rising concentration of economic mass with the level of development a. Administrative area b. Statistical area c. Land area Province, state, prefecture Broad census regions 1° longitude × 1° latitude Highest provincial share of GDP Highest regional share of total household consumption % GDP confined in 5% of land 1.0 1.0 1.0 0.9 0.9 0.9 0.8 0.8 0.8 0.7 0.7 0.7 0.6 0.6 0.6 0.5 0.5 0.5 0.4 0.4 0.4 0.3 0.3 0.3 0.2 0.2 0.2 0.1 0.1 0.1 0.0 0 ­2 3 8 13 0 1 2 3 4 5 6 7 8 0.0 0 5 10 15 20 25 30 35 GDP per capita (2000 US$, thousands) GDP per capita (2000 US$, thousands) GDP per capita (2000 US$, thousands) Source: Panel a: National accounts at national statistical office Web sites or Yearbooks; panel b: World Bank staff estimates of more than 120 household surveys in 75 countries (data set is described in detail in Montenegro and Hirn 2008); panel c: World Bank staff estimates from http://gecon.edu.yale. its land.41 The stylized pattern of rising con- middle-income country) and New Zealand centration of GDP with development using (a high-income country). Poland and New historical data is the same as that using con- Zealand have lower spatial Gini coefficients temporary data. The relationship between than richer Norway and the United States. development and economic concentration is The pattern also holds for small and large positive and roughly linear when comparing countries. developing countries with a GDP per capita of less than $10,000. But this relationship starts Divergence, then convergence-- to level off when higher-income countries are between leading and lagging areas included in the sample (figure 2.3, panel c). When production is primarily agrarian, The rising concentration of production economic activity tends to be evenly distrib- with economic development is not an artifact uted across space. Productivity differences of the number of subnational areas across are also moderate, varying naturally with countries or of the different sizes of land area soil quality and climate. But as an economy in the countries (see table 2.1). Consider Tan- develops and production expands in manu- zania, Italy, France, and Sweden, with similar facturing and services, some areas become numbers of administrative areas (21 or 22). more attractive to firms and workers. Some Tanzania's leading area of Dar-es-Salaam are endowed with natural or "first nature" generates 15 percent of national GDP, Italy's geographic advantages.43 For example, a leading area of Lombardia, 21 percent. France strategic coastal location makes an area a and Sweden, each with a higher GDP per cap- natural choice for a port (as with New York ita than Italy, also have higher concentrations and Philadelphia in the United States). For in their leading areas. others areas not so blessed by nature, their For a set of countries partitioned into economic pull might be linked to a "second five statistical areas--ranging from Argen- nature" historical accident. An example is tina to Tajikistan--the concentration of Boston, saved from economic decline by an consumption in the leading area increases influx of immigrant labor fleeing the Irish with development. Among medium-size potato famine. For Irish immigrants it was countries with about 300,000 square kilo- cheaper to travel from Liverpool to Boston meters of land area, Ghana and Lao People's than to New York. Democratic Republic (both low-income Economic development brings with it countries) have markedly lower spatial greater market integration, which facilitates GDP concentrations measured by spatial the mobility of people and capital and allows Gini coefficients42 than Poland (a lower- for greater trade, forces benefiting the leading Distance 85 Table 2.1 Administrative, statistical, and geographic area measures all point to rising spatial concentrations of economic activity with development Administrative areas Country GDP per capita Number of administrative areas Share of GDP in the leading area (%) Tanzania 324 21 15 Italy 19,480 21 21 France 22,548 22 29 Sweden 31,197 22 29 Share of household consumption Statistical areas Country GDP per capita Number of statistical areas in the leading area (%) Tajikistan 204 5 30.2 Mongolia 406 5 34.6 El Salvador 1,993 5 43.9 Brazil 3,597 5 51.6 Argentina 7,488 5 64.7 Land areas Country GDP per capita Land area (km2) Spatial Gini coefficient Ghana 211 227,540 0.48 Lao PDR 231 230,800 0.48 Poland 3,099 311,888 0.52 New Zealand 11,552 267,990 0.55 Norway 27,301 304,280 0.64 Sources: Administrative area information for Tanzania is from http://www.nbs.go.tz/nationalaccount/index.htm; information for France, Italy, and Sweden are from the Annex in Growing Regions, Growing Europe. Statistical area information is from more than 120 household surveys fielded during the 2000s for more than 80 countries (data set described in detail in Montenegro and Hirn 2008). Land area information is from http://gecon.edu.yale, which is based on 1990 information. Note: GDP per capita estimates are in 2000 U.S. dollars for the particular year of the household survey. areas. And by attracting people and firms, Across areas of the United Kingdom, the leading areas fuel agglomeration economies, coefficient of variation of GDP per capita becoming centers for innovation and growth increased by almost 40 percent between 1871 and driving the national economy. But the and 1911.45 During this period, Britain went process does not go on forever. Agglomera- from a modern-day Namibia to a Jordan or tion economies start to be offset by conges- the former Yugoslavia.46 After World War II, tion and pollution, the diseconomies of GDP per capita across areas of the United agglomeration. So the spatial concentration Kingdom displayed a slow convergence, in leading areas starts to level off. continuing until the late 1970s, when spatial What, then, of the income and welfare inequalities stabilized.47 disparities that accompany this pattern of In the United States, the dispersion of per first rising and then stable economic con- capita income across states increased between centration? Is there a tendency for lagging 1840 and 1880, coinciding with the rise of the areas to catch up with leading ones as eco- manufacturingbeltintheNorth,andtheCivil nomic development progresses? What is the War and its aftermath. The end of the Civil role of government policies in facilitating War marked the beginning of integration this convergence? between states in the North and the South, and spatial dispersion in per capita income For today's developed countries, spatial began to narrow. Because the southern states inequalities in income and welfare rose remainedmoredependentonagriculture,lag- early, followed by slow convergence ging areas of the United States suffered a set- In today's developed countries, per capita back in the 1920s because of a sharp drop in incomes initially diverged between sub- the relative prices of agricultural goods. Once national areas, and convergence began to set this shock dissipated, the slow convergence in as GDPs per capita approached $10,000, between lagging and leading areas resumed following an inverted-U relationship (see with few interruptions until the 1990s, when figures 2.4 and 2.5 and table 2.2).44 disparities among states stabilized.48 86 WORLD DEVELOPMENT REPORT 2009 Figure 2.4 Spatial inequality rose and remained high before slowly declining as economies Figure 2.5 Subnational disparities in income and approached $10,000 in GDP per capita wages persisted for more than 70 years in Canada and France Coefficient of variation of area wages or income 0.6 Area inequality Index 2.5 Sweden, 1920­61 0.5 Spain, Canada 1860­1975 2.0 0.4 United States, 1840­1960 1.5 0.3 1.0 Japan, 1955­83 0.2 Habsburg Empire, 1756­1910 0.5 France 0.1 United Kingdom, 1871­1955 0 1855 1880 1905 1930 1955 0 Year 0 2 4 6 8 10 12 14 16 GDP per capita (constant int'l Geary-Khamis $, thousands) Sources: Canada: Green 1969; France: Williamson 1965. Note: Canada data are based on provincial per capita gross Sources: United States: Williamson 1965; Habsburg Empire: Good 1986; Sweden: Williamson 1965; Spain: value added; France data are based on department agricul- Martinez-Galarraga 2007; Japan: Mutlu 1991. tural wages. Table 2.2 Spatial inequality varied through different phases of development Phase of economic development Country Spatial disparity measure Early Middle Advanced United States 1774 1790 1840 1860 Relative deviation regional GDP per capita from U.S. average 30 31 56 66 Italy 1861 1911 1936 1951 Index of regional percent agriculture labor force 6.55 9.41 12.7 14.2 Canada 1901 1911 1941 1951 Index of regional percent agriculture labor force 7.14 9.88 12.6 10.2 England 1767 1795 1867­70 1898­1914 Maximum-minimum in county agriculture wages 3s 11d 8s 2d 11s 0d 7s 4d Austria 1869 1890 1910 Maximum-minimum regional percent agriculture labor force 0.32 0.35 0.40 Spain 1860 1914 1955 1975 Maximum-minimum ratio regional GDP per capita 1.76 2.33 2.22 1.74 Australasia 1860 1880 1900 Coefficient of variation regional GDP per capita 0.30 0.35 0.10 Sources: United States: Good 1986; Italy: Williamson 1965; Canada: Williamson 1965; England: Hunt 1986; Austria: Good 1986; Spain: Martinez-Galarraga 2007; Australasia (Australia, New Zealand, and Tasmania): Cashin (1995). Note: For Spain, the maximum is the top five and the minimum is the bottom five. For England, the currency is in shillings (s) and pence (d). Canada and France also exhibit the over to 1929 and starting to fall by 1956.49 In same inverted-U-shaped pattern of ris- Italy, Germany, and Spain, the convergence ing spatial disparities in the early stages in per capita income gradually set in many of development--spanning two genera- years after these economies reached high tions--followed by slow convergence (see income--after World War II--followed by figure 2.5). In France the spatial dispersion stable income disparities (see figure 2.6). of wages across départements increased Government policies can facilitate this between 1855 and 1900, when convergence convergence. In Japan, for example, invest- set in. In Canada the spatial dispersion of ments in social services in lagging areas average gross value added between areas were increased as concentration of eco- increased between 1890 and 1910, carrying nomic production accelerated. By making Distance 87 BOX 2.4 Correcting geographic disparities in postwar Japan In 1970, Prime Minister Eisaku Sato and instrumental in mobilizing nancial welfare in less developed areas. These the Cabinet initiated the New Economic resources. The general account budget policies were e ective in corralling large and Social Development Plan and the of the central government provided investments toward achieving universal New Integrated Spatial Development Plan earmarked budget transfers to local gov- attainment of basic living standards. Per (Shin-Zenso). The objective was to address ernments in addition to nonearmarked capita income converged between lead- disparities in living standards, as a result transfers. Among the earmarked budget ing and other areas during the 1970s (see of accelerated growth in industrial areas transfers, a substantial amount was allo- the gure on the right, below). Labor around Tokyo, Nagoya, and Osaka along cated for investments in basic services migration from rural to large urban areas the Paci c Coast during the early postwar (for example, rural roads) and social insti- was pronounced throughout the 1950s years. An excerpt in the Shin-Zenso sum- tutions under cost-sharing arrangements and 1960s, but it tapered o after the marized the government's vision: with the local government. mid-1970s. The Fiscal Investment and Loan Pro- Among many problems concerning spatial Sources: Cabinet Council 1972; Hayashi 2003; disparities, disparities in living standards are gram pooled public funds from such Kamada, Okuno, and Futagami 1998; Minis- more serious than those in per capita income. sources as postal savings and public try of Finance 2008; Nakajima 1982; Okuma From this standpoint, the construction of the pension insurance premiums and then 1980; Overseas Economic Cooperation Fund basic services and social institutions must be channeled them for investments in hous- accelerated in rural towns, and new policies 1995; Policy Research Institute for Land must be adopted to improve the living condi- ing and social institutions to improve 2001; Sakamaki 2006. tions of their surrounding areas above a certain Rising investments in social services facilitate convergence in incomes minimum level. Per capita spending Relative per capita income These plans continued to provide pub- Yen (thousands) Index (average = 100) lic investment in basic services and social 120 160 institutions (for example, public utilities, 150 medical facilities, and school buildings) 100 140 Tokyo & Osaka to industrialized areas. But additional Tokyo & Osaka investments were made in the less devel- 80 130 Pacific Ocean Belt oped areas, to achieve at least a minimal 120 60 level of living standards for all places. The 110 Pacific Ocean Belt result was a rapid catch-up in investment 40 Less developed 100 areas in basic services and social institutions in 90 20 less developed areas relative to the more 80 Less developed areas industrialized areas (see the gure imme- 0 70 diately to the right). 1966 1968 1970 1972 1974 1976 1978 1980 1955 1960 1965 1970 1975 1980 Both the general account budget of Year Year the central government and the Fiscal Contributed by Keijiro Otsuka and Megumi Investment and Loan Program were Muto. the labor force more mobile, this led to Figure 2.6 Spatial disparities have narrowed slowly falling geographic disparities in incomes in Europe since World War II (box 2.4). Coefficient of variation 0.45 For developing countries, spatial Italy 0.40 disparities in living standards between Spain 0.35 subnational areas first rise and then fall 0.30 with development 0.25 Germany Comparing a large number of countries at 0.20 different levels of development reveals that 0.15 France spatial disparities in per capita product 0.10 United Kingdom and welfare diminish with level of devel- 0.05 opment (see figure 2.7). This is consistent 1950 1955 1960 1965 1970 1975 1980 1985 with most developing countries being clus- Year Sources: de la Fuente 2000; Barro, Sala-I-Martin, Blanchard, tered on the upward-sloping section of the and Hall 1991. inverted-U-shaped relationship between development and spatial inequality-- and with the developed countries on the 88 WORLD DEVELOPMENT REPORT 2009 Figure 2.7 Contemporary comparisons of countries indicate that disparity in welfare among subnational areas fall with economic development a. Statistical area b. Land area Broad census regions 1° longitude × 1° latitude Max.-min. ratio of per capita household consumption Max.-min. ratio of GDP per capita 2.5 8 7 6 2.0 5 4 3 1.5 2 1 1.0 0 0 5 10 0 5 10 15 20 25 30 GDP per capita (2000 US$, thousands) GDP per capita (2000 US$, thousands) Source: Panel a: World Bank staff estimates of more than 120 household surveys fielded during 2000s in more than 80 countries; panel b: World Bank staff estimates from http://gecon.edu.yale, which is information in 1990. downward-sloping part of the relationship. to Argentina with more than $7,500. The The conclusion is based on two sources of second source is based on the geophysi- information. The first source comes from cally scaled economic data of terrestrial more than 120 household surveys cover- grid-cells of 1° longitude by 1° latitude for ing more than 80 developing countries, 90 countries that span the full spectrum of from the Democratic Republic of Congo development, from Ethiopia with a GDP with a GDP per capita of less than $100, per capita of less than $200, to Japan with a GDP of more than $30,000.50 The household survey data offer an Table 2.3 Household survey and subnational gross product data corroborate the pattern of declining spatial disparities in welfare with development added advantage because individual house- hold consumption is a better measure of Number of Leading-lagging area disparity Statistical GDP per statistical in household consumption welfare than income. Similar households in area Country capita areas (minimum-maximum ratio) different areas of a developing country can Cambodia 234 5 1.89 have an average gap in household consump- tion of 70 percent simply as a result of loca- Bangladesh 286 5 1.73 tion.51 In Nicaragua, a six-person household Colombia 1,989 5 1.54 headed by a primary-educated 40-year-old Thailand 2,109 5 1.52 male in the lagging area of Matagalpa- Argentina 7,489 5 1.48 Jinotega consumes half of what an equiva- Canada 23,392 5 1.22 lent household consumes in the leading area of Managua. In Canada and the United Leading-lagging area disparity Land GDP per Land area in per capita gross product States a household in the lowest GDP per area Country capita (km2) (minimum-maximum ratio) capita area consumes 20 percent less than Philippines 920 300,000 5.43 an equivalent household in the highest. In Poland 3,099 311,888 4.63 Japan the area of residence means even less for the gap in consumption. New Zealand 11,552 267,990 3.35 As countries become more developed, Norway 27,301 304,280 1.78 the disparities in welfare purely attributable Japan 33,280 364,600 0.35 to location diminish.52 This pattern holds Sources: Estimates of consumption disparity are from more than 120 household surveys fielded during after controlling for the land area of a coun- the 2000s for more than 80 countries. Estimates of disparity in gross product are from, which comes from try and its number of administrative areas. information gathered in 1990. Note: GDP per capita estimates are based on constant 2000 U.S. dollars for the particular years of the surveys. Among countries partitioned into five areas, Distance 89 Bangladesh and Cambodia, both with GDP Figure 2.8 Economic growth in East Asia and Eastern Europe is faster than the world's growth per capita less than $300,53 had spatial gaps Growth rate (%) in consumption between their leading and 15 lagging areas of 89 percent and 73 percent, East Asia respectively. For Colombia and Thailand 10 (with GDPs per capita of approximately $2,000) the equivalent gaps are about 50 World percent. For Canada (with a GDP per capita 5 of $20,000) the gap is less than 25 percent. Latin America Among the medium-size countries, spatial disparities in welfare follow the same pat- 0 tern, falling across the spectrum from devel- oping to industrialized countries. The same ­5 is true for larger and smaller countries (see Eastern Europe & Central Asia table 2.3). ­100 Fast-growing countries see spatial 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 disparities in income widen Year Source: World Bank 2005e. East Asian growth has outstripped both the world economy and the growth of other developing regions. As they moved from Figure 2.9 Disparities in per capita gross product have been rising between leading and plan to market, Eastern European and lagging areas in Southeast Asia Central Asian countries have also grown faster than the world (see figure 2.8). As in Coefficient of variation 1.2 the early stages of development in today's 1.1 industrialized countries, development in East Asia, Central Asia, and Eastern Europe 1.0 Vietnam has brought widening gaps. In Southeast 0.9 Asia the disparities in incomes per capita 0.8 Thailand between leading and lagging areas has 0.7 grown wider (see figure 2.9). In China too, 0.6 the spatial dispersion in GDP per capita 0.5 Indonesia increased over the last decade (see figure 0.4 2.10). All this is consistent with the findings 1981 1983 1985 1987 1989 1991 1993 1995 1997 1999 2001 2003 2005 of the UNU-WIDER research program. Year In Eastern Europe and Central Asia, Source: Hamaguchi forthcoming. too, disparities among subnational areas in labor productivity and income wid- ened. In Russia income per capita in the Figure 2.10 Steady rise in inequality of per capita provincial gross product in China since 1990 lagging subnational area in 1985 was half the national average, and that in the lead- Coefficient of variation 0.70 ing area, twice the national average. Since then, income per capita in the lagging area 0.60 Including has fallen to a quarter of the national aver- municipalities age, while that in the leading area increased 0.50 to five times the national average.54 This divergence occurred during a reshaping 0.40 Excluding of Russia's economic geography as state municipalities industries in remote areas collapsed, and 0.30 economic activity started to respond to spa- 0.20 tial variations in market potential (see box 1956 1964 1972 1980 1988 1996 2.5). Similarly, the Czech Republic, Hun- Year gary, Poland, and the Slovak Republic have Source: Demurger and others 2002. 90 WORLD DEVELOPMENT REPORT 2009 BOX 2.5 Spatial inefficiency and the downfall of the Soviet Union The Earth hosts many vast and harsh east of the Belaya. Industrial Siberia grew found itself more resource abundant, spaces, but few governments have put in absolute terms, but its share did not but also less populated. The market as much energy into the development of exceed one- fth under the Soviet price revaluation of resources and assets such places as Russia did under the Soviet system that favored nal goods at the shrank the economic mass of distant government. expense of raw materials, transportation, zones and poles, but deteriorating infra- The e ort to develop Russia's eastern and energy (see the table below). structure did not reduce, and in some areas was substantially increased under The Soviet social infrastructure over- cases, increased economic distance. Stalin's rule. A forced industrialization lapped with industrial development. Industry-tied public services also col- attempted to shift production to the east Health centers, schools, recreational, lapsed in the 1990s, as rms were priva- and create new economic bases in the cultural, sports, and communal-housing tized or transferred to their sotscultbyt country's geographic heart. Equalization facilities--called sotscultbyt--generally to municipal authorities. For some time of economic (especially industrial) mass belonged to enterprises. This overlap under Yeltsin, the revenues of federal across Russia was seen as the way to was especially evident in large com- and regional/local budgets were o - make development uniform across space. panies in remote areas, such as the cially equal (50:50). In the 2000s, though, "Balanced industrial growth" remained transpolar city of Noril'sk. This tradition the rules were changed in favor of the a slogan for a long time. In the 1930s the was combined, somewhat paradoxi- Federation (60:40 when the external new areas received more than 50 per- cally, with a vigorous redistribution of debt payments were made, reduced cent of the central investment, nanced funds between sectoral and regional later to 55:45). But expenditures stayed mainly by expropriating wealth from departments. Pro ts were seized and at 50:50 because of growing transfers. agriculture. The new areas absorbed only then given back--not necessarily to the Today, center-region nancial rela- capital at rst. Visible e ects appeared same place--in capital goods and assets. tions are again based on the principle of during World War II, although the most The share of enterprises under the all- redistribution, though less so than in the productive zones were close to the front, Union jurisdiction reached 70 percent Soviet Union. But industry is now more like the Ural-Volga, where 58 percent of in the reigns of Stalin and Brezhnev. The fuel and material based. After decades factories evacuated from the west of the central government (Sovmin) controlled of equalization plans, the economy sees USSR were placed. less than 20 percent of industrial pro ts widening disparities in regional per capita An accounting of this centralized, direc- obtained on Russian soil. product. tive e ort to spread out economic mass Industrial deconcentration, together The gures on the next page show this is depressing. Alexei Mints, the Soviet with price system distortions and an for 1990 and 2005, using old Soviet net geographer, dismissed as propaganda the expensive arms race, would bring the material and new gross regional product claims that directed investment boosted Soviet system down. In the late 1980s (GRP) methods and prices. The two lead- backward areas and created cities "from both the elite and the masses in almost ers, Tyumen oblast in Western Siberia zero" under the ve-year plans. The real- every area or republic claimed that it bore and Moscow in the center, remained the ity was more prosaic: the "opening up" of the burdensome duty of a land that "fed same. But the gap between leading and eastern raw material elds coincided with the others." The slogan of regional khoz- lagging areas skyrocketed from 5 to 43. the growth of manufacturing in the west. raschet (self-repayment and economic With redistribution, the leading-lagging The shift eastward, Mints wrote, occurred accounting) soon grew into political sepa- gap in each area's average personal mostly in the European part.a In reality, ratism and contributed to the demise of income in 2005 was 11. Only 20 of 88 Russia's demographic and economic the Soviet Union. regions exceed the Russian average in per geocenter had moved only as far east After the Soviet Union collapsed, the capita GRP, and only 22 in income. Most as the river Belaya in Bashkiria by 1990; Russian Federation became more inte- poor areas reduced the gap in living stan- eight of Russia's 11 time zones lay to the grated with the world market. Russia dards with the help of transfers. Spatial shifts in the Russian Federation, 1900­2000 Indicator/region 1900 1925 1950 1975 2000 1900 1925 1950 1975 2000 Number of workers, millions Production, billion rubles, in 2000 prices Absolute figures 1.9 2.2 10.8 21.4 13.3 22 37 579 4,705 4,759 By type of regiona Percent Percent (in current prices) Old industrialb 64 61 42 40 33 50 65 68 42 32 New European 30 33 39 41 47 33 31 27 38 40 Eastern (Asiatic) 6 6 19 19 20 17 4 5 20 28 a. Author's calculations based on various statistical and literary sources. b. Includes St. Petersburg and suburbs, the center (including Nizhniy Novgorod) and the mid-Urals. Distance 91 BOX 2.5 Spatial inefficiency and the downfall of the Soviet Union--continued Differences in regional product widened Current prices, percent of Russian average Net material product, 1990 Gross regional product, 2005 11 macroregions, 73 regions 11 macroregions, 79 regions Index 550 500 450 400 350 Index 300 250 250 200 200 150 150 100 100 50 50 0 0 lga North East East Centre yatka Vo Urals Volga Urals Siberia SiberiaFar North Centre yatka Siberia SiberiaFar Caucasus Caucasus Kaliningrad Northwest Volgo-VC.Chernozem West East Kaliningrad Northwest Volgo-VC.Chernozem West East North North Note: Vertical lines show range of values within an area, and diamonds represent the area mean. Welfare in remote areas has become across space based on a wider sharing of largest land area, spatial policy choices less dependent on economic mass in con- oil and gas pro ts, or a forced diversi ca- and their e ciency could mean the dif- temporary Russia. The trend is not seen tion of regional economies based on mili- ference between economic progress and as satisfactory by some Russian observ- tary-industrial activities and research and stagnation. ers and policy makers, but what should development initiatives. While the debate be done about it is not clear. The policy continues, Russia's experience under the debate ranges between two polar visions: Soviet government o ers policy lessons. Contributed by Andrei Treyvish. reinforcing the redistributive system Particularly for a country with the world's a. Mints 1974, pp. 20­54. all witnessed increased spatial disparities Figure 2.11 Income disparities between areas widened as Eastern European nations moved across subnational areas since the begin- from plan to market ning of transition (see figure 2.11). Coefficient of variation The East Asian and Eastern European 0.050 countries appear to be on the rising part 0.040 of the inverted-U curve. Economic activity Czech Republic is still concentrating in a small number of 0.030 favored leading areas, with agglomeration Hungary Poland economies increasing their productivity, 0.020 wages, and income per capita. The lagging areas, insufficiently integrated into the 0.010 Slovak Republic national economy, have not yet captured spillovers from the leading areas. 0.000 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 The dynamics of geographic divergence Year in East Asia, Eastern Europe, and Central Source: Ezcurra and Pascual 2007. Asia have generally been a "race to the top." All subnational areas experienced gains in average wages and household incomes, though the biggest gains have gone to the 92 WORLD DEVELOPMENT REPORT 2009 leading areas.55 Among the poorest prov- As incomes diverge, health and inces in China, the southwest region had education converge GDP per capita growth of 7.7 percent over Many developing countries have had subna- 1979­98, the central region 7.8 percent, and tional Millennium Development Indicators the northwest region 8.4 percent.56 East across areas converge, so even though dis- Asian countries saw phenomenal declines in parities in income and material well-being poverty from more than 450 million poor widened, basic welfare has become more living on less than $1 a day in 1990 to about equal. In Indonesia the coefficient of varia- 120 million in 2007.57 For Eastern Europe tion across provinces for average years of and Central Asia, the divergence between schooling fell from 0.43 in 1971 to 0.15 in 1998 and 2003 was associated with a fall of 2000, and that for the poverty rate fell from 40 million in the region's poor living on less 0.42 to 0.35.63 In Thailand infant mortality than $2 a day, mainly because the mass of rates narrowed from a minimum-maximum poverty is in leading areas.58 gap of 6 percentage points between the lead- ing and lagging areas in 1980 to 0.7 percent- Some relatively closed or middle- age points in 2000,64 around a national income countries had incomes converge mean of six deaths per 1,000 live births. In upper-middle-income Brazil, the dis- In Vietnam the gap in malnutrition rates persion of state per capita income around between leading and lagging areas fell from the national mean fell from a coefficient of 20 percentage points in 1998 to 15 percent- variation of 0.65 in 1970 to 0.49 in 1995.59 age points in 2004, accompanying an overall Chile witnessed spatial convergence in improvement for all areas.65 In China terri- GDP per capita across subnational areas torial disparities in the human development between 1960 and 2001, when its GDP per index declined between 1995 and 2003. The capita more than doubled from $4,270 to disparity between the best-performing $10,538.60 Upper-middle-income South province (Beijing) and the worst-perform- Africa also had per capita incomes con- ing province (Tibet) declined from 0.26 in verge between its towns and cities from 1995 to 0.19 in 2003 for life expectancy, and 1990 to 2000.61 For Colombia, a relatively from 0.50 to 0.32 for the human develop- closed economy, the ratio of GDP per cap- ment index. The gap for literacy rates also ita in the leading departamento of Santafé declined between 1990 and 2003, from 58 to de Bogotá to the lagging departamento of 51 percentage points.66 The convergence of Choco fell from 10 to 6 during 1950­60 basic welfare in rapidly growing East Asian and to 3.1 in 1990.62 countries is epitomized by Malaysia (see fig- ure 2.12). Figure 2.12 In Malaysia, geographic convergence in basic welfare accompanied economic growth Diverging before converging poverty rate, 1970­2002 Converging to universal access to sanitation, 1987­2001 Households below the poverty line 80 Kedah Johor Sabah 1987 2001 Perak 70 Kelantan Kelantan Sarawak 60 Sarawak Pahang Melaka Perak 50 Pulau Pinang Perlis Terengganu Terengganu 40 Kuala Lumpur Selangor Negeri Sembilan 30 Pulau Pinang Perlis Melaka Pahang 20 N Sembilan Selangor Johar Kedah 10 Sabah Malaysia 0 Malaysia 1960 1970 1980 1990 2000 2010 40 60 80 100 Year Rural population with sanitary latrines (%) Source: Malaysia Economic Planning Unit 2008. Distance 93 For Mexican states, rates of adult literacy the benefits of development as a slow sub- and infant mortality converged from 1940 national convergence in living standards to 2002, as did life expectancy and enroll- sets in. This basic thesis holds true today. ment rates from 1990 to 2002.67 In Egypt But there are some important differences the gap in female primary school enroll- for modern-day developing countries: ment rates between the best- and worst- performing governorates narrowed from · Given the phenomenal size of today's 41 percentage points in 1995 to 25 in 2004, global market, development relies more as did the literacy rate and the gender gap in on pursuing an outward-oriented strat- literacy between 1986 and 2001.68 egy in which leading areas compete and Not all countries have experienced spa- trade globally. tial convergence in the Millennium Devel- · The rapid transformation of internal opment Indicators. Countries in South Asia economic geography--and the spatial and Africa still have wide internal dispari- disparities in today's developing coun- ties. In India and Sri Lanka the dispari- tries--will likely be greater than in ties across states remained large between industrial countries during their early 1981 and 1991,69,70 though there have been stages of development. absolute improvements both nationwide · Because redistributive mechanisms take and in the country's lagging areas. In Sri time to build and mature, labor mobility Lanka poverty was reduced in all provinces and market connectivity are more potent between 1991 and 2007, with the fastest mechanisms to integrate lagging areas reduction in its leading western province.71 into national economies. Globalization In Kenya provincial gaps in primary and technological progress in transpor- and secondary school enrollment rates tation and communication potentially remained large between 1999 and 2004, but provide a wider range of means to bridge more important, all areas made progress, the economic distance between leading including the lagging Northeast.72 and lagging areas. Global markets are more important. What's different for today's Because of greater integration today, global developers? markets are more important than domes- In The Wealth of Nations, published in tic markets than at any time in history. The 1776, Adam Smith wrote, "It is upon the market potential of leading areas is higher in sea coast, and along the banks of navigable today's developing countries than it was in rivers, that industry of every kind naturally today's developed countries during the nine- begins to sub-divide and improve itself, and teenth and early twentieth centuries, thanks it is frequently not till a long time after that to the rapid growth of trade since the end of those improvements extend themselves to the inland parts of the country."73 What Smith wrote in 1776 could apply equally to Figure 2.13 Today's developing countries face a more integrated world the spatial processes in China's modern eco- nomic development. What, if anything, is World exports as a % of global GDP different for today's developing countries? 30 In some fundamental respects, very little. Smith's key point was that a coun- 20 try's economic development, in its early stages, tends to be led by subnational areas that provide the greatest potential access 10 to markets and thus to density. But sub- national areas distant from density, inland 0 areas in Smith's example, tend to be left 1820 1870 1900 1960 1975 2004 behind. Only later in the development pro- Year cess do these lagging areas share more of Source: Chase-Dunn, Kawano, and Brewer 2000. 94 WORLD DEVELOPMENT REPORT 2009 World War II. Indeed, the growth of trade the United States also became important, has been about twice that of world income and border areas such as Ciudad Juarez, in recent decades.74 Trade as a proportion Mexicali-Calexico, Nogales, and Tijuana of world GDP is now more than 25 times had large increases in market potential and its level in 1820 (see figure 2.13). So devel- growth, whereas Mexico City had some opment under protectionist policies might depopulation and dispersion of its manu- have been a viable (if not optimal) strategy facturing activity.76 in the nineteenth and early twentieth cen- In China, during Mao's era of self- turies.75 But a protectionist strategy is much sufficiency, heavy industries were promoted less likely to be viable today, especially in in interior provinces, which received 71 the light of recent failures of such policies in percent of state investment between 1966 Latin America and Sub-Saharan Africa. and 1970. Many companies in Shanghai When a country is relatively closed, an and other coastal cities were relocated to area's market potential is determined mainly the interior and mountainous provinces of by its distance to density within the country. Guizhou, Hubei, and Sichuan.77 But since But once it is open, distance or access to inter- China has become more open to foreign national markets also becomes important, trade and investment, coastal areas flour- and border and coastal areas tend to gain ished as gateways to overseas markets, but in their shares of economic activity. Struc- many interior areas floundered. Export- tural shifts in patterns of trade can alter the oriented industries (garments, electron- topography of market potential in a country: ics, leather) are concentrated in coastal previously leading areas, perhaps favored by provinces, while domestic market-oriented policy, lose out and decline as their distance industries (metals, nonferrous smelting) to new leading areas increases. This is illus- are dispersed (see map 2.5).78 trated by Britain, China, and Mexico. The costs of transport and telecommu- Openness matters for distance. Before nications matter more. Sea coasts and navi- Mexico liberalized trade in 1985, the dis- gable rivers are natural locations for leading tance to Mexico City was the primary deter- areas because, in Smith's day, shipping was minant of an area's market potential. But the most cost-effective way of transport- with liberalization, distance to density in ing goods to domestic and international Map 2.5 Exporting industries concentrate in coastal areas to minimize distance to the global market a. International market-oriented industries b. Domestic market-oriented industries Electronics and Petroleum refining telecommunications and coking equipment Nonmetal mineral Instruments and meters products Garments and other Ferrous metal smelting fiber products and pressing Leather and fur products Nonferrous metal smelting and pressing Source: He forthcoming. Distance 95 markets. But technological progress has led Guizhou has a level ($1,653) close to the to large reductions in the cost of transport- British average in 1830.82 ing goods and in telecommunications (see Although comparisons between China chapter 6). New (non-water-based) modes and Britain need to be made with caution of transport and the information technol- because of the different geographic scales of ogy revolution have reshaped the landscape the two countries, the basic point remains. of economic density. When today's rich countries were develop- Access to knowledge is easier. So today's ing during the nineteenth and early twen- developing countries can take advantage of tieth centuries, the growth of their leading world markets of unprecedented size and areas was constrained to the rate of growth can access these markets with greater ease. of their domestic markets and the world At the same time, greater flows of foreign technological frontier. These constraints direct investment, expanding twice as fast limited the extent to which spatial dispari- as world trade, increase access to knowledge ties could increase in their early stages of at the world's technological frontier.79 For development. In sharp contrast, for today's the most successful developing countries developing countries, these constraints no (mainly in East Asia) of recent decades, the longer exist. Although the absence of these result has been national growth--driven constraints helps developing countries, the by leading areas--far faster than that of potential disparities that can arise between today's developed countries in the early leading and lagging areas in the early stages stages of their development. of development are much larger. With such rapid growth in leading Although the spatial inequality between areas, the geographic disparities in today's leading and lagging areas in today's devel- developing countries are far larger. Take oping countries will follow the same China, for example, whose GDP per capita inverted-U shaped path, the features of is roughly equivalent to that of Britain in this path will differ. The ascent is likely to 1911. London then had a GDP per capita be steeper in the initial stages of develop- around 1.7 times the national average, ment. Set against this faster rise in dispari- whereas East Anglia had a GDP per capita ties, however, is the opportunity for faster two-thirds that average.80 In China today, convergence between lagging and leading the comparable figures are 3.3 for the lead- areas as development progresses--because ing area of Shanghai and one-third for the modern information and communications lagging area of Guizhou.81 Shanghai has a technologies offer a wider range of methods GDP per capita ($16,044), roughly equiva- to bridge the economic distance between lent to the British average in 1988, while leading and lagging areas. CHAPTER 3 Division D ensityanddistance,thedimensions 600 land borders between nations (see fig- of economic geography examined ure 3.1).4 And their number may continue in the two previous chapters, mat- to increase if federated states split apart, ter for the development of countries and if minorities within nations achieve self- regions. Over the past two centuries, global determination, and if some of the remain- gross domestic product (GDP) has grown ing 70 dependencies seek independence.5 about 2.3 percent a year, an almost 50-fold This chapter shows how divisions affect increase in constant dollars.1 But growth has economic development, how geography not been uniform. Half of global GDP today and cultural history contribute to persis- is produced on just 1.5 percent of the world's tent divisions, and how countries impose land, which would fit comfortably into Alge- barriers to productive interaction with ria. This dense economic mass is home to their neighbors and the rest of the world. about a sixth of the world's people.2 Economies benefit from gradually lowering High density reflects the self-reinforcing barriers, and rich countries tend to have the benefits of proximity between economic lowest barriers to trade and factor mobil- agents across spatial scales--local, regional, ity. Countries that have integrated region- and international. Distance also matters for ally benefit from growth spillovers, larger countries and world regions. For the past home markets, and scale economies in pro- 50 years, by far the largest share of global duction and some types of public services. economic activity has been concentrated Some countries within a region may ini- in North America, Western Europe, and tially prosper more than others, but living Northeast Asia (see map 3.1). Being near standards eventually converge in regions these largest markets for products and sup- that have integrated. And in a world with plies opens great opportunities. Indeed, the economic activity and purchasing power correlation between access to markets and concentrated in a few regions, countries that economic growth is strong. have integrated globally benefit from access But it is the persistence of divisions to those markets and sources of investment. between nation-states that sets the processes This chapter makes the case for countries to of economic geography apart for countries promote such integration. and regions. The latest wave of globaliza- The main findings: tion, which began after World War II, has been associated with a borderless world. In · Divisions between countries make for 1990 Kenichi Ohmae famously pronounced thicker borders in the developing world. that "borders have effectively disappeared."3 Borders restrict the flow of goods, capi- For some world regions and some transac- tal, people, and ideas everywhere. But tions across borders, this reflects reality. But larger countries with big markets may borders, rather than disappear, have tripled get by with more restrictive borders. 96 in the past 50 years. There are now about Small countries have to worry more. Division 97 Some types of divisions, like being land- Map 3.1 Global GDP is concentrated in a few world regions, 2006 locked, are beyond the control of individ- ual countries. Others are self-imposed. And as countries develop, they gradually lower almost all types of barriers. · Economic mass is concentrated in North America, Western Europe, and North- GDP, US$ east Asia. And only East Asia has signifi- (billions) cantly increased its share of global GDP 12,000 6,000 in recent decades. This global concen- 3,000 1,500 tration matters greatly for the develop- 500 ment prospects of today's lagging world regions, and increasing their access to these large world markets must be a pri- ority for global development policy. Source: World Bank 2007j. · Within world regions, economic devel- opment tends to be accompanied by an Figure 3.1 The number of borders between nations initial divergence in living standards tripled in the past 50 years between countries, followed by conver- Number of borders gence. Basic health and education indi- 700 cators show improvements in almost all 600 world regions, but there is some diver- 500 gence in incomes between the richest and 400 poorest countries. The increasing inequal- 300 ity between countries within a region 200 reverses as lagging countries benefit from 100 0 growth spillovers from leading countries. 1820 1840 1860 1880 1900 1920 1940 1960 1980 2000 · Overcoming divisions between coun- Year tries regionally and globally is essential Source: Stinnett and others 2002. for sustained progress. This points to the importance of facilitating access to those borders are integrated in a functional global markets and promoting regional economic community (the Czech Republic integration in all its many forms (see and the Slovak Republic) or divided by con- chapters 6 and 9). flict, reducing the scope for further integra- tion (Eritrea and Ethiopia). Defining division Viewed through an economic lens, some borders are much wider than others (see Borders and divisions are not synonyms. map 3.2). The width or thickness of each National borders enclose people with shared country's borders is proportional to restric- characteristics, providing a sense of place tions that each country imposes on the flow and belonging that contributes to social wel- of goods, capital, people, and ideas with all fare. They also generate manageable units other countries.6 The wider the border, the for governing society. And well defined and more the country limits trade, travel, and the settled, they provide security and stability, flow of factors of production. yielding considerable economic benefits. Divisions, by contrast, arise when borders · Economic borders are narrow in North are poorly managed. They range from mod- America, Western Europe, Japan, Austra- erate restrictions on the flow of goods, capi- lia, and New Zealand; are wide in Asia, tal, people, and ideas to more severe divisions Africa, and Eastern Europe; and are in triggered by territorial disputes, civil wars, between in Latin America. Countries and conflicts between countries. Borders are with wide borders include emerging not a problem in themselves. But the conse- economies in East Asia and countries in quences for economic development are quite Sub-Saharan Africa, which for decades different when the countries separated by have had low growth. 98 WORLD DEVELOPMENT REPORT 2009 Map 3.2 Some borders are much wider than others Source: WDR 2009 team. Note: The width of borders is proportional to a summary measure of each country's restrictions to the flow of goods, capital, people, and ideas with all other countries. Gray areas = insufficient data. · Borders of the same width appear nar- opens its borders to benefit from interactions rower around larger countries. This with other countries, promoting further reflects the reality that large countries development. But there are exceptions. Some can often get away with more restrictive upper-middle-income countries maintain policies. Small countries depend more high restrictions--all of them oil exporters: on openness to overcome small markets Equatorial Guinea, Gabon, Libya, and Saudi and production scales. Arabia (upper right of figure 3.2). And some · Some countries with narrow borders are poorer countries have greatly reduced bor- surrounded by countries with restric- der restrictions, among them the landlocked tive policies, making it more difficult for countries of Armenia, Uganda, and Zambia, them to benefit from openness than for as well as the coastal countries of The Gam- countries in more open neighborhoods. bia, Georgia, Haiti, Kenya, Madagascar, and · This is true more for countries that are Nicaragua (lower left). open but landlocked, such as Armenia, How countries maintain divisions Uganda, and Zambia, than for those that are open and coastal, such as Chile Countries choose how permeable their or Georgia. Some coastal countries, by borders are, affecting the flows of goods, contrast, have such high restrictions that capital, people, and ideas. And the effects of they might as well be landlocked. division change as countries become more open to some flows and restrict others. Comparing border widths with eco- Goods and services. Borders reduce trade. nomic status confirms that wealthier coun- A study in the mid-1990s found that trade tries typically have lower border restrictions between Canadian provinces is, on aver- (see figure 3.2).7 As a country develops, it age, more than 20 times greater than trade strengthens the institutions that manage its between those provinces and equally distant borders and regulate the flow of goods and places in the United States. That implies a factors of production. It also becomes more "border-width" equivalent to increasing the integrated into the global economy and trade distance by 10,500 miles.8 More recent Division 99 Figure 3.2 Rich countries tend to have lower border (OECD) (see figure 3.3). Quotas, subsidies, restrictions antidumping duties, licensing, and idiosyn- GDP per capita (US$) cratic or confusing regulations affect trade as well.13 Using tariff and nontariff barriers, 25,000 poor countries restrict trade more than rich countries. They also face higher barriers to SAU GNQ LBY their exports. Nontariff barriers, on aver- 5,000 GAB age, represent more than two-thirds of total trade barriers, with higher proportions in ARM rich countries than in poor. 1,000 GEO NIC Capital. Restrictions on capital flows in HTI KEN 200514 are lower in industrial than in devel- ZMB GMB 250 oping countries (see figure 3.4) and are great- UGA MDG est in Africa, Central Asia, and South Asia. Recent empirical work--much prompted by Less More the financial crises of the 1990s--provides Border restrictions qualified evidence that financial globaliza- Source: WDR 2009 team (see note 6). tion benefits developing countries and that Note: GDP per capita is for 2005 in 2000 U.S. dollars from a series used in later sections of this chapter and based on greater financial openness does not by itself World Bank (2007j) and Maddison (2006). ARM = Armenia; contribute to more severe economic crises.15 GNQ = Equatorial Guinea; GAB = Gabon; GMB = The Gambia; GEO = Georgia; HTI = Haiti; KEN = Kenya; LBY = Libya; By reducing the cost of capital in receiving MDG = Madagascar; NIC = Nicaragua; SAU = Saudia Arabia; countries, freeing capital account transac- UGA = Uganda; ZMB = Zambia. tions increases the availability of resources estimates suggest that international borders for productive investment. It can also pro- reduce trade between industrial countries by mote portfolio diversification, thus mitigat- a still significant 20­50 percent.9 The reduc- ing risk, and encourage sound monetary tions are even larger for developing countries, management. From 1955 to 2004, freeing which tend to have higher trade barriers. capital accounts had a positive association Countries that encourage exports and are with growth in both developed and emerg- open to imports of goods and services grow ing economies.16 Liberalizing equity markets faster and reduce poverty more than coun- tries that do not encourage exports. When Figure 3.3 Tariffs are highest in Africa, South Asia, and Western Asia exports are concentrated in labor-intensive Average tariff, 2005 manufacturing, trade increases the wages for Eastern Africa unskilled workers, benefiting poor people. Northern Africa It also encourages macroeconomic stability, again benefiting the poor, who are more likely Central Africa to be hurt by inflation. And through innova- South Asia tion and factor accumulation, it enhances Western Asia productivity and thus growth.10 There may Western Africa be some empirical uncertainty about the Southeast Asia & Pacific strength of trade's relationship with growth.11 South America But essentially all rich and emerging econo- Central America & Caribbean mies have a strong trade orientation. A country's openness to trade is often Southern Africa measured by a country's sum of exports Other high-income countries and imports as a share of GDP. But a more Central Asia, Caucasus & Turkey direct measure is the average tariff rate, Northeast Asia which fell globally from close to 30 per- Eastern Europe & Russian Federation cent in the early 1980s to about 10 percent OECD countries in 2005.12 Tariffs are highest in Africa, 0 5 10 15 20 South Asia, and Western Asia and lowest in Percent member countries of the Organisation for Source: World Bank 2006f. Economic Co-operation and Development Note: The figure reflects the unweighted mean of country average tariffs. 100 WORLD DEVELOPMENT REPORT 2009 Figure 3.4 Capital restrictions are highest in Africa, South Asia, and Central Asia States or other middle-income countries.19 Migrants move for higher wages, greater Central Africa education opportunities, or a better quality Western Africa life (see chapter 5). Sending countries receive South Asia remittances, shed surplus agricultural labor, Southern Africa and benefit from return migration by those Northern Africa who have acquired skills or capital abroad. Receiving countries, many with aging popu- Central Asia, Caucasus & Turkey lations or chronic labor shortages, increase Eastern Africa their labor pool by admitting unskilled Southeast Asia & Pacific workers and their productivity by attracting Northeast Asia highly qualified migrants. Eastern Europe & Russian Federation The economic benefits from more Western Asia migration could be great.20 The pool of potential migrants is likely to remain Central America & Caribbean large given prevailing wage differentials South America between poor and rich countries, three to Other high-income countries four times those triggering the mass migra- OECD countries tion of Europeans to North America in the 0 20 40 60 80 100 late-nineteenth century.21 Yet, despite the Capital restriction Index potential benefits and the ready supply of Source: Chinn and Ito 2006. migrants, most countries restrict in-migra- added 1 percentage point to annual GDP tion, largely because of perceived negative growth.17 But short-term debt flows, which effects on domestic labor markets. include portfolio bond flows and com- Comparable information on migration mercial bank loans, can be highly volatile. restrictions is not available. But countries In countries where the financial sector is also regulate admission of short-term underdeveloped, governments and financial visitors. Each country faces a tradeoff in institutions may increase their exposure to allowing people from some nations to visit short-term debt and thus their vulnerability for business or pleasure, while deterring to sudden outflows. residents of other nations for economic, The indirect benefits of global integration political, or security reasons. This pro- and free capital flows may be greater than duces a complex system of "unequal access the direct effect of capital accumulation to foreign spaces"22 that reflects similar and portfolio diversification. Open markets restrictions for people seeking to migrate. can enforce monetary discipline, macro- Residents of richer countries face fewer visa economic stability, and financial develop- requirements than those from poorer coun- ment. They can also strengthen institutions tries (see figure 3.5). But poorer countries and governance structures. And they can also restrict entry by visitors from other increase integration with the global econo- nations. Exit can be regulated as well. Many my.18 Where markets and governance are countries make it difficult for their citizens well developed, financial globalization con- to leave.23 Passport costs across countries tributes to GDP and productivity growth are as high as 125 percent of per capita gross and reduces financial vulnerability. Where national income (GNI), and higher costs they are not, the impacts on growth are are associated with lower migration rates. ambiguous, and the risk of a financial crisis Ideas. Basic labor-intensive manufac- is high. turing is a stepping stone for countries to People. Migration flows have increased improve their economic fortunes. But to with globalization, but much less than trade maintain growth that outpaces population or capital flows. Global estimates suggest that and reduces poverty, an economy needs to 11 million people move annually for longer- move from low-margin activities to the devel- term employment or to settle in another opment and production of new or improved country. About 3.5 million of them are low- products, a process associated with moving skilled workers, many migrating to the Gulf from low-income to middle-income status. Division 101 Endogenous growth theory stresses that new Figure 3.5 Residents of richer countries face fewer visa requirements ideas support this transition, generating eco- Hard to get around: % of countries for which a visa is needed nomic rents that enable the accumulation of 100 privateandpubliccapital.China--forthepast two decades a producer of low-margin, stan- dardized manufactured goods--now exports more than $300 billion worth of information 80 and communication technology (ICT) goods a year. So far, most of these exports have been assembled from imported components, with the largest rents captured by foreign firms 60 that develop innovative technologies and control marketing and sales. Of the retail proceeds from an iPod® music player assem- bled in China, more than half goes to Apple's 40 profits and the retail and distribution costs.24 Assembly and testing account for only about 2 percent of the final sale value. Freedom of access to all types of infor- 20 mation is necessary for an atmosphere that induces innovation and productivity. Ideas and knowledge spread through the research and development (R&D) investments by 0 firms and governments and through the 0 20 40 60 80 100 global stock of existing knowledge acces- Hard to get in: % of countries needing a visa to enter sible through publications, patents, and so Source: Neumayer 2006. on.25 Governments do not restrict the flow Note: Circles are proportional to GDP per capita; visas available at the border are not included. of purely technical information, although no control. These include being landlocked, poorer countries have limited access to being in a remote location (especially if such information because of cost or lan- combined with small size), and having a guage barriers. high degree of ethnic or cultural heteroge- The link between the free flow of ideas neity within and across borders. and economic development is somewhat Landlocked. There are 43 landlocked ambiguous and not well researched. A free countries in the world. Being landlocked press generally reduces corruption and reduces growth by at least half a percentage increases public accountability.26 An indica- point.28 Boxes 3.1 and 3.2 illustrate further tor of press freedom reported annually since the costs of being landlocked. Small surprise 2002 by Reporters without Borders covers then, that many landlocked countries are freedom and security in reporting, govern- among the world's poorest. But being land- ment control of media, restrictions on Inter- lockedinitselfisnotacauseofpoverty--look net providers, and censorship of content.27 at Botswana, Luxembourg, and Switzerland. Western industrial countries generally have The problem is being landlocked with poor a high degree of freedom. Many low-income neighbors or being landlocked far from countries have high restrictions on the media markets.29 Often the two go together. Africa and Internet traffic. Significant restrictions has the most landlocked countries (15), and persist in parts of Africa, East Asia, the Mid- Eastern Europe and Central Asia the high- dle East, and the former Soviet Union. est proportion--about half (see map 3.3). Bhutan, Lao People's Democratic Republic, Some divisions are beyond the control and Nepal in Asia, and Bolivia and Paraguay of individual countries in South America are other poor landlocked Countries for the most part are free to countries. determine their openness to the outside Country size. A large land area is world. But geography and history produce often associated with abundant natural divisions over which countries have little or resources (see box 3.3). A large population 102 WORLD DEVELOPMENT REPORT 2009 BOX 3.1 A country's neighborhood matters: regional integration and growth spillovers Spillovers of growth from across borders average growth spillover between 1970 resource-poor countries of Central Africa are among the main bene ts of regional and 2000 was 15.3 to 17.0 percent. This would be left further behind. integration.a In a more integrated economic contributed to a slow, but steady, conver- If Switzerland had been subject to the space, the long-run growth prospects of gence in living standards, with the gap in same low spillovers experienced by the countries become interlinked as markets prosperity between the poorest and rich- Central African Republic between 1970 of neighboring countries become more est OECD countries closing at an average and 2000, its GDP per capita in 2000 accessible. Growth in neighboring countries rate of 1.59 to 1.85 percent a year. Along would have been 9.3 percent lower, with enhances domestic growth, which bene ts with this, the e ectiveness of growth- a cumulative GDP loss of $334 billion neighbors. This spatial multiplier enhances promoting domestic policies has been (2000 constant U.S. dollars), or 162 per- the rewards to good policy and contributes supplemented by 18.1 to 20 percent. cent of Swiss GDP (see the gure below). to convergence in living standards. In Sub-Saharan Africa the average Putting Switzerland in Africa would have cost growth spillover has been far weaker, Quantifying the benefits of growth it $334 billion signaling the relative lack of regional inte- spillovers gration despite a plethora of RTAs. The GDP per capita (constant US$, thousands) From 1970 to 2000, membership in a growth spillover is estimated at only 2.9 to 29 common regional trade agreement (RTA) 3.9 percent, implying a spatial multiplier of 28 Switzerland, actual among neighbors was associated with a only 1.01 to 1.04. This nding of virtually no growth spillover of 13.6 to 15.3 percent, 27 growth spillovers holds when neighbors so every percentage point increase in 26 are de ned by contiguity rather than RTA the average growth rate of RTA partners membership. A typical Sub-Saharan coun- 25 brought a "growth bonus" of 0.14 percent try's growth rate was basically indepen- 24 to supplement domestic growth. Associ- dent of the growth rates of its neighbors. ated with this is a spatial multiplier of 1.14 23 to 1.18, with regional integration increas- Implications for landlocked and 22 ing the e ectiveness of growth-promot- resource-poor countries in Sub- Simulation of Switzerland 21 in Africa ing domestic policies by 14 to 18 percent. Saharan Africa 20 In Europe and East Asia, where regional Under current conditions, if the Sub- 1970 1975 1980 1985 1990 1995 2000 integration has been strongest, the Saharan countries whose natural endow- Year bene ts over the past few decades have ments are most favorable sustained Source: Roberts and Deichmann 2008. been even larger. For these countries the a growth takeo , the landlocked and a. Collier and O'Connel forthcoming. provides a ready market and large labor force. Conversely, small countries lack the scale, BOX 3.2 Bolivia and Chile's border--from wide to narrow? capacity, and stock of production factors Bolivia illustrates the economic Bolivia has South America's to achieve high economic growth by them- dependence of a landlocked country second-largest natural gas reserves. selves. But as with being landlocked, size on its neighbors and how economic So economic integration could be an by itself is not a determining factor. What integration could help overcome incentive for resolving regional dis- determines economic prosperity is a coun- these divisions. After a war with Chile putes. Chile would gain from energy try's economic integration with the rest of in the late-nineteenth century, Bolivia imports from Bolivia; Bolivia would lost its access to the Paci c, and Peru, bene t from better access to ports, the world.30 Luxembourg ranks 167th in Bolivia's ally, also lost territory to which would make it easier to export. population but has the world's highest GDP Chile. Peru would likely be involved in any per capita. Fully integrated in the European Chile and Bolivia have not had agreement because it provides an Union (EU), its highly specialized financial diplomatic relations since 1978, but alternative, though less economic, sector operates globally. Small countries they are now talking. A motive for route to the coast for Bolivia and should thus favor economic integration, Chile is natural gas. Since 1995 it because any corridor through Chile because they will gain most from freer has relied almost exclusively on gas would likely pass through former from Argentina, but supplies have Peruvian territory in Chile. trade and openness. been limited by high demand in Sources: The Economist 2007b, Mal- In world regions that are more highly Argentina. inowski 2007. integrated, parts of a country therefore have less incentive to remain within a nation dom- inated by another cultural or ethnic group. Devolution in the United Kingdom and sepa- ratist movements in Spain confirm this. Sim- ilarly, the "re-balkanization" of Southeastern Division 103 Map 3.3 Forty-three countries do not have direct access to the coast Source: WDR 2009 team. Europe with the disintegration of the former contrast, have more diversified economies Yugoslavia was in part facilitated by the pros- and, being closer to rich markets, benefit pect of EU accession for the newly indepen- more from tourism and trade. dent countries. Noneconomic considerations Mauritius shows that good policy can can dominate, however. Eritrea and Timor- overcome small size and remote location. It Leste have seceded from their larger neigh- now has the second highest GDP per capita in bors (Ethiopia and Indonesia) without the Africa despite being more than 900 kilome- benefit of integration with a larger economic ters from the nearest mainland. Its location association. among the Middle East, South Africa, and Sea-locked countries. Being landlocked India allows it to capture offshoring activi- can generate an island effect, preventing a country from benefiting from neighbor- ing suppliers and markets. Small islands in remote locations suffer similar isolation; BOX 3.3 The benefits of size they are essentially "sea-locked." They face Five bene ts of being a large country: · More e ective redistributive schemes hightransportcostsforexportsandimports, to reduce gaps in after-tax incomes higher costs for energy and intermediate · Smaller per capita cost of providing between rich and poor regions. inputs, and typically higher wage costs and many public goods, such as a judi- cial system or embassies. · Better ability to provide security, rents. The problems are acute for the small as the per capita cost of defense island nations of the Pacific.31 Trade prefer- · Larger home market, which can declines. increase productivity and thus ences to support them until they become bene t economic growth. A possible disadvantage is the competitive in world markets have gener- · Stronger bu er to regional economic greater heterogeneity of preferences ated large and unsustainable inefficiencies shocks--if a region that specializes and thus the larger coordination in production. And large per capita aid in, say, agriculture su ers a recession, costs in large democracies. Diversity flows have had only limited impact on their the impacts can be reduced through also makes it harder to overcome col- lective action problems. competitiveness. Closely linking up with transfers from other regions, and wealthier "patron" countries and increasing workers can seek employment else- labor mobility may be the only strategies.32 where in the country. Source: Alesina and Spolaore 2003. Small island states in the Caribbean, by 104 WORLD DEVELOPMENT REPORT 2009 ties in manufacturing and banking, as well as is statistically significant only in countries a thriving stopover tourism industry. where one group is in the majority but the Ethnic and cultural divisions. Ethno- minority groups are still powerful--for linguistic heterogeneity imposes a coor- example, Burundi and Iraq.34 In most cases dination cost on countries, because it ethnic or cultural differences are unlikely often reflects differences in attitudes or to be the cause of conflict. But ethnic dif- interests that need to be reconciled by ferences are exploited to achieve other national governments. Consider the dif- objectives, such as gaining political power ferences in opinion about joining the EU or control over resources. Ethnicity also among the French- and German-speaking interacts in complex ways with other facets parts of Switzerland. This heterogene- of society. Autocracy, for example, reduces ity also has implications for labor mobil- growth in ethnically diverse countries more ity. For instance, the Euro zone may be a than in ethnically homogenous ones. less resilient common currency area than Linguistic diversity varies greatly the United States, because its higher cul- between world regions. The Ethnologue tural heterogeneity hinders adjustments to database includes information on almost shocks through internal migration. Ethnic 7,000 languages, including their location. heterogeneity is often associated with civil The heterogeneity of language groups is conflict and with high costs for economic very high in Africa and generally increases growth. with proximity to the Equator (see map 3.4 Empirical evidence for the impact of cul- and figure 3.6). Although empirical cross- tural diversity is mixed (see also box 3.4). country studies suggest that linguistic frac- Ethnic fragmentation is negatively associ- tionalization hurts economic performance, ated with the quality of government and a regional trading language has tradition- with economic growth.33 The relationship ally helped overcome the divisions: Hindi between ethnic heterogeneity and conflict and Urdu in a large part of South Asia, Map 3.4 Language diversity is very high in Africa Figure 3.6 Globally, language diversity is highest near the equator Latitude 90 85 80 75 70 65 60 55 50 45 40 35 30 25 20 15 10 5 ­5 ­10 ­15 ­20 ­25 ­30 ­35 ­40 ­45 ­50 ­55 ­60 ­65 ­70 ­75 ­80 ­85 ­90 0 50 100 150 200 250 Source: World Language Mapping System, Ethnologue 2004. Languages per 1 million km2 of land area Source: World Language Mapping System, Ethnologue 2004. Division 105 Indonesian and Filipino in Southeast Asia, Arabic and Persian in the Middle East, Swa- BOX 3.4 Artificial states? hili in Eastern Africa, and Hausa in West- ern Africa. English, French, and Spanish Gathered in Berlin in 1884­85, the nial origin or location in Africa. Arti - have done the same, but in many countries colonial powers determined Africa's cial borders are not associated with a higher probability of war, re ecting they are used predominantly by an edu- borders with little concern for social or economic divisions. Many borders similar results on ethnic diversity and cated minority. in the Middle East were similarly con ict found by Paul Collier.c drawn at the end of World War I.a So, avoiding economic and political Economic costs of conflict Alesina, Easterly, and Matuszeski problems associated with ethnic diver- and territorial disputes identify "arti cial states" with a sity would require cultural homogene- Impermeable borders tend to reduce eco- measure of how straight a country's ity within countries. In Africa this would nomic growth. But full political unification border is and whether these borders imply a far larger number of countries. Yet the already small size of many Afri- between countries would not necessar- partition ethnic groups into two or ily improve economic performance.35 A more countries.b Northern Africa, can countries is perhaps a more severe Northeast Asia, and South Africa have problem--it prevents countries from full merger of two countries has a positive the most arti cial (straight) borders, reaching sustainable economic scale. country size effect but an overall slightly while South Asia and Western Africa As argued in this Report, the appropri- negative impact on growth due to reduced are the most partitioned. Eastern and ate response to small size and ethnic trade with the rest of the world. Only in Central Africa are among the top four diversity is closer integration and more a few instances would both partners ben- regions in both categories. permeable boundaries. efit from full political and economic inte- Empirical analysis suggests that Source: WDR 2009 team. arti cial borders hurt economic and gration. But integration of neighboring a. MacMillan 2003. social outcomes. But this link is less b. Alesina, Easterly, and Matuszeski 2006. markets without political integration, on signi cant after controlling for colo- c. Collier 2004. average, would increase growth across countries significantly. Borders further reduce economic ben- efits where divisions are aggravated by percentage point. It causes neighbors to conflict within or between countries. Even increase their military spending by 2 per- when conflict does not involve military cent. Other costs include refugee flows action, the cost can be significant. Territo- and disruption of preferred trade routes. rial disputes impose high international eco- The civil war in the Democratic Republic nomic transaction costs because of insecure of Congo closed river access to the sea for property rights and jurisdictional and pol- timber exports from the Central African icy uncertainty. Economic models suggest Republic. that the territorial dispute between Argen- tina and Chile reduced trade between the Economic concentration two countries by $33 billion between 1950 Economic output is spatially concentrated-- and 1995.36 The competing claims between by any measure and across geographic Japan and Russia over the Kurile Islands scales. Looking at grid cells, a quarter of the lowered trade by $535 billion between 1952 world's GDP is produced on just 0.3 percent and 1995. And those between Indonesia of the land area (about the size of Camer- and Malaysia cost $11.5 billion between oon), half on 1.5 percent, and nine-tenths 1980 and 1995. Similar disputes exist over on 16 percent.39 China, Japan, and the maritime boundaries, only about one-third United States produced about half of global of which are settled by treaty.37 GDP in 2006, and the 15 largest economies When disputes turn to military confron- produced about 80 percent. tation, the costs are considerably higher-- Early in the Industrial Revolution, at the not only in loss of life, but also in economic beginning of the nineteenth century, GDP terms. The cost of a "typical" civil war is per capita in today's industrialized coun- about $64 billion, and an average annual tries was about twice that of today's devel- worldwide cost of about $100 billion far oping and emerging countries (see table exceeds global aid flows.38 A civil war in a 3.1). But total GDP in China and India, neighboring country is estimated to reduce which had far larger populations, was more a country's annual growth by about half a than twice that in today's G7 countries. By 106 WORLD DEVELOPMENT REPORT 2009 Table 3.1 The concentration of GDP and population growth shifted between 1820 and 1998 Average annual Avergage annual Share of world GDP GDP growth rate population Excess growth (%) Share of world population (%) (%) growth rate (%) rate (GDP per 1820 1950 1998 1820 1950 1998 1820­1998 capita growth) G7 22.7 50.9 45.5 13.4 18.1 11.6 2.6 0.9 1.7 China and India 49.0 8.7 16.5 56.7 35.9 37.5 1.6 0.7 0.8 Rest of Asia 7.3 6.8 13.0 8.6 15.5 19.8 2.5 1.4 1.1 Latin America 2.0 7.9 8.7 2.0 6.6 8.6 3.0 1.8 1.2 Africa 4.5 3.6 3.1 7.1 9.0 12.9 2.0 1.3 0.7 Eastern Europe and the 8.8 13.0 5.3 8.8 10.6 7.0 1.9 0.8 1.1 former Soviet Union Source: Maddison 2006. Note: The rest of Western Europe, Australia, and New Zealand are not included. Figure 3.7 Concentration increases at the global How did this concentration come about? level, then a leveling off Shares of world GDP at different levels of GDP per The concentration of economic mass in capita, 1820­1998 today's western industrialized countries and Japan has its roots in eighteenth-century % share of global GDP (log scale) economic and technological innovation. Europe'seconomicgrowthacceleratedgreatly 20 Western Western Europe "offshoots" during the Industrial Revolution, with mod- (EU12) (United States; ern manufacturing starting in Great Britain 10 Canada) in the mid-eighteenth century and gradually Northeast Asia 5 spreading across the continent. At the begin- (Japan; Rep. of Korea; Taiwan, China) ning of this process, Western Europe had less than 20 percent of global GDP.41 By the end of 2 the nineteenth century, it had more than 30 percent, three-quarters of it in the four larg- 1 0 5 10 15 20 25 30 est economies--France, Germany, Italy, and GDP per capita (1990 int'l $, thousands) the United Kingdom (see also figure 3.7). This growth occurred against a backdrop Source: Maddison 2006. of frequent conflict between neighboring countries, constant changes of alliances, the middle of the twentieth century, the G7 and mergers and disintegrations of coun- countries accounted for more than half of tries. At the beginning of the nineteenth global output (about 60 percent if the other century, Germany included about 300 indi- western industrial countries are included). vidual states. It had 1,800 customs borders, North America and Japan grew the fast- with Prussia alone having 67 local tariff est at 3.5 and 2.8 percent a year between zones.42 Only in the 1870s did Germany 1820 and 1998.40 The four largest Euro- fully integrate domestically. Even with a pean economies grew at an annual average patchwork of economic regions in Europe, of about 2 percent, not very different from trade flows had always been large, thanks to growth rates in Africa, Eastern Europe, and local or regional agreements. These expand- the smaller Asian developing countries. But ing trade links inspired the work of David while GDP growth exceeded population Ricardo, who in 1817 famously described growth by 1.7 points in the G7, it did so by the exchange of textiles and port wine only 0.8 points in China and India and by between Great Britain and Portugal in his 0.7 points in Africa. Over the 180 years to theory of comparative advantage. Ricardo's the end of the twentieth century, these dif- work motivated further trade liberalization ferent growth rates moved the concentra- by governments, most of all Britain's. tion of economic production more toward Formal economic integration did not the northern industrialized countries. begin until the middle of the twentieth Division 107 century. Motivated by political as much as Japan started to industrialize fairly late. economic objectives, six European coun- In 1820 its GDP per capita was half that in tries, accounting for about a quarter of North America and Western Europe, a ratio world GDP, joined in a treaty liberalizing that did not change until the twentieth cen- trade in coal and steel. Annual GDP growth tury. GDP growth between 1820 and 1870 accelerated in subsequent years to around was 0.4 percent a year. Industrialization 4.5 percent, up from only around 1 percent began to accelerate after the Meiji Restora- in the 35 years after World War I. Although tion in the 1860s. The fastest growth rates the relative shares of European countries in were in the second half of the twentieth cen- world GDP dropped somewhat, the com- tury. Between 1950 and 1973, as the coun- bined EU economy maintains a share of 25 try opened to the world economy, Japan's percent, largely through enlargement to its economy grew at a rate of almost 9 percent current 27 member countries. a year. By the late 1980s, its GDP per capita Europe's economic progress was exported was higher than Western Europe's. to English-speaking "offshoots" in Australia, New Zealand, and North America. Between How did the rest of the world do? 1820 and the late-twentieth century, their The share of the largest industrial economies economies grew by about 3.6 percent, almost in world GDP has fallen slightly, from 51 per- twice the population growth of 1.9 percent, cent in 1950 to 46 percent in 1998.43 Among driven by massive migration mostly from emerging economies, Eastern Europe and Europe and Asia. Their share of global GDP Russia reduced their share from almost 5 increased from 2 percent to 25 percent during percent to 2.4 percent in the late 1980s and that time, the lion's share by the United States early 1990s. The smaller shares of industrial (22 percent). Cultural proximity and close countries and Eastern Europe are largely due trade ties meant that innovations crossed the to increases in Asia (see figure 3.8). South- Atlantic quickly in both directions. east Asia and the Pacific doubled its share Figure 3.8 Only Asia's share in world GDP has risen noticeably since 1980 Shares of world GDP of developing and emerging economies, 2000 constant dollars East Asia & Pacific Europe & Central Asia Latin America & Caribbean % of world GDP % of world GDP % of world GDP Northern Asia Central Asia, Caucasus & Turkey Central America & Caribbean 6 Southeast Asia 6 Eastern Europe & Russia 6 South America 5 & Pacific 5 5 4 4 4 3 3 3 2 2 2 1 1 1 0 0 0 1960 1970 1980 1990 2000 1960 1970 1980 1990 2000 1960 1970 1980 1990 2000 Year Year Year Middle East & North Africa South Asia Sub-Saharan Africa % of world GDP % of world GDP % of world GDP 6 6 6 Northeast Africa South Asia Eastern Africa 5 5 5 Western Asia Central Africa 4 4 4 Southern Africa Western Africa 3 3 3 2 2 2 1 1 1 0 0 0 1960 1970 1980 1990 2000 1960 1970 1980 1990 2000 1960 1970 1980 1990 2000 Year Year Year Sources: World Bank 2007j; Maddison 2006. 108 WORLD DEVELOPMENT REPORT 2009 to about 1.8 percent, and South Asia's share (see box 3.5). Trade decreases with distance of global GDP rose from 1.4 to 2.4 percent. and increases with GDP, so any country will The largest increase has occurred in North- trade more with nearby countries and with east Asia since the mid-1980s, essentially in countries that have a larger GDP. Despite China, where the share of global GDP rose reductions in transport and communication from less than 1 percent to about 5.5 percent. costs, the trade-reducing impact of distance Shares in the remaining World Development increased until about a half century ago, Report 2009 regions remained essentially remaining "puzzlingly" high since then (see, unchanged despite considerably higher pop- for example, for Brazil in figure 3.9).45 ulation growth.44 This empirical evidence may be at odds with the rapidly increasing long-distance Why does this matter? The importance trade between, say, China and the United of market access States or between Japan and Europe. But The distribution of economic production this increase in trade may not be so much globally matters greatly for the development due to trade cost reductions. It is largely prospects of countries because of the interac- driven by the other factor in the grav- tion of density and distance at a global scale. ity trade relationship: economic output.46 This is demonstrated by the close empiri- China's GDP has increased, providing the cal relationship between trade as a driver economic mass to export goods to inter- of growth and two variables that define the national markets and to import consumer well-known gravity model of trade: (1) the goods, capital equipment, and intermediate distance between trading partners, and (2) inputs. Increasing trade, in a self-reinforc- their economic size as measured by GDP ing process, generates scale economies in BOX 3.5 Market access and per capita incomes Quantifying market access (sometimes the returns to market potential are increas- and distance from markets. For any given called "market potential") is not just of ing--the same amount of market poten- level, the size of the economy determines theoretical interest. Empirical studies have tial buys more per capita income--at least how well a country can take advantage of shown that market and supplier access for some countries. market access. Rich countries like Australia have a signi cant impact on growth and There continues to be a large variance of and New Zealand can compensate for a income. For instance, halving a country's GDP per capita at any given market poten- remote location by o ering a fairly large distance from its trading partners is associ- tial. Haiti's market potential is higher than market and supply capacity. ated with a 25 percent increase in per capita New Zealand's. Its proximity to the United income--more than the combined e ect of States raises its market potential, re ecting a. Redding and Venables 2004. a coastal location and open trade policies.a the interaction between economic size b. See Mayer 2008. Trade bene ts a country by raising factor incomes (wages) through expenditures by Market potential for countries has become more unequal trading partners for goods produced in that 1970 2003 country. The level of expenditures is in large GDP per capita relative to U.S. GDP per capita GDP per capita relative to U.S. GDP per capita part determined by the size of the trading partner's economy (density) and by physical 1.00 1.00 market access, largely determined by prox- imity to trading partners (distance) and the e ect of borders (division).b 0.10 0.10 Between 1970 and 2003, the distribution of per capita income spread out, re ecting greater global inequality among coun- tries--the poorest countries now have 0.01 0.01 smaller incomes relative to the United States (see the gures at the right). The dis- tribution also moves to the right, implying 0.01 0.1 1 10 100 0.01 0.1 1 10 100 that market potential is increasing almost Market potential relative to Market potential relative to everywhere as a result of global GDP U.S. market potential U.S. market potential growth. And its slope is getting steeper, so Source: Mayer 2008 for this Report. Division 109 Figure 3.9 The effect of distance between Brazil and its trading partners has remained considerable 1980 2005 Trade as a share of partner's GDP (%) Trade as a share of partner's GDP (%) 10 10 1.0 1.0 0.1 0.1 2.5 5 10 2.5 5 10 Distance (km, thousands) Distance (km, thousands) Source: IMF 2007. the trade infrastructure and services, such of economic output. As these distributions as efficient ports and frequent container change, so too do the prospects of national shipping links (see chapter 6). Larger econ- economies. These, in turn, influence devel- omies and richer countries can thus over- opment outcomes at the regional and coun- come the friction of long trade distances try levels, reflected in levels and changes in with higher economic density. income, health, and human capital. This human capital, most often considered an input contributing to human development, Divergence, then convergence is also a development outcome that raises the The changing geographic distribution of quality of life for individuals. world economic output reflects the concen- Three broad trends: tration of economic mass initially in West- ern Europe and later in North America. · A general increase in income and basic More recently, some deconcentration has living standards globally, but with some occurred as first Japan and then other econ- big exceptions. omies in the East Asia region have grown. · Considerable divergence of incomes China and India are reclaiming their posi- between the richest and the poorest tion among the countries and regions with countries, but some global convergence the highest shares of global GDP. Country in health and education. access to input and output markets influence · Some convergence within the faster the geographic distribution of absolute levels growing regions. Table 3.2 GDP per capita increased tenfold, 1500­1998 1990 international dollars 1500 1820 1870 1913 1950 1973 1998 1998:1500 Western Europe 774 1,232 1,974 3,473 4,594 11,534 17,921 23.2 Western offshoots 400 1,201 2,431 5,257 9,288 16,172 26,146 65.4 Japan 500 669 737 1,387 1,926 11,439 20,413 40.8 Asia (excluding Japan) 572 575 543 640 635 1,231 2,936 5.1 Latin America 416 665 698 1,511 2,554 4,531 5,795 13.9 Eastern Europe and the former Soviet Union 483 667 917 1,501 2,601 5,729 4,354 9.0 Africa 400 418 444 585 852 1,365 1,368 3.4 World 565 667 867 1,510 2,114 4,104 5,709 10.1 Interregional spreads 2:1 3:1 5:1 9:1 15:1 13:1 19:1 Source: Maddison 2006. 110 WORLD DEVELOPMENT REPORT 2009 General improvements 26.5 years in 1820 to 32.8 years in 1910 to Today's generation, by almost any global about 68 years in 2005.47 In the last 35 years summary measure of income and welfare, alone, average global life expectancy grew is better off than any previous generation by about 10 years. And a much larger share in human history. GDP per capita in 1990 of the world's population now has access to international dollars increased tenfold from basic education. In 1870 the mean years of $565 to $5,700 over the last 500 years, while schooling was 1.1 years, and the adult lit- population grew from 400 million to more eracy rate 25.5 percent.48 By 1929, schooling than 6 billion (table 3.2). Since 1820 output had increased to 2.5 years, and by 2000, to growth has been about 2.2 percent a year, 6.7 years, and literacy to 43.8 percent and bringing with it a considerable rise in living then to 78.3 percent (see figure 3.10). standards. Life expectancy at birth rose from Considerable income divergence between the richest and poorest countries, but Figure 3.10 Education outcomes have improved improvements in health and education Global average, 1870­2000 Over the past 500 years, per capita output Years of schooling Literacy rate (%) increased 40-fold in Japan and 65-fold in 8 90 Australia, Canada, New Zealand, and the 7 80 United States (see table 3.2).49 In Africa 70 6 it increased only threefold, and in Asia 60 5 Literacy rate (not including Japan), fivefold. Spreads 50 between the poorest and the richest regions 4 40 increased from a factor of 2 in 1500 and 5 in 3 Years of schooling 30 1870 to almost 20 by the end of the twenti- 2 20 eth century. During the past two centuries, 1 10 the Gini coefficient of inequality increased by 30 percent. Per capita income inequality 0 0 1870 1890 1910 1929 1950 1960 1970 1980 1990 2000 among world citizens increased by 60 per- Year cent, as measured by the Theil index, largely Source: Morrisson and Murtin 2005. because of income divergence between Figure 3.11 East and South Asia have been the only regions catching up countries rather than within countries.50 Average annual growth rate of GDP per capita, 1960­2006 The main story is one of an enormous increase in per capita incomes in Europe Northeast Asia and its offshoots. More recently this has Other high-income countries happened in East Asia, with Japan, whose Southeast Asia & Pacific GDP per capita has increased tenfold since South Asia 1950, and was followed by the Republic of Korea; Taiwan, China; China; and countries OECD countries in South Asia. GDP per capita in China, Northern Africa though still low in absolute terms, grew at Central Asia, Caucasus & Turkey 8.4 percent a year between 1990 and 2005. At Central America & Caribbean the low end of the income distribution, total Eastern Europe & Russian Federation GDP in the Central Africa region increased Western Asia threefold between 1960 and 2006, compared with Northeast Asia's 30-fold increase (see South America figure 3.11). With population growth out- Southern Africa pacing economic growth, per capita incomes Eastern Africa in Central Africa fell by 8 percent in constant Western Africa prices. Incomes in the poorest countries in Central Africa the world--mostly landlocked and many in ­1 0 1 2 3 4 5 6 Africa, home to the "bottom billion" of the Percent world's population--declined by 5 percent Source: World Bank 2007j. during the 1990s.51 Division 111 Between 1960 and the late 1980s, almost Figure 3.12 Life expectancy decreased significantly in many African countries every country in the world showed contin- Countries with largest increase/decrease in life expectancy, 1970­2005 ual increases in life expectancy at birth.52 In Belarus South Asia it increased from 42 years to 60, Zimbabwe and in Northern Africa from 47 years to 65. Botswana The exception was in Sub-Saharan Africa. South Africa Until the late 1980s, life expectancy increased Namibia slowly in Western, Central, and Eastern Kenya Africa and slightly faster in Southern Africa, Saudi Arabia where it rose from 46 years to about 60. Since Libya then, however, the HIV/AIDS epidemic has Egypt, Arab Rep. of Vietnam caused a large increase in mortality, bringing Oman life expectancy in Southern Africa below its Zambia level in 1960. In Central and Eastern Africa, Lesotho life expectancy is down less dramatically, Indonesia and Western Africa contained the epidemic Swaziland and saw only a slight decline in the rate of Bangladesh improvement. Nine of the 10 countries Nepal showing the worst trends are in Sub-Saharan Central African Republic Africa, and most of these are in Southern or Yemen, Republic of Southeastern Africa (see figure 3.12). Gambia, The Similar to life expectancy, global inequal- 40 50 60 70 ity in access to education fell sharply from Life expectancy (years) a Gini coefficient for years of schooling of Source: World Bank 2007j. 0.79 in 1870 to 0.39 in 2000.53 The high Gini coefficient in the nineteenth century was considerably after 1930, when primary educa- largely due to near-universal primary edu- tion was expanded in many developing coun- cation in Western Europe and its offshoots. tries.54 Between 1960 and 2000, the years of Other world regions started expanding edu- schooling among the working-age population cation much later, and inequality dropped increased across all world regions and income Figure 3.13 Education has become more equal since the 1980s Years of schooling for 15­46-year-olds (population-weighted averages) 14 All countries High-income 12 countries Middle- and lower- income countries 10 Middle East & North Africa Sub-Saharan 8 Africa Latin America & Caribbean 6 East Asia & Pacific South Asia 4 Europe & Central Asia 2 0 1960 1970 1980 1990 2000 2010 Year Source: Cohen and Soto 2007. 112 WORLD DEVELOPMENT REPORT 2009 groups (see figure 3.13).55 The ratio of highest economies should converge over time. Will to lowest population-weighted average educa- poor countries eventually catch up with the tion dropped from 9.7 years to 3.1.56 These rich? The question received considerable improvements have been fairly uniform attention among growth economists in the across regions, so the difference between the late 1980s and 1990s.57 They produced tools highestandlowestregionhasremainedessen- and techniques to analyze convergence, tially constant. Because poorer countries start relating growth to initial income, with the from a far lower level, however, their percent- expectation that lower initial status is asso- age improvements are much higher, suggest- ciated with higher growth rates. But there ing eventual convergence. has been little, if any, convergence between countries globally over the past five decades Some income convergence within (see figure 3.14). There is even some indica- faster-growing regions tion of divergence, though the trend is weak. Neighboring countries can provide mutually Within world regions, the evidence is much beneficial economic linkages, spillovers, and more differentiated. complementarities that allow whole groups Regionalintegrationandtemporaldynam- of countries to increase their incomes. If this ics make the study of convergence important. increases growth rates in poorer countries, First, economic fortunes are shaped by what neighboringcountriesdo,andsuccessfuleco- Figure 3.14 Slight global divergence in per capita nomic integration--overcoming divisions-- incomes, 1950­2006 can pull weaker countries toward incomes Countries with populations greater than 1 million that they cannot achieve in isolation. Higher Average annual growth rate of GDP per capita (%) convergence would be expected in regions 6 that have integrated. Second, in fast-growing 5 regions, there initially is divergence as the 4 leading regional economies pull away, but later there is convergence as poor countries 3 benefit from growth spillovers and begin to 2 catch up over time. 1 In East Asia, the fastest-growing world 0 region in recent years, convergence fol- ­1 lowed initial divergence. From 1950 to 1970, incomes diverged sharply as first Japan; and ­2 later Hong Kong, China; and then Singapore ­3 100 1,000 10,000 grew at very high rates (see figures 3.15 and GDP per capita, initial (constant 2000 US$) 3.16a). In the 1970s other countries joined Source: World Bank 2007j, Maddison 2006. the fast-growth club, notably the Republic of Figure 3.15 Divergence, then convergence in East Asia, 1950­2006 Countries with populations greater than 1 million, coefficient of variation and GDP per capita growth Coefficient of variation of GDP per capita GDP per capita (constant 2000 US$, thousands) 2.0 45 Japan 1.9 40 Hong Kong, China Singapore 35 1.8 Taiwan, China 30 Rep. of Korea 1.7 25 Malaysia 1.6 Thailand 20 China 1.5 15 Philippines Indonesia 1.4 10 Papua New Guinea 1.3 5 Vietnam 1.2 Mongolia 0 Lao PDR 1950 1960 1970 1980 1990 2000 2010 1950 1958 1966 1974 1982 1990 1998 2006 Cambodia Year Year Source: World Bank 2007j, Maddison 2006. Division 113 Figure 3.16 The East Asian growth experience had levels last seen in 1960. This convergence has two distinct phases much to do with market policies in China Countries with populations greater than 1 million, in 1950­70 versus 1976­92 and Vietnam as well as with a special blend of regional economic integration against a a. 1950­70 backdrop of globalization. Average annual growth rate of GDP per capita (%) There are few signs of convergence where 9 Japan growth has been sluggish and regional inte- 8 gration limited, as in Western Asia and East- 7 ern Europe (see figure 3.17). Western Asia 6 Taiwan, China includes resource-rich countries, with low 5 Korea, Rep. of Hong Kong, China and high populations, as well as resource- 4 Thailand Mongolia Papua New Guinea Singapore poor countries, such as Jordan. Low levels 3 China Indonesia Philippines of intraregional trade indicate low levels of 2 Cambodia Malaysia integration. Eastern Europe shows low varia- 1 Lao PDR Vietnam tion in per capita income until about 1990.58 0 100 1,000 After the disintegration of the Soviet Union GDP per capita, initial (constant 2000 US$) and the fall of the Berlin Wall, per capita b. 1976­92 incomes dropped drastically in some coun- Average annual growth rate of GDP per capita (%) tries and moderately in others. This diver- 8 China gence was reinforced as the western-most 7 Taiwan, China Hong Kong, countries reoriented their economic linkages Korea, Rep. of 6 Thailand China toward Western Europe, eventually joining 5 Indonesia Singapore the EU. Belarus and initially Ukraine, by 4 Malaysia contrast, maintained close links to the Rus- Cambodia Japan 3 sian Federation, which only recently began 2 Vietnam benefiting from natural resource­driven Lao PDR 1 Mongolia economic growth. Papua New Guinea 0 Philippines The southernmost economies in the ­1 Latin America and Caribbean region expe- 100 1,000 10,000 GDP per capita, initial (constant 2000 US$) rienced relatively low growth and limited convergence (see box 3.6). At the northern c. 1976­92 end of the region, in 1994, Mexico entered Average annual growth rate of GDP per capita (%) the first major regional free trade pact that 8 China includes both industrial and developing 7 Taiwan, China Hong Kong, Korea, Rep. of countries. The North American Free Trade 6 Thailand China Agreement (NAFTA) eliminated tariffs on 5 Indonesia Singapore most products traded between the United 4 Malaysia Cambodia Japan States, Canada, and Mexico. The evidence 3 since then illustrates three points about for- 2 Vietnam Lao PDR mal regional integration processes:59 1 Mongolia Papua New Guinea 0 Philippines · Formal integration followed many years ­1 of preparation, gradual informal inte- 100 1,000 10,000 gration, and domestic policy changes. GDP per capita, initial (constant 2000 US$) Mexico unilaterally reduced trade barri- Source: World Bank 2007j; Maddison 2006. ers and implemented regulatory changes Korea and Taiwan, China. Between 1976 and long before the agreement took effect. 1992, what looked like moderate divergence · The agreement led to large increases in (see figure 3.16b) actually represented two trade and foreign direct investment (FDI) groups of countries on separate but closely flows. Economic analysis suggests that linked convergence paths (see figure 3.16c). without NAFTA, Mexico's global exports Overall, this led to a strong regional conver- would have been about 50 percent lower gence as the variation among country GDPs and its FDI 40 percent lower. This likely per capita--while still large--dropped to contributed to significant poverty reduc- 114 WORLD DEVELOPMENT REPORT 2009 Figure 3.17 Western Asia and Eastern Europe have had little integration--and little convergence Western Asia Western Asia Average annual growth rate of GDP per capita (%), 1950­2006 Coefficient of variation 6 1.15 5 Oman 1.10 4 West Bank 3 & Gaza Saudi Arabia Israel Iran, Islamic Rep. of 1.05 Bahrain 2 Syrian Jordan Yemen, 1 Arab Rep. 1.00 Republic of 0 Lebanon 0.95 Djibouti ­1 ­2 0.90 100 1,000 10,000 1950 1960 1970 1980 1990 2000 2010 GDP per capita, initial (constant 2000 US$) Year Eastern Europe Eastern Europe Average annual growth rate of GDP per capita (%), 1950­2006 Coefficient of variation 4.5 0.65 Bosnia & Herzegovina 4.0 3.5 Hungary Latvia 0.63 Estonia 3.0 Bulgaria Albania Poland 2.5 Slovak Rep. Czech Rep. Serbia & Montenegro Romania 0.61 2.0 Belarus Lithuania Croatia 1.5 Macedonia FYR Russian Federation 0.59 1.0 Ukraine Moldova 0.5 0 0.57 100 1,000 10,000 1950 1960 1970 1980 1990 2000 2010 GDP per capita, initial (constant 2000 US$) Year Source: World Bank 2007j; Maddison 2006. Figure 3.18 Mexico and other LAC countries have not been catching up with the United States tion and income growth. GDP per capita GDP per capita in the largest LAC economies relative to U.S. levels in 2002 may have been as much as 4 per- Ratio of country GDP per capita to U.S. GDP per capita cent lower without NAFTA. 0.45 · Despite these positive impacts on the 0.40 Mexican economy, the agreement has not produced rapid convergence in incomes 0.35 (see figure 3.18). Mexico has avoided 0.30 major economic crises, suggesting greater Venezuela, R.B. de stability that can have significant welfare 0.25 Argentina effects.60 But its performance relative to Mexico 0.20 the U.S. economy has not differed much from that of several other Latin Ameri- 0.15 Chile can economies. Costa Rica 0.10 Brazil The large differences in economic output Colombia 0.05 will likely remain significant for some time. In fact, steady-state convergence estimates 0.00 suggest that Mexican incomes will reach 1960 1963 1966 1969 1972 1975 1978 1981 1984 1987 1990 1993 1996 1999 2002 2005 Year only about half of U.S. incomes. Among the main reasons are significant differences Source: World Bank 2007j; Maddison 2006. Note: LAC = Latin America and the Caribbean. in the quality of domestic institutions, in the innovation dynamics of firms, and in the skills of the labor force. These will all Division 115 BOX 3.6 Neighborhoods matter: Southern Cone versus Southern Europe Half a century ago the countries in the four countries' GDP per capita grew by an Europe at around 1 percent a year than in southern cone of South America-- average 1.7 percent a year. South America at 0.3 percent. Italy, Por- Argentina, Brazil, Chile, and Uruguay-- Economic dynamics in Southern tugal, and Spain bene ted from regional had per capita incomes similar to or Europe unfolded di erently. Italy was one growth spillovers, proximity to large mar- higher than the three Southern European of the founding members of the Euro- kets, and cohesion policies within a single countries with which they had strong cul- pean Community, and Portugal and Spain integrated Western European market. In tural bonds--Italy, Portugal, and Spain. joined in 1986 after emerging from a long the Southern Cone, regional integration The two groups have since followed dif- period under authoritarian regimes. From was slow, and integration with wealthy ferent growth trajectories. For most of lower levels, they grew at more than 3 markets in the Western Hemisphere was this period, the Southern Cone countries, percent a year, far outpacing Latin Amer- neglected for long periods. except Chile, followed similar protection- ica. While incomes converged in both Source: WDR 2009 team. ist policies. Between 1950 and 2006 the regions, they did so faster in Western a. Lucas Jr. 2007; The economic fortunes of Latin America and "Latin Europe" have diverged Average annual growth rate of GDP per capita (%), 1950­2006 GDP per capita (constant 2000 US$) 3.5 25,000 Portugal Spain 3.0 Italy 20,000 Italy 2.5 Chile Spain Brazil 15,000 2.0 Portugal 10,000 Argentina 1.5 Uruguay Argentina Chile 5,000 1.0 Uruguay Brazil 0.5 0 1,000 10,000 1950 1956 1962 1968 1974 1980 1986 1992 1998 2004 GDP per capita, initial (constant 2000 US$) Year Source: World Bank 2007j; Maddison 2006. Convergence in South America has been moderate; in Europe strong Average annual growth rate of GDP per capita (%), 1950­2006 Average annual growth rate of GDP per capita (%), 1950­2006 2.5 4.0 Brazil Chile Ireland Greece 2.0 Colombia 3.5 Spain Portugal Austria Ecuador Suriname Germany 1.5 3.0 Italy Norway Finland Peru Netherlands Guyana France 1.0 Paraguay Uruguay 2.5 Sweden Argentina Slovenia Belgium United Kingdom 0.5 Venezuela, R.B. de 2.0 Denmark Switzerland 0 1.5 1,000 1,000 10,000 GDP per capita, initial (constant 2000 US$) GDP per capita, initial (constant 2000 US$) Source: World Bank 2007j; Maddison 2006. benefit from closer integration with Mex- new countries or regions only occasionally ico's northern neighbors, but the process breaking into the ranks of the rich. First, will take considerable time. physical geography has helped some coun- tries become rich initially but continues Geography, globalization, to hold back others. Second, the forces of and development economic geography--starting from an Four main aspects explain the persis- initial advantage, such as technical inno- tent regional concentration of economic vation during the Industrial Revolution-- wealth over the past few centuries, with facilitated agglomeration economies and 116 WORLD DEVELOPMENT REPORT 2009 reinforced the concentration of economic differences and some of the variation in activity. Third, regional spillovers increased economic outcomes. But most of these economic activity in other countries within constraints can be overcome with enough a region, further increasing the scale and resources. They are thus a proximate rather scope of economic production. Fourth, than an ultimate cause of underdevelop- entirely new regions of economic concen- ment. High levels of malaria, for instance, tration emerged--as a response to conges- may be as much a symptom of persistent tion and a shift in established regions from poverty as a cause (see box 3.7). They are manufacturing to services, "freeing up" a grave concern for development interven- manufacturing opportunities elsewhere. tions but insufficient to explain global What does this imply for the prospects in patterns of economic wealth or to predict today's lagging world regions? future growth potential by themselves. Second-nature geography. An alterna- How much does geography matter tive but complementary explanation for today? global development patterns shows how First-nature geography. Physical endow- small initial differences between countries ments influence the development prospects and regions (for instance, natural endow- of countries. For instance, agricultural ments) can, over time, generate large dis- intensification in areas of good agroecolog- parities. A central question in economic ical endowments generates surpluses that development is how much growth is due can be shifted to more productive uses. But to differences in human and physical these assets are not distributed uniformly. capital accumulation, and how much to As Landes (1998) puts it: "Nature like life the efficiency of using these factors.63 Evi- is unfair, unequal in its favors." Research- dence from a growing number of studies ers have found a strong correlation confirms that levels of capital accumula- between economic output and geographic tion alone are insufficient to explain cross- characteristics. A simple regression of out- country differences in growth and income. put density (GDP per square kilometer) Instead, total factor productivity (TFP)-- on geographic variables--mean annual how efficiently factors of production are temperature, mean annual precipitation, combined--tends to better explain dif- mean elevation, terrain "roughness," soil ferences in growth and income between categories, and distance from coastline-- countries.64 captures 91 percent of the variability in TFP is, however, a vague concept that the density of economic production.61 A subsumes several aspects of economic pro- similar analysis explains 20 percent of the duction. Most generally, it relates to better difference in per capita output between technology for combining inputs to gener- tropical Africa and industrial regions, and ate products or services. This leads to cost 12 percent of the difference between tropi- reductions and thus increased competitive- cal Africa and other tropical regions. Cli- ness. Complementarities, spillovers, and mate also interacts with other factors, such economies of scale also explain differences as disease. Vector-borne diseases strike in TFP. Geographically, these externalities disproportionately in tropical countries, imply benefits for producers to locate close reducing productivity. Malaria is esti- to each other. Combined with scale econo- mated to cause approximately 1 million mies that favor larger production units, deaths and more than 200 million clinical the concentration of economic activities events among Africans each year.62 Other increases across geographic scales. Euro- purely geographic factors--such as being pean economic growth during the modern landlocked, which shaves half a percentage era was initiated by the industrial revolu- point off annual GDP growth, or a remote tion, which generated major technological location--were discussed earlier. advances. Improved technology and popu- Does this mean that geography dic- lation growth reinforced scale economies tates the destiny of countries? No. Physi- leading to concentrated centers of industri- cal geography helps explain initial growth alization. These centers attracted workers Division 117 BOX 3.7 The influence of first-nature geography: is it possible to eradicate malaria? The species of Plasmodia that cause cate malaria? The question has never been ter of epidemiological feasibility. Political human malaria most likely reached their satisfactorily answered at the global scale.d instability and geographic accessibility maximum global extent in 1900. Since But it is possible to start addressing are obvious examples, but these are oper- that time the a ected area has been pro- the problem. In the map below, risk is ational and not technical obstacles. gressively reduced by a regionally variable classi ed as stable if more than 0.1 case is What can be achieved with the 1.37 mixture of improving human conditions recorded per 1,000 population each year, billion people su ering stable risk? Ini- and deliberate control. The map below unstable if below this gure, and zero if tial evidence suggests that a substantial shows the di erence between the widest no cases have been recorded within the fraction of those a ected will be living in hypothesized extent of the distribution three most recent years of records. When areas of very low prevalence.g A detailed of all types of human malaria around overlaid on a population map for 2007,e investigation with mathematical models 1900a and the contemporary limits of Plas- 2.37 billion people were found to live in could estimate the impact from the exist- modium falciparum,b the most clinically areas with any risk of P. falciparum trans- ing toolkit of interventions. When this esti- severe and epidemiologically important mission. Globally, almost 1 billion people mate combined with a detailed analysis form of human malaria, in 2007. The for- lived under unstable, or extremely low, of the data on the e ciency of historical merly malarious areas are concentrated in malaria risk. Conditions of low risk are interventions, considerable insight could the temperature latitude extremes of the typical in the Americas and in South and follow. These approaches will help deter- parasite's ancestral distribution, in both East Asia but are also common in Africa. mine whether malaria is eradicable and, if the Northern and Southern Hemispheres. For 1 billion people at risk of unstable so, under what time frame and with what Researchers have documented the malaria transmission, malaria elimination resources. strong inverse correlation between the is epidemiologically feasible. Epidemio- Contributed by Simon Hay, David L. Smith, economic prosperity of nations and their logical feasibility was determined by and Robert W. Snow. contemporary malaria burden.c Richer reference to historical experience during a. Hay and others 2004; Lysenko and countries have less malaria, poorer coun- the global malaria eradication program Semashko 1968. tries more. This work also documents the and by inferring, through modeling, that b. Guerra and others 2008. many mechanisms, from individual to transmission could be interrupted by tak- c. Sachs and Malaney 2002. d. Roberts and Enserink 2007. macroeconomic, for malaria to contrib- ing insecticide-treated bednets to scale.f e. Balk, Deichmannand others 2006. ute to poverty. What if the constraint of There are many reasons in many regions f. Hay, Smith, and Snow, forthcoming. malaria were lifted? Is it possible to eradi- why elimination may not be a simple mat- g. Guerra 2008. Currently prosperous parts of the world were formerly malarious Stable Unstable Formerly malarious Never malarious Source: Malaria Atlas Project (MAP), Kenyan Medical Research Institute, and University of Oxford. 118 WORLD DEVELOPMENT REPORT 2009 and new firms, instigating a virtuous, self- ago, Japan would have seemed an unlikely reinforcing process that led to even greater source of inexpensive electronics and con- concentration. sumer goods for the U.S. market given the Development is contagious, tending to distance between the two countries. But spread across regions. Although growth the emergence of containerized shipping centers may start within specific areas in a allowed Japanese producers to be competi- country--the industrial belt in the north- tive in North American markets and later west of England or the mill towns in New in the European markets.67 The Republic England--dynamic centers tend to spread of Korea and Taiwan, China, followed in out. At the international level, growth Japan's footsteps. Manufacturing invest- spreads to neighboring states, giving rise to ments spread from there to South Asia, par- regional growth centers. With enough open- ticularly Malaysia and Thailand, and then, ness and interaction between countries, the after economic liberalization, to China. mechanisms for spreading growth are tech- nological spillovers and increasing special- What do we learn from this? ization, breaking up production processes. Size matters a lot. To generate scale This makes it more likely that some of the economies, a certain population and an eco- demand for intermediate products will be nomic mass need to be in place. In Europe satisfied from neighboring countries. This during the Industrial Revolution, a relatively can greatly expand trade, which produces large and concentrated population provided scale economies and steep increases in eco- both the labor that produced manufac- nomic productivity. The larger labor and tures and the market that consumed them. capital pools and the greater market size North America, when it shifted from natu- that emerge due to gradual improvement of ral resources to industry, had a large popula- transport links can lead to the rapid takeoff tion along its eastern seaboard, which grew of a regional economy.65 quickly with immigration from Europe and New regions of growth and wealth can elsewhere. East Asia has a vast population, emerge. This happens when growth in a core with first Japan and later China serving as regionhasreachedapointatwhichcongestion engines of manufacturing growth in the and rising wages encourage entrepreneurs to region. Each region benefited from a large seek new locations for production in nearby home market, but much of the production regions. This happened in Western Europe, was soon destined for export both within the when firms relocated manufacturing capac- region and to the rest of the world. ity to Central and Eastern European coun- Few countries have lifted their eco- tries, and in North America, when Mexico nomic fortunes based only on exports attractedinvestmentinmanufacturingcapac- of primary commodities. Botswana, a ity for the U.S. and Canadian markets. This sparsely populated country with large min- contagion model of region building would eral wealth and good policies, is one excep- suggest that all economic activity remains tion. Well-managed mineral resources can within an expanding contiguous zone--but help generate capital that can be invested in it does not. other sectors, but few countries have done Under some conditions, economic this successfully. Agriculture--important growth may leap to an entirely new region.66 for subsistence, for rural income genera- The location of this new center of global tion, and for specific regions in a coun- manufacturing depends on many factors, try--cannot by itself lift poor countries including market size, trade and transaction to middle- or high-income status. Rural costs, initial human and physical capital activities are either too small in scale to endowments, and competition from other provide sufficient surplus for export--or, potential growth regions. This leapfrogging in cases in which agricultural production model matches the emergence of East Asia has sufficient scale, it often benefits only as a global hub initially for labor-intensive a few large landowners or agribusinesses. production and later for technologically The verdict on services is still out. But it is more advanced production. Half a century unlikely that poor countries have enough Division 119 skilled white-collar workers to generate broad-based growth spillovers. India has BOX 3.8 Integration takes a long time, and its benefits do a large export-oriented service sector, but not come overnight it employs only about 560,000 of its more than 1 billion inhabitants, most in jobs In Europe, after the di usion of mod- Formal, de jure, integration thus fol- in constant-return customer support and ern industrial technology and the lowed de facto integration, providing back-office tasks.68 expansion of trade links in the early a framework and structure for deep- Manufacturing remains important. nineteenth century, it took more than ening already close relations. This 100 years before formal integration gradual process allowed institutions Each successful world region has, at some processes began in the 1950s. Even to develop and gave labor, nancial, point, made significant and broad-based then, the e orts were limited to agree- and product markets time to prepare gains with basic labor-intensive manufactur- ments on narrowly focused economic for possibly harsh adjustments, par- ing. This process initially led to a diversifica- issues between six countries. Gradually ticularly for recently joining countries tion of production as countries grew richer they expanded into additional areas with much smaller economies. Bul- and consumers demanded more varieties. of cooperation such as customs and garia and Romania, which joined in As economies in these regions expanded, nuclear energy. It took 16 years before 2007, added 8.6 percent to the EU's these agreements were consolidated land area and 6.3 percent to its popu- production and employment in individual in the European Community in 1967. lation but only 1 percent to its GDP.a countries started to specialize in what they Membership expanded slowly, with So the convergence of social and were best at, giving rise to interlinked net- three countries joining each decade economic outcomes across member works of production trading intermediate between 1970 and 2000, and nally countries will also take longer. Assess- goods among countries within the region. the addition of 12 Eastern and Central ing the bene ts from integration This is the point at which China and some European countries by 2007. Just as thus requires a long time horizon, as of the other "second-wave" economies in the initial Coal and Steel Community increased labor mobility, investment formalized long-established economic in private and public capital, and East Asia have arrived. In Europe and other and cultural ties between the member other structural changes accelerate regions that industrialized earlier, the share countries, each subsequent expansion growth in lagging member countries. of manufacturing in the economy has fallen followed a long period of ever-closer quite rapidly, with only highly specialized interaction between members and Source: WDR 2009 team. manufacturing remaining, such as machine accession countries. a. European Union 2007. tools or information technology (IT) equip- ment. In these countries, the service sector, including the research and design of prod- of bilateral agreements among a fairly large ucts that will be manufactured elsewhere, number of countries (see box 3.8). East now accounts for the largest share, by far, of Asia, by contrast, has created tightly linked employment and economic output. entrepreneurial production networks with Openness helps a lot--but it has to be relatively little formal protocol. Initial inte- introduced with care. Each of today's suc- gration in North America was facilitated by cessful regions initially developed its manu- a shared language and cultural background facturing sector behind a fairly substantive between Canada and the United States. The wall of tariffs and other protections. Only as relatively recent addition of Mexico has their economies matured and became more removed some divisions between economies dependent on foreign inputs and markets of greatly different per capita incomes. for their products did they gradually open Openness and integration are most ben- their borders and integrate regionally and eficial for smaller or landlocked countries globally. The rise of interlinked production whose access to world markets depends on networks that cross international borders neighboring countries. Luxembourg's small within each region required more coordi- size does not matter, because it is tightly nation and cooperation among countries, integrated in the European economy and not just for trade in goods and services, but thus operates more like a specialized city in also to settle on common standards and a large country. Switzerland's being land- regulations. locked has not constrained the development The process proceeded somewhat differ- of highly specialized manufacturing and ser- ently in each region, most formally within vice sectors. It can connect to world markets Europe, where the EU's political and eco- by air or through neighboring countries, and nomic integration superseded a patchwork its neighbors are significant destinations for 120 WORLD DEVELOPMENT REPORT 2009 its outputs. Integration has enabled the two overcome their significant external bar- countries to benefit from specialization and riers--the thick borders in the map that scale economies that would otherwise be opened this chapter (see map 3.2). Smaller achievable only in far larger countries. countries do not have this luxury. They To facilitate integration, industrial must learn to manage their borders more regions invested heavily in physical infra- rapidly to achieve economic integration structure that promotes intraregional trade. with their neighbors to attain competitive Initially, sea and river transport was most production scale and to access world mar- important for exporting manufactured kets. Countries and regions that do this products, requiring good coastal and river faster will have an advantage, but it will not ports. More recently, interrelated produc- be easy. By providing a vast unskilled labor tion processes require more timely avail- pool--and relatively little human or physi- ability of intermediate products, which has cal capital--countries like China and India moved a larger proportion of trade to road, can absorb new manufacturing capacity for rail, and air links. a long time. These are precisely the types of activities that might provide a path to middle income for the poorest countries. What's different for today's China also demonstrates the benefits of its developers? economic rise for its neighbors. Almost all Are the conditions today different, or is this East Asian countries have sometimes sig- just a continuing or recurring phase of glo- nificant trade surpluses with China in most balization similar to that of a hundred years manufacturing sectors.71 ago? In fact, goods and factor markets may Second, there has been an unprec- be no more closely linked today than they edented fragmentation of production pro- were a century ago. They may be some- cesses. This includes not only the intrafirm what more integrated for trade, no more division of manufacturing steps across integrated for capital, and less integrated several places, but more important the for labor.69 So how can lagging regions and intraindustry trade of increasingly special- countries join the group of leading world ized components and services, sometimes regions? Do they need to wait their turn, over long distances. Advances in communi- or are there ways for them to break out of a cations technology facilitate these complex geographic determinism? buyer-supplier networks. Although inte- Somecleardifferencesinthecurrentphase grated in global markets, production tends of globalization and economic development to be regionally concentrated. For smaller relate to the dynamics of economic geog- countries, this may be both a threat and raphy and the persisting divisions between an opportunity. The threat is that smaller countries. First, the scale and speed of eco- countries with poor infrastructure and low nomic integration in recent decades have skills will remain outside global trading been unprecedented. The economic liberal- networks. The opportunity is that, while ization in China and India, as well as in Rus- spatial concentration remains beneficial sia and South America, adds huge numbers for production, increasing specialization of unskilled workers to global production allows concentration and scale economies capacity.70 In many ways this is a reemer- within subsectors in which even small play- gence of those regions (Asia accounted for ers can carve out a niche. almost 60 percent of world GDP as recently In 1999 India's then-prime minister, Atal as the early nineteenth century). Behari Vajpayee, remarked on some of the China and India, because of the enor- same issues that have been discussed in this mous size of their home markets, are essen- chapter: "We can change history but not tially world regions of their own. With no geography. We can change our friends but formal internal divisions, they benefit from not our neighbors."72 Is he correct? On one scale economies and provide the incen- level, certainly. Countries cannot just pack tive for investors and trading partners to up and move to a better neighborhood the Division 121 way individuals can. But in an economic Many world regions continue to face and political sense, countries can change the impacts of significant division. But this their neighborhoods. Japan and the United Report shows that countries can improve States overcame deep divisions of history their economic fortunes by changing their and geography to become close neighbors neighborhoods virtually and practically. by developing extensive transport links For this, they must do two things. First, and increasing economic interdependence. they must overcome the limitations and Mexico and Turkey may be changing neigh- barriers of geography by developing close borhoods by reorienting economic ties from trade and transport links with markets and their traditional cultural backyards to more sources of investment in rich and emerging prosperous countries in another part of regions of the world (see chapter 6). And their neighborhood. European integration second, they need to seek strength in num- ended centuries of division and war. Since bers by "thinning" their borders and inte- December 2007, travel from the Portuguese grating their economies with their physical Algarve to Estonia is possible without once neighborhood (see chapter 9). showing identification. Geography in motion Overcoming Division in Western Europe The day will come when you France, you Russia, you Germany, all you nations of the continent, without losing your distinct quali- ties and your glorious individuality, you will merge into a superior unit, and you will constitute European fraternity. --Victor Hugo, from a speech at the 1849 International Peace Congress V ictor Hugo was laughed at when II. Destructive nationalism--and its Organization for European Eco- he said this, as were several of economic dimension, protection- nomic Cooperation (OEEC) in 1948 his predecessors who proposed ism--were indeed partly blamed for to implement the Marshall Plan. Its European integration. It took the catas- the disaster. Economic integration was mandate was to reduce trade barriers, trophe of two world wars to get people thus viewed as the best way to avoid particularly quota restrictions. Europe to take the idea seriously and make another war. That it should come in the early postwar years was a tariff- policy makers ready for radical change. through peaceful means and with and quota-ridden economy. Remov- The scale of devastation and misery is the main objective of maintaining ing trade barriers fostered the rapid the key to understanding the drive for peace was--and remains--a unique growth of trade. Between 1950 and integration: on top of the horrifying endeavor. In this respect, European 1958, manufacturing exports grew by death toll, the war caused enormous integration is a clear success. But it almost 20 percent a year in West Ger- economic damage. The war cost Ger- was not clear in the 1940s and 1950s many, 9.2 percent in Italy, and 3.8 per- many and Italy four or more decades that this vision of "Peace through Inte- cent in France. Additionally, average of growth and put Austrian and French gration" would succeed, particularly annual GDP growth was 7.8 percent in gross domestic products (GDPs) back because it came at the same time as the West Germany, 5 percent in Italy, and to levels of the nineteenth century.1 Cold War's division between the East 4.4 percent in France. Correlation is Overcoming division and its dra- and the West. not causality, and reconstruction was matic consequences was the objective Under American pressure, 13 a strong engine of growth. But the of European leaders after World War European countries created the rapid growth as European trade was Map G2.1 The division in Western Europe has gradually dissipated Stages of economic integration 1952 1990 2007 Source: WDR 2009 team. Geography in Motion 123 Figure G2.1 The stairway to success a custom union remove all tariffs for The institutional index of integration for the European Economic Community Six intra-EEC trade and establish a com- Institutional Index for Integration for EEC6 mon external tariff, but also a unified 100 economic area would promote free Monetary Union (1999) labor mobility, integrated capital mar- 90 kets, free trade in services, and several 80 Common Market (1993) common policies. This degree of eco- 70 nomic integration was not feasible without deep political integration. So, 60 EMS (1979) in retrospect, "using economics as a Customs Union (1968) 50 Trojan horse for political integration worked like a charm."2 As"guardians of 40 the Treaty,"the Court and the European 30 CAP (1962) Commission would control those coun- tries (especially France when de Gaulle 20 returned to power) that came to reject 10 the level of supranationality implied by 0 the Treaty. From 1966 to 1986, however, 1957 1962 1967 1972 1977 1982 1987 1992 1997 2002 the deep integration promised by the Year Rome Treaty stalled (see figure G2.1). Source: WDR 2009 team. Europeans began to erect barriers that took the form of technical regulations and standards, fragmenting markets--a liberalized was changing the minds reasons. Belgium, Italy, Luxembourg, classic reaction by lobbying industries of European policy makers. European and the Netherlands joined the project to defend their rents. integration was not just a political in 1951, and these six would become The Single European Act (1986) project--it also made economic sense. the driving force behind European relaunched the process of deepening eco- The European Coal and Steel Com- integration (see map G2.1). The ECSC nomic integration--all the more stun- munity (ECSC) was launched by showed that economic cooperation was ning given the slow disintegration during France and Germany, who invited other more feasible than political or military the 1970s. Emphasizing the mobility nations to place these two sectors under integration. of capital, the Single Act was also partly its supranational authority. The proj- The Treaty of Rome in 1957 created responsible for the birth of the Euro- ect was both political and economic the six nations of the European Eco- pean Monetary Union (EMU). Indeed, because it applied a supranationality nomic Community (EEC). The move the fixed exchange rate of the European onto two sectors that were considered committed the six to unprecedented Monetary System implied, with free strategic for economic and military economic integration. Not only would capital mobility, the loss of monetary sovereignty. This made the EMU more politically palatable for countries com- Figure G2.2 Border effects between the European Union and the United States remain more than twice that within the European Union mitted to fixed exchange rates. Overcoming division means reduc- Ratio of trade within borders to trade across borders ing the impact of borders on trade 80 flows. Has this been so in the European 70 Border EU9 United States Union (EU)? One way to answer the 60 question is to compare the volume of 50 trade within borders with the volume of bilateral trade between countries. The 40 Border United States EU9 ratio of the two is the "border effect." 30 Fontagné, Mayer, and Zignago (2005) 20 Within EU9 do this for the EU-9, the six founders 10 plus Denmark, Ireland, and the United 1980 1985 1990 1995 2000 Kingdom.The border effect for reported Year intra-EU trade fell from around 24 in Source: Fontagné, Mayer, and Zignago 2005. Note: The border effect is the reverse of the volume of trade within natural borders to the volume across borders. the late 1970s to 13 in the late 1990s--a 124 WORLD DEVELOPMENT REPORT 2009 Geography in motion substantial increase in integration (see The European regional integration sovereignty applied for membership. figure G2.2) unmatched in the world. process has spread. As the EU deepened That the EU with its unmatched supra- The border effect between the EU-9 and and enlarged, the cost of discrimina- nationality remains so attractive for the United States, while decreasing fast tory treatment (the natural implication outsiders is evidence of an enduring during the period, remains more than of any regional integration process) for success. twice that within the EU. Borders in the outsiders increased, creating a "domino EU have become thinner, but they have dynamic of regionalism."3 Even Euro- not disappeared. pean countries that most valued their Contributed by Philippe Martin. Part Two shaping economic geography In the past generation, there has been a slow revolution in economic thought, brought about by the recognition of imperfectly competitive markets, due mainly to increasing returns to scale, spillovers, and circular causation. A new way of thinking has transformed the classical analysis of industrial organization, economic growth, and international trade, and has delivered what were at rst controversial, but now widely accepted, implications for the progress of developing countries. Part two of the Report illustrates the interplay between scale economies, factor mobility, and transport costs to explain the formidable forces that shape the spatial transformation described in part one. Chapters 4, 5, and 6 are the stops in a tour of the"engine room,"each spotlighting a di erent facet of the interactions among agglomeration, migration, and specialization. CHAPTER 4 Scale Economies and Agglomeration T he most celebrated example in eco- one such occupation, fruitfully carried out nomics is perhaps the simplest. in villages. But such trades as manufactur- On the first page of The Wealth of ing and commerce can be carried out only Nations, published in 1776, Adam Smith in bigger settlements, because they require wrote of the benefits of dividing labor to access to both workers and customers. make pins. A single unskilled worker with- Caveats notwithstanding, the benefits of out the benefit of machines might make producing large quantities in a single plant fewer than 20 pins in a day. But in a pin or place have increased as transport costs factory that Smith visited, 10 workers, who have fallen in the two centuries since Smith divided among themselves the 18 operations visited the pin factory. Those who doubt the involved in making a pin, were produc- awesome potential of scale economies and ing 48,000 pins a day. Rather than strug- how access to world markets helps exploit gling to produce just a few pins a day, each them should visit Dongguan, a city half- worker was turning out almost 5,000. Later way between Guangzhou and Shenzhen in in Smith's classic work are two important Southeast China. Until the 1980s it was a qualifiers: the gains from dividing labor are collection of sleepy villages in China's Pearl limited by market size, and not all activities River delta. Since then it has rushed head- exhibit increasing returns to scale. long into the world of increasing returns The ability to transport products wid- (see box 4.1). Every year, millions of peo- ens the market, so cities are located near ple in the developing world enter this new the most natural and efficient of transport realm and the implications, for them and systems--waterways. Places blessed with for policy makers, are nothing short of this natural infrastructure often do well, revolutionary. while other places must bide their time. As This chapter summarizes the experience Smith wrote, of entrepreneurs over the last two centuries in exploiting economies of scale in produc- There are in Africa none of those great inlets, such as the Baltic and Adriatic seas in Europe, tion. It focuses on "agglomeration econo- the Mediterranean and Euxine seas in both mies," whose exploitation requires locating Europe and Asia, and the gulphs of Arabia, in areas densely populated by other produc- Persia, India, Bengal, and Siam in Asia, to ers. It next provides a brief synopsis of about carry maritime commerce into the interior two decades of work by economists seeking parts of that great continent: and the great to understand these scale economies-- rivers of Africa are at too great a distance work that has diminished the disconnect from one another to give occasion to any between research and the real world, and considerable inland navigation.1 that yields valuable policy insights. It then Besides, not all activities exhibit scale assesses whether policy makers in the devel- economies, and some do not need large oping world have been learning from this 126 markets to thrive. Subsistence farming is experience and analysis. Scale Economies and Agglomeration 127 BOX 4.1 Scale economies in an almost unreal world: the story of Dongguan, China In 1978 what today is the city of Dong- Dongguan's development since the the magnetic recording heads used in guan in China's Guangdong province 1970s, and particularly in the last decade, hard disk drives worldwide. Another pro- was but a collection of villages and small exempli es (perhaps in exaggerated duced 60 percent of the electronic learn- towns spread over 2,500 square kilome- fashion) the economic forces shaping ing devices sold in the U.S. market. A ters on the Pearl River, midway between East Asia's middle-income economies third produced nearly 30 million mobile Guangzhou to the north and Shenzhen (see the table below). phones, more than enough to provide and Hong Kong, China, to the south. The Location and favorable factor prices a mobile phone for every man, woman, area's population of 400,000 relied on undoubtedly spurred Dongguan's early and child in Peru or República Bolivari- shing and farming and--though not growth. For the rst decade and a half ana de Venezuela. the poorest in China--was not especially after China's reforms began, small and Agglomeration or external scale econ- prosperous. medium enterprises from both Hong omies are equally visible. The knowl- Today Dongguan is home to about 7 Kong, China, and Taiwan, China, were edge spillovers and lower logistics costs million people. More than 5 million of its attracted to Dongguan by plentiful sup- from locating close to input providers residents are migrants who work in the ply of land and low-cost labor, and by its and export traders have produced glob- thousands of factories that dot the city, proximity to both Guangzhou and Hong ally important industry clusters for knit- churning out a wide range of products Kong, China. Despite these factors, Dong- ted woolens, footwear, furniture, and in such huge volumes that recent media guan's rapid growth in the 1990s can toys. But the cluster that has dominated accounts have assigned Dongguan the best be understood through economies the industrial landscape of Dongguan label of "factory of the world." Dong- of scale, whether in the production of since the mid-1990s is telecommunica- guan's economy has grown at more than intermediate goods or di erentiated tions, electronics, and computer com- 20 percent annually since 1980, and in products, and agglomeration effects, ponents. Of the parts and components 2004 its gross domestic product (GDP) within and across industries. Combined used in manufacturing and processing was about $14 billion--greater than with reductions in transport costs and personal computers, 95 percent can be Iceland's. If one includes only registered improvements in logistics, technological sourced in Dongguan, and for several urban residents (as in o cial statistics), progress demonstrates that such e ects products, Dongguan's factories account Dongguan's GDP per capita of $9,000 in have emerged as important characteris- for more than 40 percent of global pro- 2004 made it the wealthiest city in China. tics of global production. duction. Even if the city's oating population of The internal scale economies are obvi- migrant workers is included, its GDP per ous. In 2005 a single plant in Dongguan Contributed by Shubham Chaudhuri. capita in 2004 was still more than $2,000. manufactured more than 30 percent of Source: Gill and Kharas 2007. Dongguan in numbers Average annual GDP growth, 1980­2005 (%) 22.0 GDP (US$, billions) 14.2 Population: registered residents (millions) 1.6 Population: total, estimated (millions) 7.0 GDP per registered resident (US$) 8,999 GDP per capita (US$) 2,070 Exports (US$, billions) 35.2 Imports (billion US$) 29.3 Government revenues (US$, billions) 1.0 Government expenditures (US$, billions) 1.2 Electricity consumption (kWh, billions) 35.2 Water consumption (ft3, billions) 1.5 Environmental impact indicators Sulfur dioxide emissions (tons, thousands) 199.4 Industrial waste water (tons, millions) 225 Sulfur dioxide emissions meeting standards (%) 92.9 Industrial water discharge meeting standards (%) 90.1 Industrial solid wastes (tons, thousands) 28.6 Industrial solid wastes meeting standards (%) 86.5 Global market share in 2002 of computer and electronics components manufactured in Dongguan (%) Magnetic heads and computer cases 40 Scanners and mini-motors 20 Copper-clad boards and disk drives 30 Keyboards 16 AC capacitators and fly-back transformers 25 Motherboards 15 Source: Dongguan Government 2005. 128 WORLD DEVELOPMENT REPORT 2009 The main findings: tate scale economies in marketing and distributing agricultural produce, medi- · Developing economies are entering a um-size cities provide localization econ- new realm of agglomeration. A century omies for manufacturing industries, and of experience indicates that as countries the largest cities provide diverse facilities develop from agricultural to industrial to and foster innovation in business, gov- service-oriented production,entrepreneurs ernment, and education services. and workers leave behind not just their vil- · Policy makers have often misjudged the lages and their agrarian occupations, but potency of market forces. Many policy also a world in which scale does not mat- makers perceive cities as constructs of ter much. More and more of them enter the state--to be managed and manipu- not just larger and denser settlements, but lated to serve some social objective. In also a world in which scale matters--where reality, cities and towns, just like firms production and distribution enjoy scale and farms, are creatures of the market. economies,especially those associated with Just as firms and farms deliver final and places.Proximity matters more,not just for intermediate goods and services, towns access to markets for goods and services, and cities deliver agglomeration econo- but also for access to ideas. mies to producers and workers. So city · A portfolio of places is needed for eco- administrators are better advised to nomic growth. Research over the last learn what their city does, and to help it generation indicates that different forms do this well, rather than try to abruptly of human settlement facilitate agglom- change the course of their city's des- eration economies for different forms of tiny. Planners and policy makers should production. A somewhat-oversimplified see their role as prudent managers of a (but not altogether incorrect) generaliza- portfolio of places, to get the most from tion would be that market towns facili- agglomeration economies. Table 4.1 A dozen economies of scale Type of economy of scale Example 1. Pecuniary Being able to purchase intermediate inputs at volume discounts Internal 2. Static technological Falling average costs because of fixed costs of operating a plant Technological 3. Dynamic technological Learning to operate a plant more efficiently over time 4. "Shopping" Shoppers are attracted to places where there are many sellers 5. "Adam Smith" Outsourcing allows both the upstream input suppliers and downstream Static specialization firms to profit from productivity gains because of specialization Localization Workers with industry-specific skills are attracted to a location where 6. "Marshall" labor pooling there is a greater concentration.a 7. "Marshall-Arrow-Romer" Reductions in costs that arise from repeated and continuous production Dynamic learning by doing activity over time and which spill over between firms in the same place The more that different things are done locally, the more opportunity there 8. "Jane Jacobs" innovation is for observing and adapting ideas from others External or Workers in an industry bring innovations to firms in other industries; similar agglomeration 9. "Marshall" labor pooling to no. 6 above, but the benefit arises from the diversity of industries in one Static location. Urbanization Similar to no. 5 above, the main difference being that the division of labor 10. "Adam Smith" division is made possible by the existence of many different buying industries in the of labor same place The larger the market, the higher the profit; the more attractive the location 11. "Romer" endogenous Dynamic to firms, the more jobs there are; the more labor pools there, the larger the growth market--and so on Spreading fixed costs of infrastructure over more taxpayers; diseconomies 12. "Pure" agglomeration arise from congestion and pollution Source: Adapted from Kilkenny 2006. a. For a formalization, see Krugman 1991a. Scale Economies and Agglomeration 129 This chapter discusses, in general terms, the implications of experience and analy- BOX 4.2 Sharing, matching, and learning sis for reshaping urbanization strategies in the developing world. Chapter 7 continues Three reasons explain why rms in a The ability to go beyond industry- this task of reframing the debate over urban particular industry often locate close speci c sharing, matching, and learning (localization economies) strategies. to each other. Geographic concentra- tion helps in-- to citywide processes (urbanization economies) requires additional mech- A guide to scale economies · Sharing. Broadening the market anisms. These include the e ects of for input suppliers, allowing them The benefits of increasing scale can be cumulative causation and the inter- to exploit internal economies of either internal or external to an individual penetration of production and trade scale in production (average costs across industries. They also include firm or farm. External economies are syn- decline as the scale of production gains from the cross-fertilization of onymous with "agglomeration economies," rises). This sharing of inputs also ideas. The concentration of workers which include the benefits of localization permits suppliers to provide highly and suppliers leads to a concentra- (being near other producers of the same specialized goods and services tai- tion of consumer demands. lored to the needs of their buyers. commodity or service) and urbanization If economies of scale are large The result is higher pro ts for all, (being close to producers of a wide range and unexhausted, and if rms can accompanied by easier access to a of commodities and services). Consump- compete not only on price but also broader range of inputs. through product di erentiation, tion externalities also are associated with · Matching. Expanding the availabil- strong centripetal forces come into agglomeration, but these are not yet well ity of the range of skills required play. In addition, by formally intro- studied in the literature.2 So, this chapter by employers to facilitate better ducing distance (the cost of shipping deals with production-related scale econo- matching to their distinctive needs. inputs and outputs), the framework mies (see table 4.1).3 At the same time, workers nd it used in this Report provides useful less risky to be in locations with insights into the centrifugal forces · Internal economies arise from the larger many possible employers. that explain spatial dispersion in a size of a plant to better exploit fixed costs · Learning. Accelerating spillovers of country. (numbers 1 through 3 in table 4.1). A knowledge and allowing workers larger steel mill can get volume discounts and entrepreneurs to learn from Sources: Gill and Kharas 2007, based on from suppliers--implying fixed costs of each other. Duranton and Puga 2004. transport and trade--and reap the ben- efits of dividing labor within the firm. · Localization economies arise from interactions, and as urbanization econo- a larger number of firms in the same mies, arising from between-industry inter- industry and the same place (numbers 4 actions.5 The reasons for producers to gain through 7 in table 4.1). Spatial proximity from proximity to others depend on the helps because immediate access to com- sharing of capital inputs, information, and petitors in the same sector allow firms labor. They also depend on improving the to stay abreast of market information matches between production requirements in negotiating with customers and sup- and types of land, labor, and intermediate pliers.4 Clustered firms can also share a inputs--and learning about new techniques larger and more dependable pool of spe- and products (see box 4.2). cialized labor. Internal scale economies are higher in · Urbanization economies arise from a heavier industries larger number of different industries in Internal increasing returns to scale are the same place (numbers 8 through 11 found in manufacturing and services, based in table 4.1). A management consulting on various sources of data. The internal company can benefit from locating near scale economies range from negligible or business schools, financial service pro- low among light industries, to high among viders, and manufacturers. heavy and high-technology industries (see Agglomeration economies depend not table 4.2). Based on engineering estimates, just on size (a big city or industry) but also a summary of sector-specific studies that on urban interactions. They are tradition- examines the minimum efficiency scale of ally classified as localization economies production and cost-saving finds significant arising from within-industry economic increasing returns in motor vehicles, other 130 WORLD DEVELOPMENT REPORT 2009 Table 4.2 Internal scale economies are low in light industries and high in heavy industries manufacturing sectors, with clothing, knit- Findings Data source ting, leather, and textiles at the lower end of the spectrum.9 Increasing output cuts costs Constant returns to scale: apparel, leather, Based on trade data (Antweiler and footwear, textiles, wood products Trefler 2002) in U.S. manufacturing, in the industries High increasing returns to scale: machinery, of middle-income countries (Chile), and pharmaceuticals, instruments, iron and steel, in the European car, truck, and consumer petroleum and coal products durables industries.10,11 Constant returns or low increasing returns to Based on engineering estimates to Based on trade data, a third of all goods- scale: leather goods, footwear and clothing, examine cost gradients and changes in producing industries have increasing timber and wood, textiles minimum efficiency scale (Junius 1997, cited from Prateen 1988 and Emerson and returns to scale.12,13 Manufacturing indus- High increasing returns to scale: Motor vehicles, others 1988) other means of transportation, chemicals, tries with the highest plant-level economies engineering, printing and publishing and industry-level externalities are petro- Low increasing returns to scale: footwear, Based on markups in manufacturing leum and coal products, petroleum refin- apparel, food products, leather industries for 14 OECD countries (Junius ing, pharmaceuticals, machinery, and iron 1997, cited from Oliveira and others 1996) High increasing returns to scale: tobacco, and steel. Industries with constant returns pharmaceuticals, office and computing include footwear, leather, textiles, apparel, machinery, railroad equipment and furniture. Low increasing returns to scale: apparel, leather Based on markups of prices over marginal Markups are another source of informa- products, textiles costs for two-digit sectors in the United States covering 1953­84 (24 sectors) tion. Because increasing returns to scale High increasing returns to scale: electric, gas, (Roeger 1995) and sanitary services, motor vehicles and confer market power on firms, markups of equipment, chemicals, tobacco price over marginal cost can be a proxy for Low increasing returns to scale: textiles, milk Based on production function estimates plant-level scale economies. Studies find a products, lumber mills, fish oil and meal products for 1963 Census of Manufacturing range of markups for U.S. manufacturing, Establishments in Norway (27 industries) High increasing returns to scale: basic metal, from 15 percent in apparel to more than 200 (Griliches and Ringstad 1971) transport equipment, cement products, fixtures, percent in the electric, gas, and sanitary ser- beverages vices. For 36 manufacturing sectors across Low increasing returns to scale: clothing, Based on cost and profit data in (167 19 member countries of the Organisation knitting, leather, textiles industries) four-digit SIC manufacturing industries for 1970 in Canada (Baldwin for Economic Co-operation and Develop- High increasing returns to scale: petroleum, and Gorecki 1986) and labor productivity ment (OECD), the highest markups are in basic and fabricated metal, transport equipment and output estimates for 90 four-digit tobacco, drugs and medicines, and office industries in Canada between 1965 and 1970 (Gupta 1983) and computing machinery--and the lowest in footwear, apparel, and wood products.14 Low increasing returns to scale: apparel, wood Based on estimates of firm-level products production function estimates for 6,665 While manufacturing data dominate plants in Chile during 1979­86 (Levinsohn the literature, increasing returns in services High increasing returns to scale: other and Petrin 1999) chemicals, food products, printing and publishing also are evident. The best-studied sector is Source: WDR 2009 team. electric power generation, where the inter- Note: OECD = Organisation for Economic Co-operation and Development; SIC = Standard Industrial nal increasing returns to scale are consider- Classification. able.15 The highest markups are in utilities transport equipment, chemicals, machinery, and sanitary services.16 Scale economies engineering, and paper and printing. In the also are found in banking and finance.17 three-digit product category, the highest A study of commercial banks in 75 coun- returns to scale are in books, bricks, dyes, tries shows that banks with larger loans and aircraft.6 By contrast, internal scale and deposits have lower average costs--and economies are negligible in rubber and plas- that banks operating in larger financial sys- tics, leather and leather goods, footwear and tems require less proportionate increases in clothing, and textiles.7 financial capital and have lower risk man- Based on cost and value added estimates, agement costs.18 different sources point to similar findings. A sample of 5,000 manufacturing firms Localization economies arise from in Norway shows evidence of scale econo- input-sharing and competition within mies at the individual industry level.8 For the industry Canadian industries at the four-digit level, Localization economies come from geo- returns to scale average 10 percent for 107 graphically concentrated groups of firms, Scale Economies and Agglomeration 131 linked by the technology they use, the mar- industry's labor force. About six of every kets they serve, the products and services 10 pairs of sheer hosiery sold in the United they provide, and the skills they require. States were knitted with Macfield yarn.22 Competitive pressures that force firms in the Together with other large North Carolina same sector to innovate or fail also lead to producers (Unifi, Regal, and Spanco), they productivity growth. Conditions tend to be make up more than three-quarters of the competitivewhenupstreamanddownstream industry's $3.7 billion worth of textured firms and associated institutions in a par- yarn products.23 The localization of the ticular industry (say, electronic machinery yarn and hosiery industries in North Car- or petrochemicals)--including universities olina is a powerful manifestation of intra- and trade associations--"cluster" together. industry external economies. Other channels for localization economies are the less easily measured "Marshall- Urbanization economies come Arrow-Romer externalities,"19 which come from industrial diversity that mainly from knowledge spillovers. fosters innovation Proximity to similar firms influences As cities grow, urbanization economies the location decisions of firms. Consider become more important.24 Urban diversity the hosiery industry in the United States. can foster the exchange of ideas and tech- Shortly after 1900 New York City became nology to produce greater innovation and the U.S. center for garment production and growth.25 Firms in different industries can distribution. But after World War II gar- share indivisible facilities or public goods, ment production moved south, to North a wider variety of intermediate input sup- Carolina.20 Many knitting and weaving pliers, a larger pool of narrowly specialized mills moved to be closer to the supply of workers, and risks. The evidence of greater yarn and to take advantage of cheaper importance of across-industry knowledge power, labor, and land. spillovers can be seen in established cities. Today, the hosiery industry, localized In fairly mature cities, such as Los Ange- in North Carolina, boasts many brands-- les, and Philadelphia, competition and city among them, Sheer Energy, Silken Mist, Just diversity help employment growth, indica- My Size, and No Nonsense--all competing tive of urbanization economies of between- in a $2 billion market. According to the U.S. sector innovation.26 On New York's Wall Census Bureau, about 150 establishments Street and in the city of London, financial producing women's full-length and knee- firms, insurance companies, and banking length hosiery in the early 2000s, half the syndicates benefit from being close to one nation's total, were located in North Caro- another. And co-location stimulates the lina. They shipped $973 million worth of growth of other specialist services, such as hosiery, about 75 percent of the national legal, software, data processing, advertising, total, and employed 13,497 people, including and management consulting firms. These 11,567 production workers.21 Adding men's clustered firms, by providing a thicker socks and stockings, more than half of a $6 market for highly educated individuals, billion industry is in North Carolina. benefit from drawing on the same large One reason textile producers went to pool of human capital. They also gain from North Carolina was to exploit productiv- the generation and diffusion of knowledge ity gains from proximity to upstream yarn amongst one another. producers. The yarn and pantyhose indus- Evidence of urban agglomeration econ- tries are tightly knit--in relationships omies comes primarily from developed delicately stitched together at each step countries.27 But there is also evidence of of production--but fiercely competitive. external economies in developing coun- Macfield, a textile giant and a leading pro- tries, wherever data are available. A survey ducer of yarns for pantyhose, socks, outer- of 12,400 manufacturing firms in 120 cit- wear, upholstery, and industrial products ies in China points to the higher produc- operates five plants in North Carolina tivity of firms in more populous cities.28 and employs about a quarter of the yarn Agglomeration economies in Indonesian 132 WORLD DEVELOPMENT REPORT 2009 people fell from just over 55 percent in 1960 to about 33 percent in 2004. Production tech- BOX 4.3 Agglomeration economies in Indonesia nology shifts away from constant returns to Much of the rigorous evidence of · Localization economies are strong increasing returns to scale. And over time, agglomeration economies comes for textiles and chemicals. scale-augmenting technical change boosts from developed countries. An excep- · Urbanization economies are strong scale economies. Imperfect and monopolis- tion is Indonesia, where recent for nonmetallic minerals and tic competition become the dominant forms research helps to identify the deter- machinery, though weak during of market structure. minants of industrial concentration. some periods. The analysis focuses on four broad The world is more urban, and the con- groups--chemicals (including Activities subject to urbanization centration of economic mass in the densest petroleum, rubber, and plastics); and dynamic economies are poor urban centers is greater as well. In 1900 the textiles (including garments, leather, candidates for policies that seek to number of people in the largest 100 cities and footwear); nonmetallic miner- spread out economic mass within added up to just 4.3 percent of the world's als (including glass, ceramics, and a country (see chapter 8 for a more population. The same 100 cities now have cement); and machinery (including detailed discussion). Such rms prefer 7.5 percent of the total, and the largest 100 electrical and nonelectrical machines, to stay put, since this helps learning, transportation equipment, and and they thrive in fairly large and cities, almost 10.5 percent. Despite ample instruments). It sheds light on how diverse cities. The agglomeration open space, almost all recent development the size and type of scale economies economies for textiles and chemicals in the United States has been less than 1 in uence the extent and pattern (largely static and local) indicate that kilometer from earlier developments.29 of agglomeration in a developing policies to deconcentrate production Even today, only about 2 percent of the country. in these industries might succeed land area of the United States is built up or Localization economies--the ben- if accompanied by improvements paved. Only agglomeration economies can e ts of locating near other rms in in infrastructure and governance in the same industry--have been more the areas chosen for relocation. The explain this extreme clustering of firms and important than urbanization econo- agglomeration economies make the workers in cities. mies for manufacturing, and static nonmetallic minerals and machinery agglomeration economies are more (essentially static and urban) likely As producers seek scale economies, important than dynamic (or learning to resist relocation to smaller urban agriculture disperses but related) externalities. The sector- centers. manufacturing clusters speci c ndings of tests for static As economies develop, farms spread out to externalities show that Source: Kuncoro, forthcoming. exploit scale economies in production. In the United States about 1,500 kilograms of agricultural products are produced annu- manufacturing between 1980 and 2003 vary ally to feed each American, whereas the over time; however, in the broadest terms, Chinese make do with about 600 kilograms these benefits are mainly static rather than per person. In 2005 the average cropland dynamic and somewhat more likely to arise in the United States was 20.4 hectares per from localization than from urbanization farmer, in Australia it was 45 hectares, and (see box 4.3). in Canada it was 47 hectares. Average farm size in Brazil is about 19 hectares.30 But A different realm scale economies in agriculture are generally Countries develop by shifting their econo- difficult to obtain in low-income countries. mies from traditional subsistence-based The cropland per farmer was a fraction of agricultural activities to higher-value manu- that in developed countries: 0.16 hectares in facturing and services. Along the way, firms China, 0.30 in Bangladesh and Indonesia, rather than farms become the dominant and 1.20 in Nigeria.31 production unit. The production of differ- As economies develop, manufactur- entiated manufactured goods and services ing and services become more important, increases as a share of the economy's out- firms cram in closer together to harness put. Between 1900 and 2000 the share of the agglomeration economies. In France, the global population in industrial or service- United Kingdom, and the United States, dominant urban localities rose from 15 per- 75­95 percent of industry is localized cent to 47 percent. The global employment (clustered or concentrated relative to over- share in agriculture among working-age all economic activity), while less than 15 Scale Economies and Agglomeration 133 percent is dispersed.32 In the United States also rises (see chapter 2).39,40 The important more than a third of aerospace engines are types of agglomeration economy change as produced in three cities: Hartford with development progresses. In particular, as an about 18 percent of total employment, and economy becomes more knowledge based, Cincinnati and Phoenix with another 18 knowledge spillovers, which require prox- percent together.33 Over time the spatial imity, become more important. Evidence concentration of industries in U.S. states suggests that knowledge industries are spa- has increased.34 Using continuous space tially concentrated.41 without considering administrative bound- Services are even more spatially con- aries and based on concentration of plants, centrated than manufacturing--for two more than half of the United Kingdom's reasons. First, they tend to use less land per 122 four-digit industries are localized, and employee. Banks, insurance companies, only 24 percent are dispersed. The rest are hospitals, and schools can operate comfort- randomly distributed.35 ably in high-rise buildings that economize Spatial clustering is more pronounced on land and allow for high density. Second, with high-skill and high-technology because of external economies, business industries (electronic computing machin- services have even greater potential for ery, process control instruments, semicon- agglomeration, as firms serve one another: ductors, and pharmaceuticals) than light every bank needs advertising, every adver- industries. This is consistent with the doc- tising firm a bank account. The potential umented findings of higher-scale effects in for codependence and agglomeration is heavier industries. High-skill and high-tech thus intrinsic to services.42 industries have more capital-intensive pro- Services are prominent among the most duction technology. They are also likely to agglomerated industries in the United benefit more from the various mechanisms States.43 Larger cities have been amassing that generate external economies (discussed service jobs from areas less than 20 kilo- earlier). meters away.44 Between 1972 and 1992, In the Republic of Korea the ranking of jobs in the United States became more industries by their localization economies spatially concentrated, driven primarily by follows the ranking of industries by their the rising localization of service activities spatial concentration across cities. Heavy in larger cities,45 as small and medium- and transport industries (metals, chemi- size counties lost jobs to the more urban cals, and transport equipment) tend to be areas.46 For instance, in Suffolk County, concentrated in a few highly specialized Massachusetts, which includes Boston, 35 cities to take advantage of local scale exter- percent of the workforce is in business ser- nalities, while traditional or light indus- vices, nearly twice the national average of tries with low scale externalities (food and 18 percent.47 In the United Kingdom nearly textiles) are more dispersed.36 High-tech 60 percent of all venture capital offices are industries (computers, aircraft, medical in London.48 London-based venture capital instruments, and electronic components) offices favor investment in London-based tend to be more concentrated than dura- small and medium enterprises to get bet- ble-good, machinery-related industries ter information: they can easily visit and (metal works, industrial, refrigeration, and monitor these enterprises. As communica- machinery and equipment).37 Cities in the tion costs fall, services become more trad- Republic of Korea have also become more able, allowing providers to take advantage specialized.38 of narrower specialization and agglom- eration economies. For instance, financial Services become even more densely services can be disaggregated into more clustered than manufacturing refined categories of retail banking, con- As countries move to a more mature phase sumer credit and financing, commercial of development, their economies become and corporate banking, investment bank- more knowledge based and service ori- ing, and so on. And within investment ented. The spatial concentration of activity banking, there is further specialization 134 WORLD DEVELOPMENT REPORT 2009 in mergers and acquisitions, corporate rumors, to the word-of-mouth learning in finance, fixed income, debt management, neighborhoods.52 and the like. Learning mechanisms also explain agglomeration in cities.53 As Alfred Mar- Cities facilitate scale economies shall implied, when knowledge spillovers of all types exist, "The mysteries of the trade become A plant in an isolated location can benefit no mysteries but are as it were in the air."54 from internal scale economies, but unless Knowledge spillovers are difficult to mea- it is situated in an area of density, it cannot sure, because they can seldom be traced enjoy the competitive benefits associated through transactions. With patent cita- with localization or urbanization econo- tions, however, it is possible to identify a mies. Towns and cities bring together paper trail for some knowledge spillovers. large pools of skilled labor and suppliers U.S. patent citations are spatially concen- of specialized intermediate inputs and trated, with citations 5 to 10 times more by doing so, enhance employer-employee likely to come from the same standard met- and buyer-seller matches. Input-sharing is ropolitan statistical area as originator pat- an important channel for agglomeration ents.55 Another strand of research focuses economies.49 Density of activity allows on workers as the primary vehicles of more refined specialization and a wider knowledge, implying that economies with variety of intermediate inputs. Averag- substantial labor mobility across industries ing across industries, a firm's relocation will exhibit a greater spread of ideas and from a less-dense location (of 499 or fewer growth.56 neighboring employees in the same indus- try) to a denser location (of 10,000­24,999 Agglomeration economies are amplified neighboring employees) results in a 3 per- by density and attenuated by distance cent increase in purchased input inten- Cities obviously reflect the demand for sity.50 The composition of a city emerges density. People choose to live close to one from the scope for agglomeration econo- another, paying high rents and tolerating mies and their interaction with other crime and congestion. This density helps aspects of microeconomic behavior. reduce distances of all types. Cities are thus Large cities with more firms allow work- a natural market creator and a conduit for ers to hedge against sector-specific risks. internal and external scale economies. Firms Smaller specialized cities expose workers to are drawn to dense areas concentrated with greater industry-specific shocks but provide people and infrastructure by the possibil- favorable match-specific advantages. In ity of serving a large local market from a both cases the concentration of economic large plant at low transport costs.57 Increas- activity lowers the search costs between ing return-to-scale production technology firms and workers, which results in fewer leads to large factories with many workers. unfilled vacancies, lower risk of job loss, The sizable workforce forms a large local and shorter durations of unemployment. market. By reducing transport costs, cities The large variety and quantity of inputs with a large local demand attract firms in to share in cities also implies better qual- different industries. So a self-reinforcing ity-matching. For instance, because of the process of agglomeration that begins with better matching possible, married couples the expanding local market further raises with university education, are increasingly industry productivity. found in large cities, up from 32 percent in Plants in dense economic environments 1940 to 50 percent in 1990.51 Cities make tend to be larger.58 As local market scale it easier for producers to find inputs and increases, firms are more likely to outsource for customers to experiment and discover their service functions to local suppliers.59 new possibilities. Examples of easy dif- This outsourcing further encourages com- fusion of information and social learning petition and diversity in the local business range from the congregation of diners in service market, which reinforces outsourc- certain restaurants, to the propagation of ing. Firms are attracted to locations with Scale Economies and Agglomeration 135 large concentrations of other firms in their Table 4.3 Scale economies amplify with density and attenuate with distance industry and with large demand.60 The large Finding Data sources and growing academic literature suggests Scale economies amplify with density . . . that doubling city size will increase pro- ductivity by 3­8 percent.61 In the Republic Doubling economic density increases 1988 data on output per worker in U.S. states productivity by 6 percent (Ciccone and Hall 1996) of Korea, a plant in a city with 1,000 work- Doubling employment density increases Data for the late 1980s on nonagricultural ers could, without altering its input mix, productivity by 4.5­5.0 percent private value added per worker in European increase output by 20­25 percent simply by NUTS regions (Ciccone 2002) relocating to a city that has 15,000 work- A one-standard-deviation increase in the Data on five traditional manufacturing ers in the same industry.62 And the spatial share of own-industry local employment industries in 224 U.S. metropolitan areas concentration of people reduces the cost of in the first period will raise that industry's between 1970 and 1987 (Henderson, Kuncoro, employment level by 16­31 percent in a later and Turner 1995) producing knowledge because information period transmission, competition, spying, imita- A 10-percent increase in local own-industry Republic of Korea city-industry data for 1983, tion, learning, innovation, and the com- employment results in 0.6­0.8 percent 1989, 1991­93 (Henderson, Lee, and Lee 2001) mercialization of new ideas are easier.63 In increase in plant output, for the same level of inputs the United States a staggering 96 percent of innovations occur in metropolitan areas.64 and attenuate with distance. Agglomeration economies are influ- Increasing distance from the city center by 1980 data for 356 new manufacturing firms in enced by geographic scope, and the den- 1 percent leads to a 0.13 percent decline in Brazil (Hansen 1990) productivity sity of economic activity and the distance between economic agents influence the Doubling the distance to a regional market Firm data in auto-component and agricultural center lowers profits by 6 percent machinery in Brazil and the United States productivity gains from scale economies (Henderson 1994) (see table 4.3). For example, doubling the Doubling travel time to a city center reduces Data for eight industries in Brazil (Sveikaukas density of economic activity in European productivity by 15 percent and others 1985) Nomenclature of Territorial Units for Sta- Own-county (lagged and contemporaneous) Plant-level data on productivity, 1972­92, in tistics (NUTS1) regions can increase total effect on plant productivity, but no effect from 742 U.S. counties (Henderson 2003b) factor productivity growth by 0.42 percent- neighboring county age points a year.65 Evidence from Brazil Effects of own-industry employment on new 12 million U.S. establishments from Dun & and the United States indicates that dou- plant openings attenuate rapidly within the Bradstreet Marketplace database (Rosenthal first five 1-mile concentric rings and Strange 2003) bling the distance to dense metropolitan centers reduces productivity by 15 percent; Source: WDR 2009 team. doubling the distance from 280 to 550 kilo- meters reduces profits by 6 percent. The introduced in The Wealth of Nations, that concept of distance can be generalized, of the "invisible hand" of perfect competi- in this context, from distance in physical tion.68 But perfect competition is an artifi- space to distance in industrial space. For cial theoretical construct: it assumes a large example, spillovers between industries are number of infinitesimal firms with negli- more likely if industries share related sci- gible influence over market prices, even in entific facilities.66 Furthermore, the extent the immediate vicinity of the firm's loca- to which distance attenuates agglomera- tion. Its assumption of constant returns to tion economies differs for different types scale further implies the so-called problem of agglomeration. For example, knowledge of "backyard capitalism."69 That is, in the spillovers that rely on face-to-face commu- world of constant returns to scale, small- nication decay more quickly with distance scale production is as efficient as large-scale than the home market effect.67 production, so every household should be producing a fully diversified range of goods A portfolio of places and services in its own backyard. Economics Adam Smith introduced scale economies, professors, when pressed by students to give factor mobility, and transport costs as a real-world example of such an industry, central to understanding the nature and would offer subsistence agriculture--small causes of the wealth of nations. But until farms producing wheat or rice, whose pro- the 1980s most economists were happier to duce could not be distinguished from those anchor their inquiries on another concept of others. Never mind that most people no 136 WORLD DEVELOPMENT REPORT 2009 longer worked on small farms in countries By the late 1980s scale economies were that had grown out of poverty. It led to con- standard features of the explanations for venient characterizations of the economy in international trade. By the early 1990s, which all firms and workers were identical, growth theorists had accepted the need to so one firm or worker could be considered incorporate imperfect competition among representative of all. Scale economies were firms into aggregate formulations of an inconvenient--they required acknowledg- economy. By the mid-1990s, theorists were ing that specialization differentiated people beginning to show how these ideas could be and products. used to understand the spatial distribution Occasionally, the contradiction between of economic activity, including the rise of internal increasing returns and perfect towns and cities. With the new economic competition would surface, but because of geography, researchers came to realize the technical difficulties it raised, it quickly that the dichotomy between internal and would be buried again.70 Then, during the external economies is often false. Why? 1970s, two economists at Princeton Univer- Because, in modeling the microfoundations sity proposed a technical solution to model of agglomeration economies, the source of increasing returns to scale, opening a door external economies have often been found for researchers to the same realm that so in the interaction of internal scale econo- many firms and workers had inhabited mies with other influences, such as trans- since the industrial revolution.71 port costs. Table 4.4 Thirty years of theoretical advance recognize the importance of scale economies Recognizing scale economies: recent Subject Main insights Key publications theoretical advances Industrial Increasing returns to scale and imperfect Spence 1976; Dixit The literature on the microeconomic organization, competition can be incorporated into formal and Stiglitz 1977 foundations of agglomeration economies 1970s economic models flourished in the last 20 years by combin- Urban External economies within cities and systems Mills 1972; Diamond ing models in the paradigms summarized economics, of cities; different levels of agglomerations are and Mirrless 1973; in table 4.4 and insights about urban 1970s related to city functions and Henderson 1974 economics that emphasize the tension International Increasing returns and imperfect competition Krugman 1980, 1981; trade, 1980s explain intraindustry trade between countries Ethier 1982; Helpman between benefits from the concentration with similar endowments; initial endowments may, and Krugman 1985; of economic activity and costs arising from through trade and specialization, influence the Grossman and that spatial concentration.72 In general, long-run rate of growth; trade unleashes forces of Helpman 1995 both convergence and divergence researchers have progressively recognized that economic growth has different impacts Economic Increasing returns-to-scale activities are Krugman 1991; geography, characterized by agglomeration and imperfect Fujita, Krugman, on firms and workers depending on their 1990s competition, while constant returns-to-scale and Venables 1999; sector and location. The underlying reason activities remain dispersed and competitive, Henderson 1999 is the love for variety in consumption and helping to explain spatial distribution of economic activity and growth of cities the economies of scale in production; the proximate reasons are product differentia- Endogenous Perfect competition and knowledge-related Romer 1986; Lucas growth, 1980s or human capital­related externalities imply Jr. 1988 tion, monopolistic power, specialization, aggregate increasing returns and explain why and location externalities. growth rates may not fall over time and why wealth The formal recognition of scale econo- levels across countries do not converge mies, externalities, and imperfect competi- Endogenous Imperfect competition explains why the incentive Romer 1990; tion makes economic theory conform more growth, 1990s to spend on R&D does not fall, and knowledge Grossman and spillovers explain why R&D costs fall over time, Helpman 1991; Aghion closely to the world in which policy makers resulting in more and better products that fuel and Howitt 1992 live. The policy implications of this work growth arise from the way economic production Endogenous Imperfect competition and Schumpeterian entry Aghion and Howitt relates to trade, ideas, and cities. growth, 2000s and exit of firms, with entrants bringing new 2005; Rossi-Hansberg technologies, explain how a country's growth and Wright 2007; Intraindustry trade. The main insight and optimal policies vary with distance to the Duranton 2007 · technology frontier; knowledge accumulation in coming from a formal recognition of cities leads to growth increasing returns to scale and product Source: Adapted from Gill and Kharas 2007. differentiation is that trade may take Scale Economies and Agglomeration 137 place between economies that are similar Smaller cities specialize, receiving in factor endowments: both interindus- industries as they mature and relocate try and intraindustry trade may profit- Even after controlling for natural com- ably take place. The main implication is parative advantage, externalities are still that countries may, in theory, encourage important in explaining the patterns of some activities and ensure comparative specialization and diversity among cities advantage. (see table 4.5). The production of nontradi- · Idea-driven economies. The insight is tional items is more concentrated in diverse that the nonrival nature of ideas makes U.S. cities, while standardized traditional them different from other factors of pro- goods are concentrated in smaller special- duction, such as capital, land, and labor, ized cities. Similarly, in Japan, smaller cit- in that the market may underinvest in ies are specialized, while low-tech activity the creation of new ideas. The main and standardized high-tech production implication is that governments should, processes are located offshore. Likewise, in theoretically, subsidize some strands of the Republic of Korea, large cities are more research and development (R&D), such service oriented and smaller cities, manu- as those that will ensure the continuance facturing oriented.78 of the comparative advantage a country Mid-size cities tend to specialize in has acquired in certain areas. mature industries, not new ones, and larger · City-based growth. The main insight is cities specialize in services not manufactur- that activities that display increasing ing.79 Improved infrastructure and falling returns generated by factors external to transport costs have encouraged standard- a firm tend to be concentrated in cities, ized manufacturing production to move while those displaying constant returns out of high-rent centers to smaller cities. remain more dispersed. The main impli- cation is that policies to keep cities busi- Table 4.5 Agglomeration economies vary by city size and profile, and by the industry life cycle ness-friendly and livable become more Main finding Data important as economies develop. Localization economies are more important for heavy Data for two-digit manufacturing Urban systems exhibit some stylized industries; urbanization economies are more important industries in Japan (Nakamura for light industries 1985) patterns. Larger cities tend to be more diversified and service oriented: they Localization economies become less important, giving Cross-sectional data for the way to urbanization economies, as cities expand in size United States and Brazil innovate, invent, breed new firms, and (Henderson 1986) expel mature industries.73 Smaller cities Scale economies from labor pooling are stronger Annual firm employment data tend to be industrially specialized: they in newer and expanding markets, while those from for four U.S. metropolitan areas produce or manufacture and receive relo- knowledge spillovers and specialized asset-sharing are and three two-digit industries cated industries from diversified cities.74 more important in mature markets (Hammond and Von Hagen 1994) The relative city-size distribution and For mature capital goods industries, there is evidence Panel data of 742 urban counties of localization economies but none of urbanization for 1970­87 (Henderson, Kuncoro, industrial concentration in specific cit- economies; for new high-tech industries, there and Turner 1995). ies tend to be stable over time. An urban is evidence of both localization and urbanization system tends to be made up of a few large economies diversified cities and many smaller, more For all industries both localization and urbanization Data for five traditional and three specialized, cities. 75 effects are important. For traditional industries most new high-tech manufacturing effects die out after four or five years, but for high-tech industries in 224 metropolitan The stylized observation in most coun- industries, the effects can persist longer. The biggest areas between 1970 and 1987 tries is an urban hierarchy of a few large effects are typically from conditions of three to four (Henderson 1997) cities and many smaller cities with var- years ago, in the county and metropolitan area ied economic functions.76 At the global The historical industrial environment of cities matters. In Growth data for the largest level, "world cities" at the top of the hier- fairly mature cities urbanization economies encourage industries (1956­87) in 170 U.S. industrial growth cities (Glaeser and others 1992) archy, such as New York, London, Paris, and Tokyo, are characterized by a diverse For high-tech industries a 1-standard-deviation increase City-industry data for the Republic in diversity of the local manufacturing base increases of Korea, 1983, 1989, 1991­93 industrial structure, predominantly ser- productivity by 60 percent, but diversity has no effect on (Henderson, Lee, and Lee 2001) vice based, and a labor force with a wide standard industries (such as textiles, or food). range of skills.77 Source: WDR 2009 team. 138 WORLD DEVELOPMENT REPORT 2009 Production in large cities focuses on ser- in production. This pattern became even vices, nonstandardized manufacturing, more marked during the 1990s.86 and R&D.80 The relocation of manufactur- Many business and economic historians ing to the suburbs has been documented in have argued that the extra costs of coordi- Colombia, Indonesia, the Republic of Korea, nating and monitoring multilocation firms and Thailand.81 It is common to find that relative to integrated firms have come down services do not deconcentrate from city cen- significantly following key developments in ters to their surrounding suburbs.82 transport and communication technolo- gies, as well as new management practices.87 Large cities diversify, incubate new Technological progress in transport and ideas and firms, and push out mature telecommunication made it less costly for industries firms to separate their production facilities New firms often start in diverse cities, but from their headquarters and management they move to specialized ones after they facilities. Firms can locate their production mature. Of all new plants in France, for facilities in environments with same-sector example, 84 percent were created in cities specialization, and their headquarters in a with above-median diversity.83 Some 72 metropolis with a concentration of busi- percent of firm relocations are from an area ness service employment. Furthermore, with above-median diversity to an area with the reduced communication costs that above-median specialization. In the United make transportation of service industry States almost all product innovations are in outputs (through electronic transmittal) metropolitan areas. Industrial diversity and cheaper did not imply the "death of dis- city size are both good for innovative out- tance" and the fading of cities into obscu- put.84 Trial plants are based in large cities rity, contrary to many predictions.88 In this in Japan, but mass production plants are context, while distance has become less in small cities or rural areas. Young firms important for transmitting information, it appear to need a period of experimentation has become more important for transmit- to determine their ideal production pro- ting knowledge. Telecommunication can cess.85 In the early learning phase, diver- be a complement to, but it is certainly not sified cities act as "nurseries" for firms to a strong substitute for, face-to-face interac- try out a variety of processes. Once a firm tions, which involve several forms of com- identifies its ideal process, it can begin mass munication simultaneously, notably body production in specialized cities, where all language and verbal conversation (see box firms share similar processes or specializa- 4.5).89 The geographic distribution of com- tions (see box 4.4). mercial Internet domains suggests that the The different economic functions that Internet is a complement to face-to-face cities serve can be seen in the clustering interactions (primarily within-city) as well of headquarters from different sectors and as a substitute for longer-distance commu- concentrations of business services in a few nication, such as phone or postal mail.90 large cities while production plants from each sector congregate in smaller special- Activities that cities specialize in are ized cities. In 1950 there was little differ- stable, and so are city-size distributions ence across U.S. cities in their proportions Externalities imply that history matters. of managers and production workers. That is, modern-day location patterns for Although the largest cities already housed an industry are strongly influenced by the more managers, there was no clear ranking historical industrial environment of cities by city size. By 1980, however, the differences and thus by the localization economies. across cities had increased substantially, and Such intangibles include the local stock a clear ranking by size had emerged. Larger of knowledge relevant for an industry or cities had become specialized in manage- a labor force with specific acquired skills. ment and information-intensive activity, Two otherwise-identical enterprises in the which benefit from face-to-face contacts, same city could benefit differently from the and smaller cities had become specialized local agglomeration depending on how long Scale Economies and Agglomeration 139 BOX 4.4 When sowing and reaping happen in different places: rising interdependence of cities Urban specialists and economists have tories as the market adjusts to the whims services or advertising agencies; for pro- long debated whether specialized or and fancies of fashion. duction facilities, this means places with diverse cities are more conducive to Clusters of similar rms are sometimes other such plants. Headquarters are usu- growth. Cities that are narrowly special- promoted as the best environment for ally in bigger cities, because professional ized create greater economies of agglom- innovation. But studies nd instead services tend to exhibit greater economies eration, so a rm's productivity increases that diverse metropolises do better in of agglomeration, are less land-intensive, with proximity to similar rms. Mean- breeding new products and processes. and employ highly educated employees while, a diverse mix of activities makes For example, the adoption of computer- willing to pay for big-city amenities. If land cities more likely to grow, particularly in controlled machinery for cutting metals markets work well, the ensuing increase in new sectors. The main conclusion: both has been faster in situations in which land prices prompts production establish- diversity and specialization are important, many rms (ranging from furnace ments to relocate to smaller, more special- but at di erent points in a rm's life cycle. manufacturers to aircraft producers) have ized towns and cities. A "balanced" urban system is not one in similar technical needs but are not direct Cities in the United States provide a which all cities are similarly specialized or competitors. Firms for which innovating is good illustration. In 1950 the ratio of man- diversi ed, but one in which both diversi- important (such as electronics producers) agers to production workers was similar ed and specialized cities coexist. prefer diversity during the early innova- across cities of di erent sizes. By 1990, For young rms, urban diversity is more tive phases, and then they relocate to however, cities with between 75,000 and important. A new businessman may not specialized cities for mass production. For 250,000 people had 20 percent fewer know all the details of the product to be manufacturing and services, unlike agri- managers per production worker than made, what components to use, where to culture, "sowing" and "reaping" can take the national average; cities with 1.5 to 5 source them, which workers to hire, and place in di erent locations. million people had 20 percent more man- how to nance the venture. Firms using Just as product development and mass agers per production worker; and those similar technologies in di erent sec- production increasingly take place at larger than 5 million people were 50 per- tors are more likely to share information di erent locations, so too does manage- cent above the national average. A similar about new practices and technologies ment and production. Half a century ago trend can be seen in other countries such than rms in the same sector. For rms in the di culties associated with managing as France and Germany. more standardized or mature industries, businesses from a distance made rms Policy makers should be aware of these urban specialization is more important. keep their headquarters and manage- developments. Since this growing interde- These rms typically bene t less from ment o ces close to their factories. Fall- pendence manifests itself in plant reloca- the exibility from urban diversity, and ing transport and communications costs tions away from large cities, governments by locating in a specialized environment; have made it much easier to manage pro- may be tempted to take away resources they can better reap the bene ts of urban duction from far away (see chapter 6). from them. This would kill the goose that agglomeration economies. For example, As a result, many rms have separated lays the golden eggs, since such reloca- auto rms in Detroit lower their costs management and production spatially, tions to smaller specialized cities are just a by sharing parts suppliers, and garment searching for the best possible conditions later part of a life cycle of rms that large, manufacturers in cities like San Pedro Sula for each. For headquarters, this means diverse cities helped give birth to. in Honduras bene t from thick labor mar- locations with other headquarters where kets that help workers move between fac- these rms can, for example, share legal Contributed by Diego Puga. each has been in the city. Similarly, two oth- known as Zipf's law: a city's population erwise identical cities would offer different size relative to the primate city is inversely types of external economies depending on proportional to its rank in the national their histories.91 hierarchy of cities.92 There is also persis- The influences of history and special- tence in the industrial concentration in ization are consistent with the observed specific cities.93 stability in the relative city-size distribu- Among mature industries the persis- tion and the industrial concentration in tence in employment patterns across cities specific cities over time. Within countries is high over time, and the convergence in the relative sizes of cities tend to remain individual industry employment across cit- unchanged. Among urban specialists, ies is slow. This persistence occurs despite this phenomenon is often represented as a high plant and employment turnover rates recurring relationship between a city's size for individual manufacturing industries, relative to the largest city in the country, and despite strong evidence that plants 140 WORLD DEVELOPMENT REPORT 2009 began in Boston in 1924, when the Mas- sachusetts Investment Trust was founded. BOX 4.5 Cities continue to thrive as telecommunication Today, Boston is still home to almost costs fall a third of U.S. employment in mutual As telecommunications improve, co-authorships all rose. Better tele- fund and asset management services. The cities become more important as a communications increase long-range Hartford insurance industry began even platform for interactions and knowl- interactions, but not at the expense earlier, in the late-eighteenth century. edge transfers. Recent studies in the of local interactions. Local merchants insured each other's United States and Japan document As ideas become more complex overseas trading expeditions by sharing the complementary roles of telecom- and di cult to communicate, the their profits and losses. These informal munications and face-to-face interac- value of intensive face-to-face inter- tions: people closer to one another action rises, and cities become even arrangements eventually grew into large physically call each other more often. more important. And if cities are insurance companies, starting with the One interpretation is that face- centers of telecommunication tech- Hartford Fire Insurance Company in 1810. to-face interactions generate more nology, improvements in information Other major Hartford insurers, including demand for telephone interactions. technology will increase their eco- Aetna, Connecticut General, and Travel- Since the mid-1980s, when faxes and nomic role. The rise of the New York ers, were founded in the early and middle e-mail became prevalent, business multimedia industry may signal the 1800s. Hartford is still known today as travel has risen more than 50 percent. comparative advantage of large cities Another evidence of increased face- in facilitating the di cult information the "insurance city," with a wide range to-face interactions with falling tele- ows in cutting-edge industries. In of related services such as life insurance, communication costs is the phenom- the developing world, the rise of Ban- medical insurance, fire/marine/casualty enal growth of co-authored articles galore is a case in point. insurance, and pension funds.98 in economics--from 12 percent in the 1960s to 56 percent in the 1990s. Sources: Gaspar and Glaeser 1998; Huber Apprehension of market forces Local, out-of-state, and international 1995; Sassen 1991; and Gottman 1977. Over the past century, producers and work- ers in the developing world have sought, and often found, their fortunes in towns and relocate as local wages and demand condi- cities. In the past three decades, researchers tions change.94 Historically, some cities have have analyzed and increasingly understood undergone major sectoral overhauls, but the gains from urban agglomerations of all they have tended to be the exceptions.95 shapes and sizes. But it is not yet clear that The persistence of an industry's employ- policy makers appreciate the sheer strength ment concentration in specific cities, which of these market forces and the benefits that implies the "lock-in" of industrial structure, come from harnessing them. can be explained by localization economies. More than half the developing world's These cities can better compete for and, governments surveyed in 2005 by the UN over time, retain plants and employment in Population Division expressed a desire that industry. A larger scale of own-indus- to make major changes to the spatial dis- try activity historically means that firms in tribution of their populations. Almost that locality today will operate more pro- three-quarters of developing country offi- ductively with greater accumulated knowl- cials expressed a strong desire to imple- edge about technology, sources of supply of ment policies to reduce migration into different quality inputs, and local culture urban areas or to take actions to reverse and its effect on the legal, business, and rural-urban migration trends.99 Many in institutional climate. These localization developed countries are equally fearful advantages are relevant for more traditional of urbanization in developing countries. manufacturing industries.96 They explain "The explosive growth of cities around the the longevity of many industrial clusters in world--especially the rise of huge, nation- certain locations--such as the world-class sized Third World metropolises--has cutlery cluster in Solingen, Germany, since U.S. scientists and officials worried. Chief 1348.97 among their concerns: "megacities increas- There is also evidence of persisting con- ingly will serve as incubators of diseases, centration of particular services in specific economic disruptions, and endless politi- cities. The American mutual fund industry cal crises."100 This worry was reflected in Scale Economies and Agglomeration 141 the goal of the 2006 World Urban Forum, Figure 4.1 The urban premium for household consumption can be considerable held to discuss "mega-cities with mega Maldives problems."101 The prevalent view was that Bolivia "cities in the developed world have histori- Angola Rwanda cally been engines of economic growth. Guinea But many cities in the Third World are so Peru Burundi dysfunctional that they have become drags Paraguay on economic progress."102 Gambia, The Mexico Some of the favored solutions: slow the Senegal massive migration to cities, decongest the Honduras largest cities in the developing world by Cambodia Mali establishing new cities, and make the big- Vietnam gest cities centers for cleaner high-technol- Brazil El Salvador ogy activities. These solutions all represent Guatemala a potentially costly misreading of the mar- Nepal Zambia ket forces that drive the spatial transforma- Djibouti tions for economic development. Chad Bhutan Uganda A misplaced fear of urbanization Colombia Ghana Economic activities in urban areas account Thailand Mauritania for as much as 80 percent of GDP in more Philippines urban and industrialized countries. The Cameroon urban share of economic activity in less Burkina Faso South Africa developed countries is about 50 percent. Morocco Just the 10 largest metropolitan areas in Bangladesh Sri Lanka Mexico, which account for a third of the Russian Federation country's population, generate 62 percent Malawi of its national value added.103 In Vietnam, Madagascar Ethiopia where the share of the urban population is India Indonesia 30 percent, the share of cities in national Tanzania output is 70 percent. In China 120 cities Nicaragua Swaziland account for three-quarters of the country's Côte d'Ivoire GDP.104 Clearly cities make a dominant Timor-Leste contribution to economic production, even Guyana Moldova in poor and middle-income countries. Costa Rica There is also ample evidence that urban Egypt, Arab Rep. of Chile areas in developing countries, including Ecuador those in the poorest countries in Africa, Benin Nigeria deliver external economies. Consumption Mongolia in urban and rural households in a broad Georgia Bulgaria cross-section of developing countries shows Pakistan that people with similar observable charac- Armenia Romania teristics enjoy higher consumption attribut- Turkey able purely to their urban location. The gains Jordan Croatia range from 2 percent in Hungary, the Krygyz Ukraine Republic, and Poland, to 30 percent in Costa Mozambique Rica, Ethiopia, India, Romania, and Tanza- Congo, Dem. Rep. of Tajikistan nia, and to more than 80 percent in Angola, Poland Bolivia, and Rwanda (see figure 4.1). Hungary Kyrgyz Republic These magnitudes make it futile for 0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.0 policy makers to try to restrict the flow of Ratio of urban consumption premiums people to urban areas. Even when restric- to rural consumption levels tions have stemmed migration flows, the Source: WDR 2009 team calculations, using 120 household surveys in 75 countries. 142 WORLD DEVELOPMENT REPORT 2009 economic costs have been high. China's attractive to investors, a city must satisfy policies to restrict rural-urban migration the demands of its dominant or growing until the late 1990s stunted urbanization, industries for both real estate and facili- with between half and two-thirds of Chi- ties. For example, professional services and nese cities remaining too small. For the financial services require large amounts of typical city in China, being too small is office space, which can be more efficiently estimated to result in a loss of about 17 per- provided vertically in high-rise office cent in net output per worker; for at least a buildings. Manufacturing requires large quarter of the cities, these losses may range amounts of land for factories to produce between 25 and 70 percent.105 goods, and for warehouses to store products and materials. And the recreation, tourism, A misplaced preoccupation with size, and entertainment sectors require highly not function, of cities visible, pedestrian-friendly areas of cities A city's prospects for prosperity and even and retail space. survival are determined by how nimbly The ability and ease of a city to adapt its the same piece of land is adapted to chang- land to different uses according to chang- ing market demands. Given that land is an ing market needs will enable its sustain- immobile factor critical to the production able growth. The last 800 years in Hong of any activity, the real estate choices that Kong, China, and the last 300 in New York cities provide influence the magnitude of show the importance of markets in signal- external economies and the nature and ing and implementing this urban renewal specialization of city economies. To be (see boxes 4.6 and 4.7). In New York the mercantile trade grew out of the early shipping industry. In turn, the mercantile trade industry would help give birth to the BOX 4.6 Hong Kong, China: market forces led the way, city's modern finance industry. Traders in government followed New York City in the late-nineteenth cen- Hong Kong, China, with a land area ties. For many decades, the private tury thrived by sharing access not only to of about 1,000 square kilometers, sector showed more commitment to physical transportation infrastructure (the less than a quarter the size of Rhode and interest in urban redevelopment. harbor, canals, and railroads), but also to Island, started out as a shing village. Between 1960 and 1980 the gov- intermediate inputs of specialized services In the 1200s Hong Kong, a hilly and ernment experimented with urban not available elsewhere (such as scheduled barren island, saw its rst population renewal and comprehensive redevel- sailings, wholesalers, and ship brokers). boom as Chinese ed the mainland to opment to improve environmental escape war and famine. People made conditions, tra c circulation, and Later, these inputs to trade became the a living on salt production, pearl community facilities. Over subse- foundations for shared inputs in finance, diving, and shery trades. Between quent decades, the exibility in land with maritime insurance underwriting the 1650s and the 1800s, Hong Kong use planning and the participation the subsequent basis for other forms of was also a military outpost and naval of the private sector would prove investment.106 base, and its economy continued to crucial to satisfying the demands on Cities that provide fluid land and property rely on trade. By the end of World land for housing, commerce, industry, markets and other supportive institutions-- War II in 1945, the population in Hong transport, recreation, and commu- Kong, China, had been reduced to nity use. This combination enabled such as protecting property rights, enforcing less than half the prewar total of 1.6 Hong Kong, China, to ourish into the contracts, and financing housing--will more million. regional center of business and nan- likely flourish over time as the needs of mar- In the 1950s and 1960s, Hong Kong cial services that it is today. kets change. Successful cities have relaxed took up manufacturing buttons, Consistent with the tradition of min- zoning laws to allow higher-value users to arti cial owers, umbrellas, textiles, imal government intervention in Hong bid for the valuable land--and have adopted enamels, footwear, and plastics. Kong, China, the private sector has flexible land use regulations to adapt to their Squatter camps provided homes for been the driving force behind urban the masses. The camps led to disas- transformation. The government con- changing roles overtime. ters--like the Shek Kip Mei re--until tracted out urban redevelopment to a The benefits of agglomeration econo- the governor responded by putting specialist organization dominated by mies arise from the density of economic up multistory residential buildings. private development interests. activity. These are the advantages for an Conditions in public housing were information technology startup locat- basic, with communal cooking facili- Source: Adams and Hastings 2001. ing in Silicon Valley or a bookstall owner Scale Economies and Agglomeration 143 BOX 4.7 Reinvention and renewal: how New York became a great city New Amsterdam was founded as a Dutch New York characterized SoHo as a com- a generation, the transformation in the colony in 1614. It passed into British hands mercial slum. But today the area, once meatpacking district was as stark as the and became New York in 1664. Manhat- called Hell's Hundred Acres, is a busy contrast between night and day.b Today, tan, the Bronx, Brooklyn, Queens, and commercial and retail district and home more than 35 wholesale meat companies Staten Island were brought together to New York University.a still operate there. But the area is now in 1898 in the form we know today. Wall Street. The nancial district is also home to world-class restaurants, art Throughout its history, New York has con- one of the city's best-known and old- galleries, a fashionable retail corridor, tinually rebuilt, reinvented, and renewed est neighborhoods. Today's Wall Street and night clubs that take advantage of itself. Once a fur-trapping and shipping neighborhood is part of Manhattan Com- the enormous former factory spaces. hub because of its natural harbor, New munity District 1, which extends south Real estate prices have skyrocketed. Mr. York City is today a global nancial cen- from Canal Street to the tip of Manhattan Inconiglios paid $50 a month when he ter and a regional powerhouse in mass at Battery Park and includes Governor's moved to the meatpacking district. In media, arts, information and communica- Island. It is home to the New York Stock 2007 the Carlyle Group and Sitt Asset tion technology (ICT) innovation, and Exchange and the NASDAQ, the world's Management acquired a pair of buildings medical research. The New York metro- two largest stock exchanges. on West 14th Street for $70 million.c politan area is home to more than 18.7 The street's name was originally De Wall Williamsburg. This neighborhood million people with a GDP of $1,133 bil- Straat in reference to the Walloons, Belgian reinvented itself from a booming trade lion, making it the second-largest urban farmers who were the majority of the port to a rich industrial town after the agglomeration in the world, after Tokyo. residents living in New Netherland around Civil War. With the construction of the New York had a gross metropolitan prod- Fort Amsterdam in 1630. The beaver belt Williamsburg Bridge in 1903, many Jewish uct of $950 billion in 2005, making it the was the single most important commodity families who lived in Manhattan's Lower largest regional economy in the United in New Netherland. Trade encouraged new East Side crossed the East River to a bet- States. If it were a country, New York City activity in the production of food, timber, ter life in Williamsburg. When industries would be the world's seventeenth larg- tobacco, and eventually slaves. In the late- left the area in the 1960s and 1970s, Wil- est, ahead of Switzerland. At more than eighteenth century there was a button- liamsburg became an immigrant ghetto. $56,000, it has the second highest per wood tree at the foot of Wall Street under But the cheap rent also made the neigh- capita production in the world. which traders and speculators gathered borhood an artistic hub. The neighbor- A tour of four neighborhoods reveals to trade informally. In 1792 this arrange- hood evolved into a mix of Italian, Polish, the city's versatility and vibrancy. ment was formalized with the Buttonwood Hispanic, and Hasidic residents. In 2005 SoHo. In the 1700s SoHo was farmland. Agreement, which laid the groundwork for New York City approved zoning changes By the early 1800s it was primarily resi- the New York Stock Exchange. that would allow for open spaces, parks, dential, inhabited by the wealthy and, Meatpacking district. In 1969, when a ordable housing, and light industry. soon after, by the middle class. In due Vincent Inconiglios moved to a loft on Today, prices average $700­$900 per time rapid development attracted many Gansevoort Street in the meatpacking square foot, and prominent waterfront businesses. Hotels, theaters, stores, man- district, it was a no man's land. The neigh- developments range in the millions.d sions, minstrel halls, casinos, and brothels borhood was de ned by a stench that appeared along Broadway. Starting in overpowered the senses. Down the street Sources: Seeman and Siegfried 1978; Shaw the 1880s the textile industry settled in from Mr. Inconiglios was a pickle fac- 2007; Biedermann 2007; Lynch and Mulero 2007. the area. By the 1950s artists ocked to tory, and an importer of Spanish melons a. Seeman and Siegfried 1978. the area because of low rents, a result of occupied the shop downstairs. The area b. Shaw 2007. people, industry, and commerce shifting was teeming with barrels of bones, meat, c. Biedermann 2007. uptown. In October 1962 the City Club of and men in bloody white coats. Within d. Lynch and Mulero 2007. locating close to other bookstalls on Dada- they are moving to has lower wages and bhai Naoroji Road in Mumbai, India. cheaper land. While the financial sector of London is But bigger city size and economic den- largely concentrated in a few square miles sity bring their own problems. For people of the City and Canary Wharf, financial and firms, city living comes at a price in firms also benefit from being located both developing and developed countries. anywhere in Greater London. Firms ben- Traffic in central London moves at only efit from locating close to other firms in 11 miles per hour107--the same speed as either the same or different industries and horse drawn carriages a hundred years ago. unless all of them move together, they will Beijing is notorious for its pollution-in- become less profitable, even if the location duced smog. Land in Mumbai is among the 144 WORLD DEVELOPMENT REPORT 2009 most expensive in the world. High levels of that reduce the gains from agglomeration crime are an accepted feature of city living economies. around the world. Millions of city dwellers But restricting the growth of cities is live in overpopulated slum housing, with not the answer. There is no evidence that little or no access to basic amenities and the agglomeration economies of megaci- services. These are the costs of density, the ties have been exhausted. Indeed, evidence diseconomies of agglomeration. suggests that the growth of vehicles in the The main source of diseconomies is the developing countries is increasing with per paucity of land in places where agglomera- capita income along a path similar to that tion economies take hold. Land is limited followed by the richer countries.108 The and as economic growth occurs, it has to problem has more to do with the spatial be used with increasing intensity. Take structure of the city and investments in Manhattan in New York City, which has an infrastructure. Vehicle ownership is ris- area of less than 35 square miles. In 1800 ing 15 to 20 percent annually in much of it had a population density of just under the developing world.109 But most coun- 3,000 people per square mile. By 1850 this tries have not matched this growth with a had risen to about 23,500, peaking in 1910 parallel expansion of transportation infra- with a population density of more than structure, so traffic congestion is severe. 100,000. Today, the population density is Cities in developing countries only devote about 70,000. With land in fixed supply, its half as much land space to roads as in the use eventually can offset any further ben- United States. But it is not just a matter of efits from agglomeration economies. The increasing this capacity. In cities such as way to offset the fixed supply of a factor of Bangkok and Manila, it is the management production is to substitute other factors for and the use of road space that is important. it, and the rise of skyscrapers in many large Part of the problem is that in many cities urban areas is an illustration of this substi- the responsibility for road infrastructure tution of capital for land. The building of has devolved from central to local govern- subway systems in many of the developed ments, which do not always have the neces- countries' larger cities is another example. sary resources. But such substitution has its limits, and the Combined with the differing propensi- increasing shortage of land in cities leads to ties of industries to benefit from agglomer- higher rents and congestion costs for work- ation economies, the resulting constraints ers and firms. explain why the spatial distribution of Better transport can, by reducing the economic activity within a country is not economic distance to density, in essence restricted to a single center, but rather make land a less-binding resource. Indeed, consists of multiple centers of differing with the long-term decline in transporta- sizes. For policy makers the challenge is to tion costs, cities have expanded. In 1680 best relax the constraints generated by the London was only 4 square miles and, congestion and overcrowding of land and because of the difficulties of traveling, resources so that the benefits of agglom- more than 450,000 people were crammed eration can be maximized. In many cases into this small area. By 1901 the city had these constraints have been tightened by expanded to 24 square miles, and the misguided land use policies and planning average population density had fallen to failures, only adding to congestion (see 79,000. In 2001 London's 627 square miles chapter 7). had a population density of 13,203 people per square mile. An expanding city meant A misplaced fascination that millions of commuters have to be with "new" cities transported from the suburbs, large vol- The land Chicago was built on is not all umes of retail goods have to be delivered that different from the more sparsely to shops, and manufactured products developed places around Lake Michigan. have to be shipped out. All of this leads Yet the difference in economic production to congestion or diseconomies of scale and household earnings between Chicago Scale Economies and Agglomeration 145 and other settlements on the lakefront for economic reasons, but some were cre- in Wisconsin and Indiana is stark. And ated for political reasons. Have these new along the 10-hour drive through Texas towns and cities, met their goals? Generally on Interstate Highway 75, wages and land not. rents spike in Fort Worth, Austin, and San Antonio and drop off sharply at points in · New cities do better when they are between. It is hard to reconcile these huge located near larger successful cities. But differences in economic density with the they often suffer from the same govern- minor differences in physical geography; ment-related failures that led the gov- it is as if the areas of Fort Worth and San ernment to establish them, especially the Antonio cast a shadow over the points in failure to manage large cities well. That is, between. A better understanding of eco- governments that do badly in managing nomic geography, characterized by exter- large old cities also do badly in managing nal economies, is required to harness small new cities. economic forces. But it is not always obvi- · New cities attract residents, sometimes ous that developing country governments even more than anticipated, but often understand economic geography or appre- not the people intended. That is, gov- ciate these forces. ernments can set up (noncapital) cit- A survey of new city initiatives in the ies, and they sometimes become viable, Arab Republic of Egypt, Brazil, Hun- but not for the reasons the government gary, India, and República Bolivariana de envisaged. Venezuela is sobering. Brazil transferred · These cities attract people because of its capital city from the coast to the mid- the circular causation that the new eco- western interior more than 900 kilometers nomic geography emphasizes: work- away. República Bolivariana de Venezuela ers and entrepreneurs come to seek picked Ciudad Guayana in the 1950s, a city markets, and then more people come in the southern part of the country, to be because this is where the markets are. the industrial "growth pole" of the central But there may be huge opportunity and southern region and to attract people costs, because the counterfactual could and jobs from the already rapidly growing be more organic growth of settlements. metropolitan region in the north. In many That is, it makes sense for private agents formerly planned economies, the more to come to these cities since others are common practice was building industrial already there, but large efficiency losses towns to accelerate industrialization. In may result from the country's point of Hungary, Dunaujvaro was designed as a view. Once a "bad" location is picked, "steel town," Tiszaújváros as a "chemistry it may not fail entirely because of cir- town," and Kazincbarcika as a "mining and cular causation, but that means the eco- heavy industry town." The Soviet Union nomic costs of the mistake are greater, built Magnitogorsk into a steel town in an not smaller, since the country will pay area with huge reserves of iron ore to chal- these costs for a long time. lenge its capitalist rivals. · New noncapital cities that seem to suc- Some new towns were built around met- ceed are those where the purpose and ropolitan areas to alleviate the pressures location are chosen over time by markets that the large cities faced. Navi Mumbai and in cases in which the government was established in 1972 with the hope of hastens the pace of growth by coordinat- developing a twin city for Mumbai, and to ing investments in infrastructure, hous- decongest Mumbai. Egypt started a com- ing, and general governance. prehensive new town construction program around Cairo and away from Cairo to create For these reasons, cities and towns should a "new population map of Egypt" starting be seen as market agents that, just like firms in the 1970s, and the construction is still and farms, serve market needs. ongoing. Many of these cities were created CHAPTER 5 Factor Mobility and Migration T he largest movement of people in the conditions.2 In a globalizing economy capi- world occurs every year around the tal is mobile and will move quickly. Labor, beginning of February, as millions in by contrast, tends to be less mobile for cul- China travel to be with their families for the tural and linguistic reasons. Second, a strong Lunar New Year. In 2006, to mark the Year policy consensus supports the free flow of of the Dog, about 11 million people traveled capital for foreign direct investment, even if out of Shanghai alone, and 10 million trav- this consensus is not always fully manifested eled into the city; 60 million people traveled in the policies of the many countries where on the last day of the festivities. In Febru- external and internal obstacles remain. ary 2008 ice and snow storms frustrated the Relative to capital, labor is subject to more plans of an estimated 200 million people political restrictions and to explicit and trying to travel across China to be with implicit barriers. Some novel insights come their families for the New Year. Similarly, in from considering agglomeration economies the United States during the Thanksgiving and human capital together. Based on these period, millions take to the roads, airports, insights, this chapter makes a case for facili- bus and train stations. The number of trips tating the voluntary movement of people. longer than 50 miles increases by half, with Textbooks teach us that the factors of about 10 million people a day traveling over production--capital and labor--move the holiday weekend, almost twice the daily to places where they will earn the highest average during the rest of the year.1 The ris- returns, and that these are the places where ing volume of holiday travelers in almost each factor is scarce. But by recognizing that every country, rich or poor, is a telling reflec- increasing returns to scale are important, tion of just how many people live and work policies can be made better. Unlike unskilled in a place other than where they were born. labor and physical and financial capital, This chapter is about the mobility of skilled labor--embodying human capital, labor and capital, how their movements a person's education, and endowment of help to concentrate economic activity, and skills and talent--earns higher economic how these flows mitigate differences in wel- returns where it is abundant, not scant. This fare that can accompany economic concen- explains the clustering of talented people tration. It emphasizes movements of labor, in cities, the migration of entrepreneurs especially, for two reasons. First, although to leading areas within countries, and the many countries and regions are still thirsty rising number of skilled migrants moving for investment, national reforms and inter- to wealthy countries, all places where their national agreements since the 1970s have skills seem plentiful but are nonetheless eliminated most restrictions on the flow of highly rewarded. Recognizing the growing capital. The scarcity of capital in some places benefits to human capital in areas where it now has less to do with actual barriers and has already accumulated changes the think- 146 more to do with unfavorable investment ing about how governments should try to Factor Mobility and Migration 147 raise growth and achieve spatial conver- theories viewed the outmigration of gence in living standards. skilled people as both socially traumatic But policy makers in many developing and an economic loss. New theories rec- countries--particularly in South Asia and in ognize that migration, when driven by Sub-Saharan Africa--have been conditioned economic forces, is a positive and selec- by an early literature on migration to worry tive process. The interactions between about the specter of rising urban unemploy- agglomeration and labor migration ment, overburdened city services, social ten- power places forward. sions in economically vibrant areas, and a · The policy challenge is not how to keep "brain drain." As a result, many countries still households from moving, but how to restrict the movement of people. Yet direct keep them from moving for the wrong and indirect restrictions, although not effec- reasons. Instead of trying to fight the pull tive at stemming the flow of people, create of agglomeration economies on workers unnecessary friction and impose the cost of and their families, governments should forgone opportunities for economic growth work to eliminate the factors that push and convergence in living standards. people out of their home areas. By doing Although researchers are now less skep- so they can improve the quality of migra- tical about the benefits of labor migra- tion and encourage economic growth. tion, policy makers in both developing Labor mobility driven by economic rea- and developed countries are not so sure. sons leads to greater concentration of What can they learn from each other? This people and talent in places of choice and chapter documents the disconnect between adds more to agglomeration benefits in the implications of recent research and these places than to congestion costs. the migration policies in developing and developed countries, showing how they are From mercantilism to globalization changing. to autarky, and back again Keep in mind three points: Restrictions on the flow of capital, labor, · The facts about labor migration can be and goods fragmented the world economy surprising. Although international migra- betweenthetwoworldwars,butglobalization tion still captures the greatest attention in picked up speed after the end of the Cold War the media, by far the largest flows of people in 1990, loosening the restrictions and inte- are between places in the same country, grating the world economy. Capital mobil- and not from villages to cities, but from ity within and across countries increased. economically lagging to leading rural International labor mobility--particularly areas.3 Although the movement of people unskilled labor--declined after the mass to cities is on the rise, particularly in South movements in the nineteenth century and and EastAsia,the most sustained pattern of only recently began to rise. But the mobility internal mobility within developing coun- of people within countries has accelerated. tries has been from lagging rural areas, So, for the movement of labor over the last like Western Kenya and Bihar in India, to century, distance has diminished, but divi- leading rural areas in those countries, like sions not only have increased (many more the Central Highlands and Punjab, and a borders) but indeed may have become more large share of this migration is temporary.4 obstructive (many more restrictions). And when people move across national borders, they do not go far.5 Most interna- Capital flows--up sharply since the 1970s tional migration takes place within world The mobility of capital across borders, par- regional "neighborhoods," particularly ticularly investment capital, has increased between developing countries.6 since the 1970s. The world is returning to · Movements of capital and labor are an age of capital mobility abandoned at the driven by the benefits of agglomeration. onset of World War I. From 1880 to 1914, a Early migration theories were based on growing share of the world economy oper- surplus labor, fixed "exogenous" rates ated under the classic gold standard and a of growth, and job creation--and these global financial market centered in London. 148 WORLD DEVELOPMENT REPORT 2009 The gold-standard fixed exchange rates and national aims, domestic policy goals, and underpinned a stable and credible regime "beggar-thy-neighbor" trade strategies that that enforced discipline on countries. Interest encouraged strict capital controls. Interna- ls. rates tended to converge and capital to flow tional capital flows petered out, and invest- with relative eaease across borders, const nstraineded ment from abroad was viewed with suspicion. only by the limits of technology. Many rapidly So prices and interest rates across countrie fell ic ies industrializing countries outside Europe--inrop out of sync. Even during the Bretton Woods the Americas and in Asia--took part in an era from the end of World War II to 1971, increasingly global economy.7 re lob as countries attempted to rebuild the global The fluid economic environme onom ronment was economy, fears of mobile capital that had mobi destroyed by two world wars and the global wo taken hold during the interwar years proved economic retraction in between. From 1914 difficult to dispel. Indeed, capital controls to 1945, monetary policy was used to pursue were sanctioned to prevent currency crises. we Figure 5.1 International capital flows have surged since the 1970s Gross private capital flows Percentage of GDP 35 World Sub-Saharan Africa South Asia 30 Middle East & North Africa Latin America & Caribbean 25 Europe & Central Asia East Asia & Pacific 20 15 10 5 0 1970 1975 1980 1980 1985 1985 1990 1995 2000 2005 Year Source: World Bank 2007j. Figure 5.2 A large share of capital now as direct investment flows as direct investment Foreign direct investment, net inflows Percentage of GDP 5.0 World Sub-Saharan Africa South Asia 4.5 Middle East & North Africa Latin America & Caribbean 4.0 Europe & Central Asia East Asia & Pacific 3.5 3.0 2.5 2.0 1.5 1.0 0.5 0.0 1970 1975 1980 1980 1985 1985 1990 1995 2000 2005 Year Source: World Bank 2007j. Factor Mobility and Migration 149 But the growing volume of trade made an estimated 48 million emigrants, between it difficult to constrain the flow of capital, 10 and 20 percent of the population, left and in the early 1970s, the constraints began Europe (see table 5.1).12 Unlike international to loosen. Fixed exchange rates were aban- migration today, the movement of people doned, creating an international economic across borders in the first and second periods environment that could accommodate capi- of labor migration was not driven by a lack tal flows and market development. Political of economic growth or development in the stability, structural reforms, and regula- sending countries. Indeed, the first coun- tory structures lowered the risks to foreign try to industrialize and the most advanced investment in developing countries, and at the turn of the twentieth century--Great capital markets responded with enthusiasm Britain--was by far the largest sending coun- (see figures 5.1 and 5.2). By 2000, capital try. Economic analysis shows a positive cor- mobility returned to levels seen in 1914.8 relation between emigration and the extent of Capital has become the most mobile fac- industrialization in the sending country.13 tor of production. Converging real interest A long period of autarky and economic rates, declining spreads between deposit and nationalism began in 1910. Unprecedented lending rates, and shrinking risk premiums restrictionswereplacedontrade,investment, on the sovereign debt of developing coun- and immigration, stifling the international tries are evidence of an international envi- movement of capital and labor. The trickle ronment in which capital can go where it of international migrants consisted mainly wants to, even if it does not always go where of refugees and displaced persons, unrelated people wish it would. Indeed, recent com- to economic development. parisons of the marginal product of capital The postindustrial period of migration between high-income and lower-income began in the 1960s, characterized by new countries show little evidence of friction preventing the flow of capital to poor coun- Table 5.1 In the late-nineteenth century most international migrants came from better-off Europe tries. Instead, the lower capital ratios in poor Top sending countries in 1900 and 2000 countries are explained by lower efficiency Percentage of Top emigrant- Percentage of and a lack of complementary factors.9 Top emigrant-sending sending country's sending countries sending country's countries in 1900 population in 1900 in 2000 population in 2000 Labor flows across borders--blocked for British Isles 40.9 Mexico 10.0 much of the twentieth century Norway 35.9 Afghanistan 9.9 In a pattern similar to that of capital flows, Portugal 30.1 Morocco 9.0 from a peak in the late-nineteenth cen- Italy 29.2 United Kingdom 7.1 tury, the mobility of labor across borders declined for most of the twentieth century, Spain 23.2 Algeria 6.7 with the rise of economic barriers at the Sweden 22.3 Italy 5.7 onset of the Great Depression and World Denmark 14.2 Bangladesh 5.0 War II. Geographers have long recorded Switzerland 13.3 Germany 4.9 the movement of humankind, from the Finland 12.9 Turkey 4.5 earliest migrations out of Africa to Europe and Asia,10 to the resurgence of movement Austria-Hungary 10.4 Philippines 4.3 across borders. They categorize the modern Germany 8.0 Egypt, Arab Rep. of 3.5 history of international migration into four Netherlands 3.9 Pakistan 2.4 distinct periods: mercantile, industrial, Belgium 2.6 India 0.9 autarkic, and postindustrial.11 Russian United States 0.8 During the mercantile period, from 1500 Federation­Poland 2.0 to 1800, the movement of people around the China 0.5 France 1.3 world was dominated by Europeans. Agrarian settlers, administrators, artisans, entrepre- Europe 12.3 neurs, and convicts emigrated out of Europe Japan 0.9 in large numbers. During the industrial Total (Europe and Japan) 11.1 Total (of countries 1.9 period that followed--sometimes referred to listed) as the first period of economic globalization, Sources: Massey 1988, Parsons and others 2007, in Ozden and Schiff 2007. 150 WORLD DEVELOPMENT REPORT 2009 forms, no longer dominated by flows out of the top, each accounting for between 3 Europe. People began to move from lower- million and 4 million emigrants. income countries to wealthy countries, with The volume and flow of international a surge in migrant labor from Latin America, migration is no longer mainly associated Africa, and Asia. In the 1970s, countries that with population growth or demographic had been major sources of migrating labor pressure. Unlike the 1960s and 1970s, to Northern Europe and the Americas-- international immigrants are not from such as Italy, Portugal, and Spain--began the poorest, least developed countries. to receive immigrants from Africa and the Voluntary international movements of Middle East. The growing wealth of oil-rich people tend to originate from countries countries in the late 1970s made economies with rapidly growing economies and fall- in the Persian Gulf new destinations. And by ing fertility rates. Emigration today is the the 1980s, migration to East Asian countries outcome less of desperation and more of spread beyond Japan to Hong Kong, China; integration.15 the Republic of Korea; Malaysia; Singapore; The pattern of international migration is Taiwan, China; and Thailand. also shifting, from South-North to South- Today, about 200 million people are for- South.16 Although the top three receiving eign born, roughly 3 percent of the world countries are members of the Organisation population.14 The flows of new interna- for Economic Co-operation and Develop- tional migrants have varied--from a ment (OECD)--the United States, Ger- 2-percent increase between 1970 and 1980, many, and France, in that order--Côte to 4.3 percent from 1980 to 1990, and to d'Ivoire, India, the Islamic Republic of Iran, 1.3 percent from 1990 to 2000. Poor and Jordan, and Pakistan are now among the middle-income countries now send the top 15 destinations. But migration of labor most emigrants, led by Bangladesh, China, from the low- and middle-income countries the Arab Republic of Egypt, India, Mexico, of the South to the wealthy countries of the Morocco, Pakistan, the Philippines, and North is still large, 37 percent of interna- Turkey (see table 5.1). But Italy, Germany, tional migrants in 2000. Movement between and the United Kingdom still rank near Northern countries made up 16 percent of Table 5.2 Close to home: the largest international flows of labor are between neighboring countries Percentage of world migrants recorded as a bilateral movement between pairs of countries/regions, circa 2000 Destination countries/regions Countries/ UE15 & AU & HI regions of origin USA Canada EFTA NZ Japan MENA LAC ECA MENA AFR EAP SAS Total USA n.a. 0.16 0.34 0.04 0.02 0.03 0.43 0.04 0.05 0.03 0.15 0.02 1.29 Canada 0.54 n.a. 0.10 0.02 n.a. 0.01 0.02 0.01 n.a. 0.01 0.01 0.01 0.74 EU15 & EFTA 2.22 0.98 5.59 1.13 0.01 0.14 0.68 0.78 0.16 0.39 0.20 0.19 12.47 AU and NZ 0.06 0.02 0.16 0.23 n.a. n.a. n.a. 0.01 n.a. 0.01 0.03 0.01 0.55 Japan 0.28 0.02 0.06 0.02 n.a. n.a. 0.04 0.01 n.a. n.a. 0.05 0.01 0.50 HI MENA 0.10 0.03 0.06 0.01 n.a. 0.12 n.a. 0.02 0.72 0.01 0.04 0.03 1.14 LAC 10.22 0.36 1.45 0.05 0.13 0.10 2.07 0.17 0.08 0.14 0.14 0.25 15.15 ECA 1.27 0.39 4.75 0.26 n.a. 0.92 0.07 16.98 0.33 0.34 0.18 0.41 25.88 MENA 0.47 0.17 2.85 0.10 n.a. 1.49 0.04 0.16 1.79 0.28 0.05 0.12 7.52 AFR 0.41 0.12 1.58 0.10 n.a. 0.25 0.02 0.11 0.18 7.00 0.03 0.16 9.97 EAP 3.32 0.71 1.09 0.63 0.54 0.48 0.06 0.14 0.14 0.09 3.86 0.27 11.32 SAS 0.83 0.31 1.13 0.12 0.01 2.66 0.02 0.13 2.07 0.14 0.37 5.67 13.46 Total 19.71 3.25 19.14 2.72 0.74 6.22 3.45 18.56 5.53 8.44 5.10 7.15 100 Source: Parsons, Skeldon, Walmsley, and Winters 2007. Notes: AFR = Africa; AU = Australia; EAP = East Asia and Pacific; ECA = Europe and Central Asia; EU15 = European Union 15; EFTA = European Free Trade Association; HI MENA = High-income countries in the Middle East and North Africa region; LAC = Latin America and the Caribbean; NZ = New Zealand; SAS = South Asia; n.a. = not applicable. Factor Mobility and Migration 151 migration and that between Southern coun- of immigrants to the United States, 20 per- tries accounted for 24 percent, with Argen- cent to France, and 10 percent to Germany tina, China, Côte d'Ivoire, India, the Islamic come from countries with which they share Republic of Iran, Jordan, Pakistan, and a border, 81 percent of immigrants to Côte South Africa as important destinations. d'Ivoire, 99 percent to the Islamic Republic There is a strong tendency for labor to of Iran, and 93 percent to India are from move between countries in the same world neighboring countries. neighborhoods, particularly for South- International migrants tend to stay South migration (see table 5.2). Migration within regional neighborhoods, particularly of labor is usually from countries with a in developing world regions, most notably in shared land border.17 While only 30 percent Sub-Saharan Africa (see figure 5.3). Almost Figure 5.3 Migrants from East Asia, Latin America, and the Middle East and North Africa go mainly to OECD countries, but most in South Asia and Sub-Saharan Africa stay close to home Destination of emigrants from countries in East Asia & Pacific Destination of emigrants from countries in Europe & Central Asia Percent Percent 100 100 90 90 80 80 70 70 60 60 50 50 40 40 30 30 20 20 10 10 0 0 OECD Other Within Other Unidentified OECD Other Within Other Unidentified countries high-income region developing countries countries high-income region developing countries countries countries countries countries Destination of emigrants from countries in Latin America & Caribbean Destination of emigrants from countries in Middle East & North Africa Percent Percent 100 100 90 90 80 80 70 70 60 60 50 50 40 40 30 30 20 20 10 10 0 0 OECD Other Within Other Unidentified OECD Other Within Other Unidentified countries high-income region developing countries countries high-income region developing countries countries countries countries countries Destination of emigrants from countries in South Asia Destination of emigrants from countries in Sub-Saharan Africa Percent Percent 100 100 90 90 80 80 70 70 60 60 50 50 40 40 30 30 20 20 10 10 0 0 OECD Other Within Other Unidentified OECD Other Within Other Unidentified countries high-income region developing countries countries high-income region developing countries countries countries countries countries Source: Ratha and Xu 2008. 152 WORLD DEVELOPMENT REPORT 2009 BOX 5.1 Regional labor mobility has been falling in Sub-Saharan Africa The rate of labor migration within devel- to move within a country: to pursue job The Southern African Development oping regions is highest in Sub-Saharan opportunities and to diversify risks to Community (SADC), a loose alliance of Africa, but it has fallen since the 1960s. income. Indeed, the economic rationale nine countries of Southern Africa formed More than 60 percent of emigrants from for movement from a lagging to a leading in 1980, coordinated development proj- Sub-Saharan countries move to other area of the same country is virtually indis- ects to lessen economic dependence on countries in the region. The higher rate of tinguishable from that for moving across South Africa during the Apartheid era. labor movement within the region rela- a border in a region like Sub-Saharan Part of this alliance was a provision for the tive to other developing world regions is Africa, where these movements are over ow of labor between member countries. partly a consequence of the large number relatively small distances and for the most The recent anti-immigrant violence in of land borders, but also of the relative part unmonitored. But many migrants South Africa is a setback for regional inte- permeability of these borders and the dif- also move across borders within a frame- gration and migration. culty of monitoring the ow of people work of formal agreements between Kenya, Tanzania, and Uganda have crossing them, despite numerous legal countries. Since the 1960s, a web of formed the East African Community (EAC), restrictions. bilateral and multilateral agreements has a regional intergovernmental organiza- Migrants represented just over 3.5 grown in an attempt to reap the bene ts tion for interterritorial cooperation with percent of the population in Sub-Saharan and control the costs of labor mobility roots extending to 1948 before indepen- Africa in 1960 but only 2.3 percent by within subregional neighborhoods. dence. The EAC, gaining strength as a 2000. In 1960 the stock of migrants rela- In West Africa governments have framework for economic integration since tive to the population was much higher in attempted to manage population move- 1999, recently introduced East African Southern Africa than in other corners of ments within the Economic Community passports and temporary passes to speed the region, but it has since fallen to about of West African States (ECOWAS), which the movement of labor. the level of migrants in Western Africa has had the most in uence on the ow The movement of labor across borders (see the table below). In Eastern Africa and composition of migration in Sub- in Sub-Saharan Africa's neighborhoods and Central Africa the stock of migrants Saharan Africa. Established in 1975, could be encouraged. During economic has fallen signi cantly. ECOWAS includes a protocol allowing the contractions, policy makers in these neigh- Voluntary migration across borders free movement of people and the right borhoods feel the same xenophobic politi- in Sub-Saharan Africa is motivated by of residence and establishment for the cal pressures as governments in rich coun- the same reasons that prompt people citizens of its member countries. tries do to favor native workers and ration public services to nonnatives. Less than one-third of governments in Sub-Saharan Sub-Saharan Africa's stock of migrants has fallen since 1960 Africa have rati ed the International Con- Per 1,000 population, by regional neighborhood vention on the Protection of the Rights of Neighborhood 1960 1970 1980 1990 2000 All Migrant Workers and Members of Their Eastern Africa 37.3 31.6 35.3 31.2 17.9 Families. To really reap the bene ts from labor mobility for faster economic growth Central Africa 40.7 44.2 35.9 20.6 16.0 with convergence across Sub-Saharan Afri- Southern Africa 49.7 40.6 33.3 34.5 30.6 ca's regional neighborhoods, much more can be done to welcome migrants and Western Africa 28.0 27.3 34.6 28.5 30.0 open channels for the ow of remittances Sub-Saharan Africa 35.6 32.8 35.0 29.0 23.0 to their home countries. Source: UN Population Division, in Lucas 2006. Source: Lucas 2006. 17 percent of recorded international migra- act as economic engines of growth in devel- tion around 2000 occurred within Europe oping regions--to Côte d'Ivoire in West and Central Asia, though a large part of this Africa, to South Africa in Southern Africa, resulted from border changes and changes to Thailand from countries in the Greater in the definition of who was "foreign born" Mekong Region in South Asia (see box 5.2), in these countries. The second highest rate and to Argentina from Bolivia, Chile, Para- of labor mobility between countries in the guay, and Peru. Distance is not the whole same region was for Sub-Saharan Africa (see story. Divisions, in the form of language and box 5.1). culture, also determine the pattern of inter- Cross-border migration within subre- national migration, with more than half of gional neighborhoods flows to countries that migrations occurring between countries Factor Mobility and Migration 153 with a common language. Of course, a com- mon language and other cultural factors BOX 5.2 Cross-border migration in the Greater Mekong reinforce the neighborhood effects. Subregion Immediately after World War II--when economies were growing rapidly, wage The Greater Mekong Subregion Remittances from Thailand to inequality was falling, and the volume of (GMS), with 315 million people, com- Cambodia, Lao PDR, and Myanmar labor movement across borders was low-- prises Cambodia, the Lao People's are estimated at $177 million to $315 international migration was not really a Democratic Republic, Myanmar, Thai- million a year. In Cambodia they land, Vietnam, and the Guangxi and are important for 91 percent of the thorny political issue. But after 1975--as Yunnan provinces of China. Despite households interviewed in one of the growth in high-income countries slowed, marked disparities in economic main sending provinces. wage inequality increased, and the volume development among its members, Much of this migration, however, of international migrants swelled--immi- the subregion is extremely dynamic, will remain irregular and unregu- gration became a heated topic of debate in with annual growth rates averaging lated, increasing the vulnerability electoral politics. Indeed, selective "man- above 6 percent in recent years. of migrants, the majority of whom aged immigration" policies first introduced Thailand's higher wages, faster do not use social services because growth, and more favorable social they fear deportation. One of the in Australia and Canada in the 1980s are and political climate attract people biggest problems is ensuring access becoming popular in other high-income trying to escape poverty in Cambodia, to schooling for children, who also destination countries.18 Lao PDR, and Myanmar. For Thailand su er from a lack of health care. With the return to globalization since the migrants are a reservoir of cheap For the same reasons that migrant the end of the Cold War, the movement of and exible labor and a boost to its adults rarely receive health treat- labor across borders resumed, but govern- competitiveness in some sectors. And ment, migrant children rarely receive ments still restrict the number and influ- Thailand alone is estimated to have 1.5 vaccinations. million to 2 million regular and irregu- Despite the bene ts of labor mobil- ence the characteristics of immigrants. This lar migrants from the GMS. Removing ity, facilitating legal ows of people contrasts sharply with the "first era of glo- them could reduce Thailand's GDP by has been slow. Sending countries balization" in the nineteenth century, when around 0.5 percentage points a year. generally lack the capacity to man- the flows of labor were free of obstruction. By some estimates, more than half age the mass export of labor and to Restrictions on immigration arise and are of migrants enter Thailand holding protect the rights of their nationals sustained by wage inequality in receiving legal documentation and then over- abroad. Receiving countries have countries, rather than by unemployment stay, becoming illegal. Migrants are fairly weak migration frameworks, disproportionately young, of working often implemented hastily as an or absolute wages. They are more likely to age, and male. Those from Myanmar "after-the-fact" response to large be tightened when international labor flows are, on average, less educated and less numbers of migrants. The absence of increase and to be loosened in periods of literate than the average for the popu- an adequate legal and policy frame- domestic support for trade.19 lations of origin, indicating a push to work, typical of regional neighbor- migrate, or negative self-selection. hoods in developing country regions, Internal labor mobility--growing But self-selection is positive among increases the costs (and risks) of rapidly, despite restrictions migrants from Cambodia, who have migration and reduces its bene ts. slightly higher education attainment With improvements in transport tech- than the population back home. Source: World Bank 2006e. nology and infrastructure, the mobility of labor within countries rose steadily throughout the twentieth century, accel- erating in its last two decades. The vol- ume and velocity of internal voluntary migration, of concern to policy makers for in most of the developing world,21 except decades, are growing despite predictions to in Latin America, where movements the contrary.20 Declining agriculture and between cities dominate.22 Rural-to-rural rising manufacturing have changed the migration, difficult to document, has been distribution of labor in low-income and largely ignored.23 emerging middle-income countries since Migration from rural areas to cities has the mid-twentieth century in South Asia been gaining importance since the mid- and Sub-Saharan Africa, and long before 1970s, especially in the urbanizing econo- in East Asia and Latin America. Migration mies of South and East Asia, with the rapid of labor from lagging to leading rural areas rise of manufacturing and services. In remains the dominant internal movement India, where movements from poor to rich 154 WORLD DEVELOPMENT REPORT 2009 BOX 5.3 From facilitating to restricting to (again) facilitating labor mobility in China In the second half of the twentieth cen- millions back to the countryside. Then, · Rural-to-urban migration must be con- tury, China undertook some of the most with the Great Leap Forward (1958­60), trolled strictly. active internal migration policies ever the government abandoned all attempts · Movements between settlements of observed, initially to great economic to control the ow of labor, again seek- similar population size need not be bene t, but increasingly to the detriment ing to accelerate industrial development, controlled. of growth and development. Now, these motivating another surge of workers to policies are changing again. China's cities. By 1960 China's urban popu- · Flows from larger to smaller settle- In the 1950s the government sought lation had doubled from that in 1949. ments or between rural areas should be to stimulate industrialization through In the 1960s and rst half of the 1970s, encouraged. policies that encouraged rapid urbaniza- the urban population fell, a consequence tion. Households were given incentives of the relocation of intellectuals and Under the hukou, each individual has to move to cities, and rural workers urban elites to the countryside during the an o cial place of residence, and the responded en masse, answering the Cultural Revolution (1966­76). Then, with documents verifying residence are similar demand to participate in reconstruction an abrupt shift in policy in 1976, the ow to a passport. People are allowed to work and industrial development. As a result of people to the cities surged anew. legally, to receive social security bene ts of these e orts, the urban population of In the early 1980s the government including health coverage, and to access China had by 1953 grown by a third, to 78 became particularly preoccupied with the food rations only in their place of resi- million. The rst Five-Year Plan (1953­58) speed of urbanization. Although rural-to- dence. A change in o cial place of resi- promoted urban development, creating urban migration was responsible for only dence can be granted only by permission, forces that pulled people to the cities, 20 percent of the growth of China's cities similar to a local authority granting a visa. complemented by the collectivization of from 1949 to 1980, evidence of burdened But some forms of legal temporary migra- agriculture and the establishment of the infrastructure and services in Shanghai tion would be allowed to meet shifts in commune system. and Beijing occupied the attention of labor demand. In apparent response to a larger-than- policy makers. As the government's preoccupation expected ow from villages, the govern- The hukou household registration system with the size of China's cities and the ment tried to stem the ow, centralizing became the main policy tool to regulate the pace of urban growth changed, the hiring, restricting travel, and rationing ow of workers. It has four tenets: hukou was tightened or loosened--for grain in cities. But these measures failed to example, by relaxing the residency slow the out ow of Chinese rural workers, · Migration, especially to urban areas, requirement to receive food rations and the pressure on cities grew so much should be allowed only if compatible or extending the rights of temporary that the government mobilized to move with economic development. migrants. Despite the controls, lax China's industrial growth and concentration has been accompanied by massive movements of workers Internal labor migration between 1995­2000 Migration from Migration from Migration within the western region central region coastal region to to coastal region coastal region HEILONGJIANG JILIN WESTERNWESTERN REGIONREGION LIAONING LIAONING LIAONING MONGOL BEIJING BEIJING BEIJING NEI Beijing Beijing Beijing TIANJIN TIANJIN HEBEI HEBEI TIANJIN HEBEI COASTAL GXIA COASTALCOASTAL REGIONREGION SHANXI SHANDONG QINGHAI IN SHANDONG N SHANDONG Migration totals GANSU JIANGSU JIANGSU SHAANXI HENAN WESTERNWESTERN JIANGSU 2,000,000 REGION SHANGHAI ANHUI SHANGHAI SHANGHAI HUBEI 1,000,000 SICHUAN ZHEJIANG ZHEJIANG ZHEJIANG CHONGQING 500,000 COASTALCOASTAL JIANGXI COASTAL COASTAL HUNAN 250,000 REGIONREGION REGIONREGION 100,000 REGION GUIZHOU FUJIAN FUJIAN FUJIAN TAIWAN TAIWAN TAIWAN YUNNAN GUANGXI GUANGDONG GUANGDONG GUANGDONG HONG KONG HONG KONG HONG KONG MACAO MACAO MACAO HAINAN HAINAN HAINAN Source: Huang and Luo forthcoming, using data from the Population Census of China. Factor Mobility and Migration 155 BOX 5.3 From facilitating to restricting to (again) facilitating labor mobility in China--continued enforcement allowed large ows of to the pool of surplus labor. And China's information, and job search assistance, migrant workers to settle in cities under industries are in constant need of low- and recourse to legal action in cases of "temporary" status. Indeed, in the past cost labor. employer abuse. 30 years, the labor force requirements to Recognizing the growth dividend from Recent research suggests that the fuel China's spectacular growth perfor- allowing labor to ow freely, the govern- restrictions have taken a toll: many Chi- mance have relied on migrants who are ment has been loosening the hukou in nese cities are smaller than they should temporary under law but in fact perma- recent years, even facilitating migration. be.b In many areas, such as Chengdu and nent residents. Migration restrictions have declined. The Chongqing, governments are again facili- Today the movement of people from labor market has become more e cient. tating a rapid rural-urban migration (see rural areas to cities is again surging. And mobility decisions are much more chapter 7). One in ve rural workers migrates, and responsive to economic factors.a Begin- migrants account for a third of urban ning with pilot programs in selected employment. In 2005 average incomes in municipalities, migrants from rural areas Source: WDR 2009 team. cities were three times the rural average. will be given access to health and social a. Poncet 2006. The mechanization of farming has added protection services, training, labor market b. Au and Henderson 2006a. rural agricultural areas have historically Skills--the motor of internal and been the dominant form of internal migra- international migration tion, movements from villages to cities have A rush of labor matching the migrations out increased sharply in recent years. Migra- of Europe at the turn of the twentieth cen- tion from the poor Indian state of Bihar tury has not yet taken place, and perhaps it has doubled since the 1970s, mainly to cit- never will. But unlike the flow of unskilled ies, and not to the agriculturally prosper- labor, that of skilled labor--with human ous states in India's Northwest, as before. capital--has been on the rise. Globalization In Bangladesh two-thirds of all migration and selective migration policies are likely to from rural areas is to cities. And in China, ease travel for skilled labor within countries with the easing of residency restrictions, and across borders. migration from rural areas to cities now Within countries, education attainment predominates (see box 5.3). continues to determine who moves and Uniform measures of internal migration who does not--certainly from rural areas are rare. Because there are so few household to cities. People with more education are surveys that regularly measure labor mobil- more likely to migrate in their own country ity, and the questions asked about migra- (see figures 5.4 and 5.5).24 Many temporary, tion vary, comparable indicators can be seasonal migrants with little or no educa- calculated for only a few countries. Ques- tion also migrate.25 But education boosts tions about migration are more likely to the velocity of labor mobility, by opening be asked in countries that experience large employment opportunities farther afield movements of labor and with governments and shortening the job search at migrants' that are concerned about migration. Among destination.26 the countries included in table 5.3, for Education also increases the likelihood example, are stable nations (such as Argen- of people moving abroad. The interna- tina and Costa Rica), where migration is tional migration of skilled workers rela- more likely to reflect economic motives, as tive to that of unskilled workers has been well as countries that have recently experi- rising since the 1970s for every developing enced conflict (such as Bosnia and Herze- world region (see figure 5.6). The high- govina, the Democratic Republic of Congo, est proportions of skilled emigrants (as Rwanda, and Sierra Leone), where internal a percentage of the educated workforce) mobility is also due to flight from the threat are from Africa, the Caribbean, and Cen- of violence. tral America. Many Central American 156 WORLD DEVELOPMENT REPORT 2009 Table 5.3 Rates of labor mobility vary widely across countries in the developing world Internal Recent migrants Internal Recent migrants migrants (less than five Recent migrants (less than five Recent Country and year of (% of years, % of total migrants Country and year of (% of years, % of total migrants survey (ranked by working-age working-age (% of internal survey (ranked by working-age working-age (% of internal stock of migrants) population) population) migrants) stock of migrants) population) population) migrants) Bosnia and 52.5 12.8 24.5 Costa Rica 2001 19.9 2.5 12.4 Herzegovina 2001 Brazil 2001 19.5 3.3 16.7 Paraguay 2001 39.0 7.3 18.7 Sierra Leone 2003 19.0 3.7 19.3 Bolivia 2005 37.7 5.0 13.3 Nicaragua 2001 18.6 3.1 16.9 Morocco 1998 33.4 6.0 18.1 Guatemala 2006 17.5 3.3 19.1 Azerbaijan 1995 33.2 19.4 58.4 Haiti 2001 17.5 2.8 15.8 Honduras 2003 29.0 5.5 19.2 Argentina 2006 17.2 1.4 8.1 Venezuela, R. B. de 28.3 3.0 10.7 Kyrgyz Republic 1997 16.2 4.7 29.2 2004 Romania 1994 15.1 1.9 12.8 Congo, Dem. Rep. 27.1 7.8 28.9 of 2005 Croatia 2004 14.7 1.2 8.0 Dominican Republic 26.9 4.0 14.9 Bulgaria 2001 14.3 1.4 10.0 2004 Cambodia 2004 14.2 2.8 19.4 Armenia 1999 24.5 22.4 91.7 Tajikistan 2003 9.9 1.5 15.7 Mauritania 2000 24.2 2.9 12.0 Mongolia 2002 9.8 0.0 0.4 Albania 2005 23.9 4.1 17.3 Kazakhstan 1996 9.3 1.4 14.7 Ecuador 2004 22.7 5.3 23.4 Madagascar 2001 9.3 0.0 0.0 Vietnam 1992 21.9 3.1 14.3 Mozambique 1996 8.1 0.0 0.2 Rwanda 1997 21.5 5.9 27.6 Malawi 2005 2.7 1.1 43.2 Colombia 1995 20.1 5.3 26.3 Micronesia 2000 1.2 0.3 23.6 Source: WDR 2009 team, estimates using household surveys. Note: Internal migrants are individuals who are not living in the same district in which they were born. This definition does not count returnees as migrants--that is, persons who moved away from their place of birth in the past, but returned by the time of the survey. Recent migrants migrated in the five years before the year of the survey. Figure 5.4 In Latin America and the Caribbean internal migrants are more educated than those who stay behind Education comparison between internal migrants and nonmigrants at place of origin and at time of migration Percentage 40 35 30 25 20 15 10 5 0 Costa Rica Honduras Venezuela, Paraguay Guatemala Nicaragua Dominican Brazil Haiti R.B. de Republic Education level attained Migrants, primary Migrants, secondary Migrants, post-secondary Nonmigrants, primary Nonmigrants, secondary Nonmigrants, post-secondary Source: WDR 2009 team drawing from selected household surveys. Factor Mobility and Migration 157 and island nations in the Caribbean had Figure 5.5 Internal migrants are more schooled than workers in the places to which they move more than half their university-educated Comparison between internal migrant and nonmigrant workers at migrant's place of destination citizens living abroad in 2000. And close to 20 percent of skilled workers have left Tajikistan Sub-Saharan Africa.27 This could be taken as evidence that human capital is becom- Colombia ing more mobile internationally--or that Venezuela, R.B. de "selective" immigration policies in wealthy countries are biasing the composition of Albania international migration toward those with skills. But the increase in the migration of Congo, Rep. of Nonmigrant, primary completed skilled labor is due to the rise in higher Migrant, primary completed levels of education worldwide, most nota- Croatia Nonmigrant, secondary completed bly for countries sending the majority of Migrant, secondary completed international migrants. In relative terms, Mongolia the cross-border movement of skilled labor Malawi has remained fairly constant as a share of the stock of skilled labor in sending coun- Brazil tries. Rather than human capital becom- ing more mobile, more human capital is Vietnam simply available, propelling larger volumes of migration.28 Argentina The rapid development of telecommu- Micronesia nications and other forms of information and communication technology has sepa- Haiti rated the mobility of human capital from the mobility of labor. In a trend likely Bolivia to accelerate, more services in the pro- duction processes of industries based in Dominican Republic wealthy countries are being located "off- shore" in low- and middle-income coun- Kyrgyz Republic tries, where human capital is cheaper. Paraguay What began with the export of software development and maintenance services Bulgaria from the Indian city of Bangalore to firms across the world has developed into Madagascar a burgeoning trade in services requiring a wide array of skills, from simple cus- Costa Rica tomer communications--particularly from countries like India and the Philip- Guatemala pines, where English is widely spoken--to Honduras financial accounting and computer main- tenance. There is, as yet, no evidence that Nicaragua the export of "disembodied" human capital over telephone lines and the Inter- Cambodia net will substitute for the flow of skilled workers. But by creating the possibility Mozambique of separating human capital from labor, 0 10 20 30 40 50 60 70 80 90 100 information and communication tech- Percentage nology has further increased the mobility Source: WDR 2009 team drawing from selected household surveys. of skills relative to people. 158 WORLD DEVELOPMENT REPORT 2009 Figure 5.6 Migration from developing countries is becoming more skilled Labor mobility and voluntary migration for economic gain are the human side of Latin America & Caribbean Sub-Saharan Africa Percentage of emigrants Percentage of emigrants the agglomeration story. 100 100 As for firms, localization and urbaniza- tion economies arise from knowledge spill- 80 80 overs between people in proximity. For this 60 reason, people are often more economically Low skill 60 productive when they are around others, 40 40 Medium skill especially for people with skills. Migra- tion and agglomeration feed off each other. 20 20 Highly skilled Whether this is agglomeration in leading 0 0 places within a geographic area, in leading 1975 1980 1985 1990 1995 2000 1975 1980 1985 1990 1995 2000 areas within countries, or in leading coun- Year Year tries within regions, human capital flows to Europe & Central Asia Middle East & North Africa economically leading places. At every spa- Percentage of emigrants Percentage of emigrants tial scale, migration is the way that people 100 100 who invest in education and skills realize 80 80 the returns on their investment. An important insight of the agglomera- 60 60 tion literature--that human capital earns higher returns where it is plentiful--has 40 40 been ignored by the literature on labor 20 20 migration. Preoccupied with urban unem- ployment and squalor in the fast-growing 0 0 cities of the South, early research on labor 1975 1980 1985 1990 1995 2000 1975 1980 1985 1990 1995 2000 Year Year migration advocated restrictions.29 Govern- ments often acted on these prescriptions, East Asia & Pacific South Asia Percentage of emigrants Percentage of emigrants instituting migration abatement policies, 100 100 but to little effect: flows from the country- side to cities and from lagging to leading 80 80 provinces continued unabated. The only effect of restrictions may have been forgone 60 60 economic growth and slower spatial con- 40 40 vergence in living standards. Increasing returns to scale and spillovers 20 20 from clustering--especially human capital- related spillovers--make clear the growth 0 0 1975 1980 1985 1990 1995 2000 1975 1980 1985 1990 1995 2000 and welfare payoffs of policies that facili- Year Year tate movements of labor from lagging to Source: Docquier 2006. leading places. The implications for policy are powerful. Rather than an impending destructive tide of humanity, the swelling Labor mobility: learning from a flows of people from villages to cities could generation of analysis be a boon for economic growth and the Until recently, two strands of economic convergence of welfare. Moving from the literature--labor migration and economic local spatial scale, to the national, and then growth, each with a powerful impact on to the international, the benefits from clus- developing-country policy makers-- tering are the same--and the problems fac- have evolved along separate tracks and ing policy makers grow and become more diverged. But as shown in the arguments complex. Put plainly, they do not want to and evidence on agglomeration econo- lose people and human capital. mies in chapter 4, the two phenomena Economic theory now recognizes that are closely intertwined in the real world. governments should not try to hold on to Factor Mobility and Migration 159 people. The pull of agglomeration forces in surplus. But in the growing modern in prosperous places is simply too strong manufacturing sector it is not. So, in rural for any opposing measure to be sustained. areas, every additional worker, irrespective Another aspect of the pull of agglomeration of innate talents or education attainment, has been well studied by economists but is has zero marginal productivity, but each often overlooked by governments. Migrants potentially has a positive marginal produc- who move to cities, to leading areas, or to tivity in manufacturing. This opens a gap leading countries are rarely disconnected in earnings and an incentive for labor to from their home places. Most migrants migrate from rural areas to cities in search maintain strong and active links with their of manufacturing jobs. Agriculture supplies home communities and send remittances. an unlimited labor force to manufacturing, And they do much more than remit capital. and the transfer of labor between the two They send back information and technical sectors takes place through rural-urban assistance, and when a place is ready, they migration. This migration continues until often bring back ideas, knowledge, expecta- the "disguised unemployment" of workers tions of good governance, and links to lead- in rural areas is absorbed into manufactur- ing markets. Sending governments that put ing in urban areas. the right policies in place can capture these Movements from rural to urban areas benefits for faster growth and faster conver- were considered desirable when accom- gence in living standards. panied by economic growth. In what has become known as the Todaro class of migra- Migration theory now recognizes the tion models,33 prospective migrants decide benefits of agglomeration whether to move by comparing the expected Economists' notions about what motivates future income streams they could earn in people to move and what such movements the city and in the rural home, after taking mean come from theories of economic into account the costs of actually moving growth and convergence. Whether couched and searching for a job.34 A key feature of in a classical framework30 or in the recent the early Todaro models is that the econo- models of "endogenous growth,"31 where my's rate of growth--and by extension the people are free to move, they will move to rate of employment creation in the mod- compete away differences in wages between ern manufacturing sector--was assumed locations. Since higher wages at the desti- to be constant and set independent of the nation reflect an initial shortage of workers model. This classical framework--with an relative to capital--or a large endowment exogenously determined rate of economic of capital per worker--the arrival of new growth and constant rate of employment migrants will slow the accumulation of creation in the manufacturing sector-- capital per worker and the growth of wages. explained rising urban unemployment in In contrast, the accumulation of capital cities like Nairobi. But it also created what per worker in the places migrants leave came to be known as the Todaro Paradox: will speed up as they go, accelerating wage any policy to improve urban economies growth for workers who stay behind. By this could lead to more urban unemployment mechanism, incomes in different locations because the improvements would induce are predicted to eventually converge. even more migration from rural areas. The first theories of labor migration Few economic models have had as much originated in the analysis of economic impact on policy makers in developing growth in developing countries.32 These countries as these early labor migration early theories partitioned a developing theories. Across the developing world, but economy into a traditional agrarian rural especially in South Asia and Sub-Saharan sector and a modern manufacturing sec- Africa, the Todaro Paradox provided a basis tor centered in urban areas. The main idea for strong disincentives and even outright was that with economic development, par- restrictions on the movement of labor.35 ticularly with progressive mechanization The Todaro model suggested that prohibit- of agriculture, labor in rural areas is always ing internal migration over and above what 160 WORLD DEVELOPMENT REPORT 2009 labor movements. It minimized differences in the appetite for risk among prospective BOX 5.4 Labor and social policies restrain migration in migrants. It did not account for differences Eastern Europe--not good for growth in education attainment and how these Internal migration increased in unemployed workers close to the differences can influence job searches. It several countries in Eastern Europe retirement age receive preretirement ignored pertinent motivations and house- and Central Asia at the beginning bene ts linked to their pensions. Low- hold characteristics that could influence a of the transition away from planned income households are also eligible family's choice of who will migrate. And it economies. But this may have been for guaranteed temporary social neglected the possibility of migration for a one-time phenomenon. Much of assistance bene ts. Housing policies jobs in the urban informal economy and the sudden increase in migration in may discourage migration. During the Commonwealth of Independent the transition, homes typically were the pull these could exert independently of States appears to have been driven transferred to their occupants at little the modern manufacturing sector.36 by the return of people to their eth- or no cost. So the cost of remaining In parallel, some economists in the mid- nic homelands and the departure of in one's home is low. At the same 1980s began to think differently about eco- workers from areas they had been time, rent control discourages new nomic growth, mainly by reformulating the sent to by central planners. More construction, driving up the cost of way classical growth models treated techni- than 1 million people relocated from housing in regions that are expanding cal progress. Human capital and ideas were Siberia and the Russian North and Far economically. The high cost of hous- East to the more central parts of the ing in economically prosperous places different from other factors of production-- Russian Federation, about 12 percent can whittle away at the income gains they exhibited increasing returns to scale.37 of the populations of these areas. workers might expect from migration. And because the generation of ideas and These movements may have run Also discouraging migration are human capital are in essence social activ- their course. Migration has slowed uniform national minimum wages ities--clustering people in a way that has despite di erences in income and unadjusted for costs of living, col- no comparison in the process of accumu- the quality of life. Internal migrants in lective bargaining arrangements, lating physical capital--these models could the Czech Republic, Poland, and the and job protection laws. In other Slovak Republic represent less than regions where the informal economy explain why cities are important. They also 0.5 percent of the working popula- is dominant, labor market regulation could explain why human and financial tion, much less than 1.5 percent in is less binding. But in the formerly capital do not move from where they are Germany, and nearly 2.5 percent in planned economies of Central and already abundant--rich countries, leading France, the Netherlands, and the Eastern Europe, minimum wage and areas in countries, and cities--to where they United Kingdom. job protection regulations matter. In seems to be scarce--poor countries, lagging People of working age in Europe's Poland, where the minimum wage areas, and rural communities.38 economically depressed areas mainly is relatively high, national wage- don't move because extensive setting appears to inhibit the migra- If there are external effects from cluster- unemployment bene ts and social tion of workers from economically ing human capital, cities can jump-start assistance reduce the pressure to depressed areas. Elsewhere in the and maintain economic growth. Although migrate from declining areas. Under region, where legislated minimum urban specialists had long held this view,39 Poland's unemployment insurance, wages are relatively low, they do not it was sufficiently novel for economists. quali ed workers receive fairly gen- appear to have a similar e ect on Researchers in urban economics enthusias- erous bene ts for periods ranging internal labor mobility. tically took up the hunt for the theorized from six months in areas with low unemployment to 18 months in areas positive external effects from human capi- Sources: Dillinger 2007, Paci and others with high unemployment. In addition, 2007. tal spillovers. Theoretical and empirical studies sought to quantify what happens to productivity, wages, and land prices when the aggregate stock of human capital in a is required for full employment in manu- city increases.40 Evidence began to emerge facturing could increase national welfare of social returns to education accruing to because output in both agriculture and specific geographic areas, supporting argu- manufacturing can be maintained at opti- ments in favor of a greater concentration of mal levels. economic activity, if not the clustering of In the late 1970s and early 1980s, how- labor specifically.41 ever, economists began to question the These arguments did not themselves classical models, pointing out the weak- spill over into the mainstream labor migra- nesses of the Todaro framework, which tion literature until the turn of the century. failed to capture the dynamic nature of This should come as little surprise: the Factor Mobility and Migration 161 fundamental assumptions for the classical first time.43 And these are only the flows of migration models are at odds with those remittances that governments and research- embraced by the new growth theorists and ers can observe--just a fraction of what is by those emphasizing agglomeration econ- actually sent through formal and informal omies (see box 5.5). channels. Allowing the freer flow of skilled and unskilled labor across national borders Migration, growth, and welfare: would probably do more to reduce poverty divergence or convergence? in developing countries than any other sin- In a world with increasing returns to scale, gle policy or aid initiative.44 will selective, voluntary migration lead to economic divergence or convergence? A large volume of empirical work from devel- oped and developing countries bolsters BOX 5.5 From Lewis to Lucas: the economic perspective on an emerging consensus that governments migration has changed should not see voluntary internal popula- tion movements as a threat. Indeed, internal The insights from economists that have retical study of migration from rural to migration offers societies an opportunity had the greatest impact on how policy urban areas in low- and middle-income for economic growth and the convergence makers view migration share similar ori- countries.a He posited a transfer of of welfare. gins, in theories of economic growth. labor from a traditional sector, employ- The evolution of economic thought ing a land-intensive technology, to a In contrast to the emerging consensus on on migration--and particularly on the modern human capital-intensive sec- migration within countries, the benefits and growth payo from clustering labor tor, with an unending potential for eco- costs of international migration are still the and talent in cities--spans the work nomic growth. In Lucas's model, cities subject of debate. The preeminence of place of two Nobel Laureate economists, W. are places where new immigrants can in determining the return on an individ- Arthur Lewis and Robert E. Lucas, Jr. accumulate skills required by modern ual's investment in human capital is most Lewis laid the foundations for production technologies. In the con- dramatically observed in the difference the the study of labor migration with clusion to the paper, referring to the his two-sector model of economic attraction to cities driven by gains from simple act of crossing a border can make growth in developing countries. But agglomeration, he writes: to earned income. An adult male Bolivian theorists studying economic growth "Even in the rapidly growing economies with nine years of schooling in Bolivia will since Lewis took a di erent path from of the post-colonial world, the passage earn roughly US$460 per months in dol- those who used his insights to focus from a 90 percent agricultural economy to lars that reflect purchasing power at U.S. narrowly on labor migration. one that is 90 percent urban is a matter of decades. Since everyone has the option to prices. But a person with the same educa- The classical migration models migrate earlier rather than later, something tion, talent, and drive would earn about inspired by Lewis assumed an exog- must occur as time passes that makes the enously determined and constant city a better and better destination."b 2.7 times that much if he worked in the rate of economic growth. In sharp The new insight from theories that United States. A similar Nigerian educated contrast, the new growth theorists-- acknowledged spillovers from cluster- in Nigeria would earn eight times as much inspired by Lucas's contention that ing human capital is that, while the by working in the United States rather than there are positive external spillovers returns to scale in agriculture are con- in his native country. This "place premium" from clustering human capital-- stant, the returns to scale in manufac- is large throughout the developing world.42 internalized growth in models that turing and services are increasing. The Although the benefits to an individual allowed for increasing returns to policy implications of adopting one scale. The classical theories modeled from migrating from a poor country to a view or the other are profoundly di er- each additional migrant as lower- ent. A policy maker persuaded by the wealthy country are clear, is the acceler- ing the probability of employment, classical view would restrict the move- ated flow of skilled labor out of developing contributing to urban unemploy- ment of labor, particularly ows of countries more likely to help or to hinder ment, and raising congestion costs. migrants from villages to towns and cit- their growth and convergence prospects? The new growth theorists and later ies. In contrast, a policy maker who rec- The answer is disputed. the proponents of urban agglom- ognizes the external bene ts of human But what is not disputed is the growing eration economies could imagine in capital would do exactly the opposite, that migrant an additional source of volume of internationally remitted earn- facilitating migration and clustering, human capital to drive the agglom- particularly of workers with skills. ings, which now outpace all other capital eration engine of growth. flows to poor and middle-income countries. In 2002 Lucas bridged the gaps Source: WDR 2009 team. In 2007 the flows of remittances to many between these diverging strands of a. Lucas Jr. 2004. developing countries surpassed those of the development literature, in a theo- b. Lucas Jr. 2004. foreign direct investment and equity for the 162 WORLD DEVELOPMENT REPORT 2009 Labor migration promotes growth. Research in Bangladesh, China, the Phil- Within countries, the accumulated empiri- ippines, and Vietnam suggests that internal cal evidence shows that labor migration migration has helped to drive growth.51 In increases the earnings prospects of people Brazil internal migration has raised pro- who move. It also shows that labor migra- ductivity by allowing producers to reap the tion contributes to aggregate growth by benefits of agglomeration.52 Conversely, improving the distribution of labor, driv- in China, restrictions on the movement of ing concentration. And by clustering skills labor impede growth by constraining city and talent, migration drives agglomeration size. Because Chinese cities are kept artifi- spillovers. In the United Kingdom the esti- cially smaller than they might otherwise be, mated long-run wage premium for men the country has experienced welfare losses who migrate is about 14 percent, and for from forgone higher growth rates.53 And in women about 11 percent.45 Wage premi- India past restrictions on the movement of ums ranging from 7 percent to 11 percent labor may have kept the size of Indian cit- have been found among internal migrants ies inefficiently small, at a cost in forgone in the United States.46 These gains for indi- growth (see box 5.6).54 Internal migrants vidual migrants translate into gains for the are clearly economically active. In 24 of the broader economy. In many countries high 35 countries with comparable survey indi- rates of internal labor mobility have been cators, migrants are as or more likely than associated with periods of sustained eco- locally native people of working age to be nomic growth, as in the United States from employed (see figure 5.8). 1900 to 2000,47 Brazil from 1950 to 1975,48 In today's developed countries, lead- Japan from 1950 to 1975,49 the Republic of ing and lagging areas converged. As for Korea from 1970 to 1995,50 and China from earnings and living standards between the 1980 to 2005. Among a selection of develop- leading and lagging areas, historical evi- ing countries with comparable measures of dence of the impact of the internal labor internal migration drawn from household movement during the nineteenth century surveys, a positive association is found in today's developed countries supports between internal labor mobility and eco- convergence. nomic growth (see figure 5.7). At the start of the nineteenth century, the majority of nonindigenous people in the United States lived on the eastern seaboard. Figure 5.7 Internal labor mobility and economic growth often go together By century's end, more than 2 million square miles had been added to the country's origi- Migrants as % of population 30 nal land area (see "Geography in Motion 1: Overcoming Distance in North America"). Robust institutions were critical in settling 25 such a large land mass. The U.S. Constitu- tion along with the Northwest Ordinance 20 (1787) provided the framework for trans- forming unsettled areas into states. Factor 15 mobility was enhanced by the commerce clause in the Constitution, explicitly pro- hibiting state governments from restraining 10 trade across state boundaries. State and local governments provided public goods and 5 infrastructure to attract settlers. In the 1820s real wages of "common" (unskilled) non- farm labor were about 33 percent higher in 0 ­15 ­10 ­5 0 5 10 the Midwest than in the Northeast. Between Growth of GDP per capita in year before survey 1820 and 1860, the Midwest's share of the Source: WDR 2009 team estimates using selected household surveys. unskilled northern labor force rose from Note: Marker shows land area of country. Marker in the upper-left-hand corner is República Bolivariana de Venezuela in 2004. 23 to 45 percent. As the Midwest's share of Factor Mobility and Migration 163 BOX 5.6 Implicit barriers to mobility: place-based entitlement and divisions in India Policy barriers to internal mobility in India tion as internal migrants--people living movement are recorded within the same are imposed by omission rather than by and working in a place other than where district. The ow of migrants across state commission, exemplifying the implicit they were born. The share of migrants is lines is a trickle. Since 2001 there has obstacles to migration in many develop- larger in cities (about 40 percent of the been a slowdown in permanent or long- ing countries. Current policies do not population) than in rural areas (about 30 term migration (see the map below). The allow communities to fully capture the percent). But by far the largest ows of share of lifetime (permanent) interstate bene ts of labor mobility. The costs and migrants--within districts, across districts, migrants--at about 4 percent--is much risks of migration would be signi cantly and across states--are from lagging rural lower than the total migrant population. lowered by greater exibility in the way to leading rural areas. Since the 1960s Most of these permanent migrants live in households use public services and rural-to-rural migration ows typically cities. In addition to geographic distance, social entitlements, and in the deploy- have been more than twice the volume as the strong di erences in culture and lan- ment of targeted assistance for mobile the next largest ows, from rural areas to guage can discourage movement far from populations. Negative attitudes held by cities. Rural-rural migration accounted for a person's home place. government and ignorance of the ben- roughly 62 percent of all movements in Although o cial data sets indicate e ts of population mobility have caused 1999­2000. Workers from lagging states a slowdown in permanent rural-urban migration to be overlooked as a force in like Bihar, Orissa, Rajasthan, and Uttar migration, microstudies nd that circular economic development. Pradesh routinely travel to the developed migration is emerging as a dominant Recent evidence shows that population green revolution states of Gujarat, Maha- form of migration among the poor. mobility in India--having stabilized in the rashtra, and Punjab to work on farms. Short-term migrants have been esti- 1970s and 1980s--is rising. India's 1961 In India both distance and division mated to number 12.6 million but recent census classi ed 33 percent of the popula- limit labor mobility. The highest levels of microstudies suggest that the gure is 30 million and rising. The economic bene ts of migration Internal migrants in India flow to prospering Delhi and Maharashtra are not always recognized by policy Internal migrant flows reported in 2001 census makers. Two forms of policy have been MAHARASHTRA HIMACHALHIMACHAL DELHI HIMACHALHIMACHAL PRADESHPRADESH PRADESHPRADESH attempted to counter migration in India. PUNJABPUNJAB UTTARANCHALUTTARANCHAL PUNJABPUNJAB UTTARANCHALUTTARANCHAL UTTARANCHAL The rst response has been to increase HAHARYANRYANADELHIDELHI HARYANA HAHARYANRYANA HARYANA SIKKIMSIKKIM SIKKIMSIKKIM rural employment, in an attempt to stem 131,895 DELHIDELHI movement out of rural areas. This policy RAJASTHAN RAJASTHAN UTUTTARTAR UTTAR RAJASTHAN RAJASTHAN UTUTTARTAR UTTAR PRADESHPRADESH PRADESHPRADESH implicitly assumes that deteriorating BIHARBIHAR BIHARBIHAR agriculture leads to out-migration and JHARKHANDJHARKHAND JHARKHANDJHARKHAND GUJARAT GUJARGUJARAT WESTWEST MADHMADHYA PRADESH MADHYA PRADESH YA PRADES MADHYA PRADESH MADHMADHYA PRADES YA PRADESH WESTWEST BENGALBENGAL BENGAL BENGALBENGAL BENGAL that improved employment opportuni- GUJARAT GUJARGUJARAT ties in lagging rural areas can reduce CHHHHAATTTISGARH TISGARH CHHATTISGARH TTISGARH ORISSAORISSA ORISSAORISSA or reverse migration. These measures CHHA C MAHARASHTRA MAHARASHTRAMAHARASHTRA 12,505,916 include the recently introduced National Rural Employment Guarantee Program, ANDHRA ANDHRA ANDHRA ANDHRA PRADESHPRADESH PRADESH PRADESH which promises 100 days of wage labor GOA KARNATAKA KARNKARNATAKATAKA GOA KARNATAKA KARNKARNATAKATAKA to one adult member in every rural PUNJAB HIMACHALHIMACHAL UTTAR PRADESH HIMACHALHIMACHAL household who volunteers for unskilled PRADESHPRADESH PRADESHPRADESH PUNJABPUNJAB 1,712,627 PUNJAB PUNJAB UTTARANCHALUTTARANCHAL UTTARANCHAL UTTARANCHAL work, numerous watershed development HAHARYANRYANA HARYANA SIKKIMSIKKIM HARYANA HAHARYANRYANA SIKKIMSIKKIM programs that aim to improve agricultural productivity, and programs to develop DELHIDELHI DELHI DELHI UTUTTARTAR UTTAR RAJASTHAN RAJASTHAN UTUTTARTAR UTTAR RAJASTHAN RAJASTHAN PRADESH PRADESH PRADESHPRADESH small and medium towns. MEGHALAYA MEGHALAYA 8,969,367 BIHARBIHAR BIHARBIHAR The second policy response is implicit. JHARKHANDJHARKHAND JHARKHANDJHARKHAND Because of the perceived negative MADHMADHYA PRADESH MADHYA PRADESH YA PRADES WESTWEST WESTWEST BENGALBENGAL BENGAL MADHYA PRADESH MADHMADHYA PRADESH YA PRADES BENGALBENGAL BENGAL GUJARGUJARAT GUJARAT GUJARAT GUJARGUJARAT e ects, local governments remain hos- CHHATTIS ISGARRHH CHHATTIS TTISGGAARRHHORISSAORISSA ORISSA ORISSA tile toward migrants, while employers CHHA CHHA MAHARASHTRAMAHARASHTRA MAHARASHTRA MAHARASHTRA routinely disregard laws to protect their rights and needs. In many cases welfare ANDHRA ANDHRA PRADESH PRADESH ANDHRA ANDHRA policies and social services are designed PRADESH PRADESH for a sedentary population. This is best GOA KARNA AKA KARNATAKA GOA KARNA AKA KARNATAKA exempli ed by location-speci c entitle- 12,505,916 Number of within-state migrants Migration totals ments to social services, housing subsi- 1,000,000 100,000 dies, food rations, and other public ame- 50,000 500,000 nities especially important to working Source: WDR 2009 team, based on census data from the Census of India. poor people. 164 WORLD DEVELOPMENT REPORT 2009 Figure 5.8 Internal migrants are more likely to work than natives large in the mid-nineteenth century, eroded Comparison of internal migrants and native workers as labor poured into the city. More gener- % employed ally, wage differences across French dépar- 0 10 20 30 40 50 60 70 80 90 100 téments narrowed significantly over the Malawi second half of the nineteenth century. Bosnia and Herzegovina Across Canadian cities, wage differences Azerbaijan evolved precisely as would be expected if Mauritania Canadian migrants treated the entire coun- Armenia try as a single labor market.56 And across Morocco England wage differences did not erode Tajikistan much during industrialization. This was Haiti not because the labor force was immobile, Micronesia as history shows English labor moved freely Bulgaria from low- to high-wage areas. Rather, it was Kazakstan because the forces that sustained industrial Croatia development in various parts of England Mongolia were so persistent that internal migration Albania in any one period was not enough to cause Venezuela, R.B. de wage gaps to decline.57 Guatemala Throughout the twentieth century, labor Dominican Republic mobility continued to be integral to con- Honduras Nicaragua centration and convergence across the two Sierra Leone geographically expansive and economically Romania prosperous countries of North America. Colombia In the United States labor mobility was Argentina greater than in Europe, distributing labor Ecuador from low- to high-wage states, converging Brazil state per capita incomes through most of Bolivia the twentieth century.58 With the notable Paraguay exception of the U.S. South (see box 5.7), Costa Rica regional convergence in wage rates coin- Mozambique cided with cross-regional labor market Madagascar institutions and information flows.59 Simi- Congo, Dem. Rep. of larly, in Canada, labor mobility continued Vietnam to narrow per capita income differences Kyrgyz Republic among Canadian provinces from 1910 to Cambodia 1921. And when internal migration petered Rwanda to a trickle between 1921 and 1960, conver- 0 5 10 15 20 25 30 35 40 gence in income also slowed.60 % unemployed Disparities in income and welfare Employed natives (top axis) Unemployed natives (bottom axis) between places were higher in the 15 coun- Employed migrants (top axis) Unemployed migrants (bottom axis) tries of Western Europe where labor mobil- Source: WDR 2009 team using household surveys. ity has been much lower than in North America throughout the twentieth cen- the labor force increased, its wage advantage tury. The variation in employment across eroded to roughly 17 percent in the 1850s, subnational areas of the United States is and to 10 percent in the four decades after much lower than that across subnational the Civil War. State data on farm wages point economic areas of the European Union to a long-term narrowing of geographic wage (EU). In parts of the EU where employ- differences in response to internal migra- ment is highest, the employment rate is 60 tion, a process that can be dated back even percent higher than where it is lowest. In before the Civil War.55 the United States the difference is only 22 In France wage differences between the percent.61 The greater equity in employ- Paris metropolitan area and the countryside, ment outcomes in the United States is a Factor Mobility and Migration 165 direct consequence of higher labor mobil- ity and a more tightly integrated national BOX 5.7 Why did the U.S. South take so long to catch up? labor market.62 Division. Convergence, after divergence, in developing countries. More recent empir- Researchers have long questioned ers found opportunities by moving ical studies show the positive impact of why wages and incomes in the U.S. westward. net migration on income convergence in South were so di erent from the rest Even when the ows of migrants Japan.63 So do estimates from India, the of the country for so long. Although from the South to the North began to wages and incomes in the South grow, wage and income di erences Russian Federation, and the United King- dom.64 These studies may underestimate have caught up, particularly since the persisted. Scholars explain that the Great Depression, the process seems Great Migration of rural southern the full impact of migration on the con- to have been remarkably slow before blacks to northern cities involved a vergence in living standards by failing to the 1940s (see "Geography in Motion disproportionately educated seg- take into account the differing skills of 1: Overcoming Distance in North ment of the population. After World migrants flowing in and out of areas, and America"). War II, the selective migration of by ignoring the indirect negative impact Before the New Deal, the southern African Americans moderated, with labor market was isolated from the return migration associated with eco- of housing and labor market rigidities. rest of the country, with large wage nomic growth in the South. Analysis that distinguishes between (1) gaps. Yet, there was little migration With the gradual buildup of infor- the flow of skilled and unskilled migrants out of the southern states, even mation and migration corridors and (2) the levels of human capital in among African Americans who suf- between North and South, the elimi- migrants' places of origin and destination fered the lowest wages and the most nation of legal racial discrimination in shows a large impact of labor mobility on social discrimination and political the wake of the civil rights movement, convergence. disenfranchisement. Before the 1920s and the improvement in the educa- demand for low-skilled labor in the tion of African Americans students Much of the empirical evidence of the industrializing North was satis ed with the racial integration of schools, impact of internal migration from low- by migrating workers from Europe. the wages and incomes in southern and middle-income countries is consis- Employers in northern factories states gradually converged with those tent with expected convergence, after an showed a preference for low-skilled in northern states. Indeed, since the initial divergence. As Japan grew, regional European immigrants over blacks, 1970s, labor migration between the income inequality followed a bell-shaped and the abundance of European North and the South and other areas curve, initially increasing in 1955­61, but migrants made it a ordable to of the United States has been a safety indulge this preference. This would valve easing economic pressure dur- then falling in 1961­75. Research attributes the convergence to labor migration.65 In change with the restrictions on ing recessions. The di ering impact movement across borders, tightened of economic downturns across areas China after the economic reforms of 1978, in the 1930s. of the country have spurred large income differences between subprovin- Until World War II, there were few movements of workers from states cial areas initially widened after economic established ows of either informa- where the economy was contracting reforms from 1978­96, but later declined tion or labor between the South to other, more prosperous areas. In with greater integration and internal labor and the North. Given the cumulative the early 1990s, a sizable number of migration.66 In India and Indonesia higher dependence of migration corridors, workers migrated from the Northeast this impeded the movement of to states in the South. internal labor mobility is associated with low-skilled workers of all races out lower income inequality,67 and migra- Sources: Rosenbloom and Sundstrom of southern states. Southern work- 2003; Margo 2004; and Vigdor 2006. tion increased equality in Mexico. But in Chile, lower-than-expected rates of inter- nal migration may be to blame for high Saharan Africa, remittances account for income inequality.68 And in Brazil, where 15 percent of rural income. In Uganda both social inequality and spatial mobility labor mobility has a positive effect on of labor are high, some research suggests household expenditure. Comparable that inequality would be even higher if not households enjoy much higher per capita for internal migration.69 spending if they migrate within their own Most early research on internal migra- district or to another district than if they tion and convergence across areas within stay in their native area. The incomes of countries focused only on a small part Ugandans who migrate from lagging to of migration as an economic force: labor leading areas are 10 to 60 percent higher market adjustments and changes in wage than nonmigrants in origin or destination differences between areas. The impact areas.70 In Tanzania, in the Kagera region of remittances was ignored. Yet in Sub- between 1991 and 2004, internal migration 166 WORLD DEVELOPMENT REPORT 2009 added 36 percentage points to consump- and increase the investment in education.73 tion growth.71 And in East Asia remittances from migrant Remittancesfrominternalmigrationhave family members increase investment in a positive impact in other developing regions education and capital-intensive household as well. In Bangladesh temporary migrants to enterprises in the Philippines.74 In China Dhaka send up to 60 percent of their income the ministry of agriculture expects that the to family members in their home places, remittances of migrant workers will soon covering a large share of the household be more than earnings from agriculture for budgets of migrant-sending households. In rural households. several Latin American countries remitted For the communities left behind, inter- earnings not only augment the consumption nal migration is critical for overcom- of receiving households, but also lower the ing poverty and smoothing household incidence of poverty in their communities72 consumption in the wake of unexpected shocks. Indeed, World Development Report 2008 Agriculture for Development identifies Table 5.4 Most migrants move for economic reasons, but many are pushed out by internal migration as an important "path- poor services way out of poverty" for rural households Percentage of internal migrants reporting reason for migration that can no longer rely solely on agriculture To seek For education, for their livelihood.75 Even in situations in employment health, or better For sociopolitical which supporting the permanent resettle- or join family living conditions or other reasons ment from villages to cities may be fairly Malawi 95 1 4 costly, within the means of only better- Morocco 91 2 7 off households, seasonal and temporary Romania 87 10 3 migration can more immediately mitigate Ecuador 86 12 2 downward shocks to consumption in rural areas than even the best-designed social Nicaragua 84 5 11 assistance program.76 Albania 82 11 7 Members of rural households in Ban- Mozambique 81 4 15 gladesh migrate to cities to diversify house- Vietnam 80 7 13 hold income when harvests are lower than Armenia 78 1 21 expected. Internal migration in China raises the consumption of households in Tajikistan 78 10 12 migrants' home communities, and the Kazakstan 77 14 9 increase is greater for poorer households. Bolivia 77 17 6 And the out-migration of Chinese work- Dominican Republic 76 21 3 ers allows those who remain in rural areas Cambodia 75 2 24 to work more. The gains associated with Paraguay 74 24 2 internal migration increase housing wealth and consumer durables as well as agricul- Guatemala 72 24 4 tural production.77 Indeed, the selective Bulgaria 71 28 0 phenomenon that determines who moves Congo, Dem. Rep. of 69 6 25 seems to work both ways. Those with higher Kyrgyz Rep 69 28 3 academic achievement choose to migrate to Madagascar 62 28 11 jobs in China's cities, and people who are better at farming choose to stay.78 Bosnia & Herzegovina 55 1 43 Pulled or pushed? The development Rwanda 54 5 41 benefits of migration are seen when peo- Azerbaijan 44 5 52 ple move voluntarily. Large numbers Mongolia 41 28 30 of people--particularly in the poorest Sierra Leone 23 3 74 countries--are also forced to move by deteriorating living conditions and con- Mauritania 23 74 4 flict. People are "pushed" off their land Source: WDR 2009 team estimates using household survey data. when agriculture is in decline, by the Note: "Sociopolitical" refers to different circumstances and events, depending on the country and year specified, that lead to involuntary internal displacement. pressures of population growth, and when Factor Mobility and Migration 167 environmental change makes continued In many developing countries, schools, cultivation of certain areas no longer health care centers, hospitals, and public viable. Historically, droughts have had and private amenities are located in areas sudden and prolonged impacts on the dis- of economic activity. With a concentration tribution of the population, particularly in of economic mass, public services can be Sub-Saharan Africa79 and South Asia.80 withdrawn from smaller towns and villag- Another important "push" that pro- es.81 Several studies document the migra- pels internal migration--mundane, but tion to large economic centers by people in no less critical--is the lack of adequate search of better education and health ser- public services (see table 5.4 and box 5.8). vices.82 This movement, though voluntary, BOX 5.8 Migrating to economic density: rational decisions or bright lights? Twenty percent of poor men born in Brazil's the possibility of earning higher wages. belief has resulted in deterrents ranging Northeast--one of the country's lagging But since many migrants do not nd jobs from disincentives to draconian regula- areas--now live in its prosperous South- after moving, this attraction may be irra- tions to limit the movement of people. east. A large demographic shift occurred tional. Some policy makers in developing Recent empirical evidence from four from villages to towns and cities in the countries believe that rather than adding decades of Brazilian census data shows 1970s, and from towns to cities in the 1990s. to the economy in their new neighbor- something di erent. Working-age men Economists have long argued that hoods, migrants subtract from them by migrated not only to look for better jobs migration decisions are motivated by worsening the problems of livability. This but also to get better access to basic public services such as piped water, electricity, Brazilians move to the prosperous Southeast and health care. Results from models of Internal migration of adults, ages 25­35, based on reported region of birth in 1991 and 2001 migration behavior that focus only on the censuses migrant's desire to move in search of better jobs can be biased, because places with better public services also have more job opportunities. Firms like to locate where BRAZIL workers would like to live. By ignoring the importance of public services, some econometric estimates may overstate a migrant's willingness to move in response to wage di erences. N O R T H To determine how much public services matter, a rich data set of public services at the municipality level was combined 762 (1991) 788 (2000) with individual records from the Brazilian NORTHEAST census to evaluate the relative importance of wage di erences and public services in the migrant's decisions to move. Pre- CENTER-WEST 45,672 (1991) 57,131 (2000) dictably, wage di erences are the main factor in uencing migration choices. For the better o , basic public services are 2,839 (1991) Brasília 4,117 (2000) not important in the decision to move. But for the poor, di erences in access to basic public services mattered. In fact, SOUTHEAST poor migrants are willing to accept lower Population that did not migrate wages to get access to better services. A 1991 2000 Brazilian minimum wage worker earning 300,000 14,697 (1991) R$7 per hour (about US$2.30 in February 17,228 (2000) 200,000 1991 2000 2008) was willing to pay R$420 a year to SOUTH have access to better health services, R$87 100,000 15,000 10,000 for better water supply, and R$42 for elec- 30,000 5,000 tricity. Poor migrants are rational. 2,000 Contributed by Somik Lall and Christopher Source: WDR 2009 team, based on census data from the Instituto Brasileiro de Geografia e Estatística. Timmins. 168 WORLD DEVELOPMENT REPORT 2009 Central America it was 16.9 percent, and from the Caribbean it was 42.8 percent.84 BOX 5.9 Too early to tell? The impact of African emigrants Critics of the "brain drain­brain gain" on Africa debate point out that it ignores real-world The brain drain is debated most port of American philanthropists, patterns of international migration. Skilled heatedly for Sub-Saharan Africa. Con- he returned to Ghana to found the workers do not "drain away" as much as cern is justi ed, but emigrants and country's rst nondenominational "circulate" among countries in the world diaspora groups have contributed school, which would later become economy. The benefits of attracting and much to Africa's development. the University of Ghana. Many of the retaining skilled people do not have to be Early accounts of voluntary migra- region's seats of learning have similar distributed in a zero-sum game among tion from Africa tell of small numbers origins. of Africans going abroad to study Most of Africa's independence countries. In addition to the large flows in religious schools and monaster- leaders were part of what might of international remittances, many skilled ies in Europe in the 1700s. Many of have been termed a brain drain in migrants work hard to return to their coun- these students translated the Bible the 1930s and 1940s. Kamuzu Banda, tries with improved prospects as entrepre- into African languages. In doing so, Jomo Kenyatta, Julius Nyerere, and neurs, armed with capital, new skills, and they produced some of the earli- others were from a generation of stu- ideas. Several political, academic, and busi- est attempts to introduce written dents who emigrated to the United ness leaders in developing countries began text to what were predominantly States and Europe and formed plans oral-language traditions. One such to ght for independence. as emigrants (see box 5.9). Cross-country student, Jacobus Capitein, who emi- The economic and social contribu- research on the determinants of economic grated from what would become tion of these emigrants to their coun- growth has not found evidence of a nega- Ghana to study in the Netherlands, is tries of origin are di cult to quantify tive impact associated with the emigration credited for spreading the use of the but impossible to deny, and have of people with skills.85 written word in his native country. made all the di erence to the devel- Another Ghanaian, Kwegyir Aggrey, opment prospects of Sub-Saharan Practical policies for managing from a later generation of emigrants, Africa. traveled to the United States to study migration at Columbia University. With the sup- Source: Easterly and Nyarko 2008. Not everyone chooses to migrate. Moving can be a costly, difficult, and disruptive decision. Indeed, a generation of research is more likely to add to congestion costs in shows that the movements of labor--from cities than to agglomeration benefits. villages to towns, between towns and cit- International brain drains--or gains? ies, across borders in the same region, and There is concern about the volume of from poor to distant wealthy countries-- skilled workers leaving Sub-Saharan Africa are selective. Migrants are not the same as and the Caribbean. As a percentage of the people who stay behind. And while many total stock of highly educated people, the individuals move in search of a better job number of skilled emigrants looks high. On or higher education, many others--par- the whole, though, most skilled migrants ticularly those in the rural areas of low- to high-income countries come from the and middle-income countries--seek basic larger middle-income countries like Bra- schooling and health care for their families. zil and India. Migration prospects in these But this migration is economically ineffi- countries induce more human capital accu- cient. By overlooking the provision of basic mulation, increasing not only the number social services in outlying areas--such as of skilled migrants but also the skills of the schools, primary health centers, and even global workforce generally.83 A "brain gain" basic public infrastructure--policy mak- is likely when the rate of emigration of ers can unwittingly influence the choice to skilled workers from a country is between migrate, motivating households to move 5 and 10 percent. Concern arises for the for reasons other than to exploit economic stunted development prospects of some opportunities. While the move is welfare countries in Sub-Saharan Africa, Central improving for these families, the economy America, and the Caribbean, where the may end up worse off. emigration of skilled labor is much higher. By focusing more attention on provid- In 2000 the rate of skilled emigration from ing education, health, and social services Sub-Saharan Africa was 13.1 percent, from in outlying, economically lagging areas, Factor Mobility and Migration 169 Table 5.5 What does a practical policy toward migration do? Recognize agglomeration benefits. Migration of unskilled labor Migration of skilled labor Internal migration Neutral, but discouraging if agglomeration economies Strongly supportive, particularly to capture agglomeration are unlikely. Policies should encourage migration gains where these are likely. Invest in services in peripheral for economic reasons and discourage migration in areas to build portable human capital. Increase the flow of search of public services. Remove explicit and implicit labor market information, so migrants arrive better informed restrictions as well as place-based service entitlements. of employment possibilities. Cross-border migration within Supportive, particularly for welfare and diversification Supportive, where markets in regional neighborhoods are regional neighborhoods gains from remitted earnings. integrated and gains from agglomeration can spill over to the sending country. Cross-border migration Supportive, particularly for welfare gains from remitted Neutral, as there is a possible foregone agglomeration outside regional earnings. from an accelerated brain drain, but possible gains from neighborhoods knowledge transfer of return migrants, and strong incentives for human capital investment from the prospect of migrating. Source: WDR 2009 team. governments can go a long way toward over other investments, governments can eliminating some of the reasons households eliminate some of the reasons households are pushed to migrate. These efforts can, are pushed to migrate. These efforts can in turn, improve the quality of migration. shape the composition of migration in a Labor mobility that leads to greater concen- way that growing concentrations are more tration of people and talent in locations of likely to add to agglomeration economies, choice will contribute more to agglomera- rather than pile up congestion costs. tion benefits than it adds to congestion. By recognizing the selective nature of The impact of policies on the welfare of voluntary labor migration, and the implica- migrants and the broader economy should tions of increasing returns to scale, the eco- fuel skepticism of attempts to restrict nomic arguments and empirical evidence labor mobility. Encouragingly, there is a in this chapter support a more positive view growing shift away from restrictions on of labor mobility than that held by policy population mobility and toward facilita- makers in poor and middle-income coun- tion and encouragement. But other than tries in the past. From this perspective, a allowing people to move and settle where practical policy stance will differ according they will earn the highest return on their to the human capital endowment of pro- labor and human capital, can govern- spective migrants and whether the agglom- ments do more to help capture the ben- eration spillovers from clustering talent can efits of agglomeration? be captured and taxed by governments (see Migration results from forces that "pull" table 5.5). as well as those that "push" individuals to All the evidence on the benefits of educa- leave. One big pull is the agglomeration tion suggests that policy makers should be economies in cities. But people are also concerned about the rapid loss of talent to pushed out by the lack of social services. In countries far outside their regions. But the Africa disparities in school enrollment and potential costs in forgone human capital neonatal care between cities, towns, and from outright restrictions on skilled emigra- villages are attributable to the near absence tion are high. A far more practical and sus- of schools and health facilities in outlying tainable policy stance would operate along areas.86 Evidence from Central Asia shows two tracks. First, raise the private, individual that in the isolated parts of Tajikistan, costs of acquiring human capital to match schools are inadequately heated, drinking the private individual returns from migra- water is scarce, and arrangements to clear tion of skilled workers abroad. Second, reap garbage and sewage are lacking.87 In China the benefits from diaspora communities in the government is emphasizing a more even the world's prosperous places, by encourag- distribution of basic services to address the ing their economic and political participa- gaps in living standards between the coast tion at home, and by making it easy for them and the interior. By prioritizing education, to retain citizenship, vote, and eventually health, and social services in outlying areas resettle if they so choose. CHAPTER 6 Transport Costs and Specialization T he sharpest insights sometimes transport costs fell low enough that small come from piecing together bits of differences in products and tastes fueled information that separately can be trade between similar countries, at least innocuous and unsurprising. In the mid- in Europe and North America. Neighbors 1970s overseas transport costs had fallen to traded different types of beer and differ- a fraction of what they were in 1900, thanks ent parts of cars, such as wheels and tires. to such inventions as steam power and the Trade in parts and components grew to take telegraph. And the share of trade between advantage of specialization and economies neighboring countries in Europe had risen of scale. The first wave of globalization was relative to their trade with countries more characterized by "conventional," inter- distant. In 1910 British exports were spread industry trade that exploited differences in quite evenly between Europe (35 percent), natural endowments, the second by a "new Asia (24 percent), and other regions (31 international trade" driven by economies of percent). By 1996, 60 percent of Britain's scale and product differentiation. exports went to Europe and only 11 percent Transport costs and scale economies to Asia.1 interact to produce the trade flows observed Singly, neither fact is surprising. in the past half-century.3 The main insight Together, they are exactly the opposite of from research is that the relationships what standard economics would predict. between transport costs, production loca- After all, transport costs should be a larger tions, and trade patterns are nonlinear. part of the cost of goods shipped from Falling transport costs first led to countries half a world away than for goods traded trading more with countries that were dis- with neighbors. So a fall in transport costs tant but dissimilar. When they fell further, should have meant more trade with distant they led to more trade with neighboring partners than with neighbors, not less. countries. Similarly, when transport costs What had happened? fell from moderate levels, production con- Research in the 1980s provides the centrated in and around large markets. answer.2 Two waves of globalization--a In East Asia, as the costs of transporting euphemism for falling transport and trade goods by sea and air fell, the production of costs--were responsible. During the first manufactured goods spread from Japan to wave from about 1840 to World War I, neighboring economies such as Hong Kong, transport costs fell enough to make large- China; the Republic of Korea; and Taiwan, scale trade possible between places based China. Production then moved to South- on their comparative advantage. So Britain east Asia, and now it has moved to China. traded machinery for Indian tea, Argentine With a fall in telecommunication costs, beef, and Australian wool; trade increased large cities in the United States and Europe between distant and dissimilar coun- reaped the rewards of growing markets. 170 tries. During the second wave after 1950, But as the costs of telecommunications fell Transport Costs and Specialization 171 further, services such as accounting and call 50 percent, when the share of intermedi- centers moved to smaller cities in Europe ate inputs in value added is 70 percent.5 As and North America, and then, as they fell transport costs fall, then, trade in intermedi- further, to cities in distant India and the ates would also increase rapidly. Philippines. "Circular causation" also affects trans- Intraindustry trade--the exchange of port. Trade volumes influence transport broadly similar goods and services--is costs. On the trans-Pacific route, cost dif- perhaps the most important economic ferences between a "Panamax" unit of 4,000 development since World War II. Coun- tries trade Samsung, Motorola, and Nokia Figure 6.1 Intraindustry trade is becoming more important for all types of goods, but not in all world regions phones; casings for television remotes; and Evolution of global intraindustry trade, by 3-digit product group, 1962­2006 buttons and stitching for textiles. Such a. By type of goods trade is now more than half of global trade, Grubel-Lloyd Index, 3-digit up from a quarter in 1962. The share of 0.6 intraindustry trade has gone up for all types Intermediate goods of goods and services, from such primary 0.5 goods as oil and natural gas, to such inter- mediate inputs as auto parts and computer 0.4 Final goods help-lines, to such final goods as food and beverages (see figure 6.1, panel a). 0.3 This is important because of the border- Primary goods related divisions identified in chapter 3. These divisions are barriers to movements 0.2 of capital and labor. If all that countries could trade were final goods, such as televi- 0.1 sions and cars, then convergence in living standards would be slow at best. With trade 0 in intermediate inputs, the potential for 1962 1966 1970 1974 1978 1982 1986 1990 1994 1998 2002 2006 specialization and trade increases signifi- Year cantly. The efficiencies generated through b. By region specialization and scale economies in pro- Southern Africa duction and transportation have indeed Western Africa 1962 benefited the world. But these benefits have 1975 not been shared evenly (see figure 6.1, panel Central Africa 1990 b). East Asia, North America, and Western 2006 Northern Africa Europe account for much of the world's Central Asia, Caucasus & Turkey intraindustry trade. This chapter explains why these regions Eastern Africa account for this trade and what this means Western Asia for developing countries. In good measure South Asia the reasons have to do with the interactions Central America & Caribbean between scale economies and transport costs. Transport and trade costs influence South America trade volumes. A 10-percent increase in trade Eastern Europe & Russian Federation costs is estimated to reduce trade volumes Northeast Asia by 20 percent.4 Trade in intermediate goods Southeast Asia & Pacific is especially sensitive to transport costs. If the share of imported intermediate inputs Western Europe in final demand is large, small changes in Australia & New Zealand transport costs can have large effects on the North America volume of trade flows--the "trade friction" 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 increases. For instance, a 5-percent increase Grubel-Lloyd Index, 3-digit, intraregional trade in transport costs can produce trade friction Source: Brülhart 2008, for this Report. equivalent to an ad valorem tax of almost Note: The Grubel-Lloyd Index is the fraction of total trade that is accounted for by intraindustry trade. 172 WORLD DEVELOPMENT REPORT 2009 TEU (20-foot equivalent units, a measure of and Western Europe. Global air and Inter- shipping tonnage) and a mega post-Pana- net traffic maps show a similar imbalance. max unit of 10,000 TEU are 50 percent. But These developments should be disconcert- exploiting these cost advantages requires ing for developing countries not integrated large trade volumes and high capacity, into these self-reinforcing production and because economies of scale are available trade networks. Scale economies in produc- not just in the production of goods and tion and transport will make it more diffi- services but also in their transport. It costs cult, not easier, for developing countries to about $400 to ship a container to the United enter these highly competitive markets. States from China, about $800 to ship from A world of nonlinear relationships India, and $1,300 to ship from Sierra Leone.6 and cumulative causation is a world with China's enormous trade is almost certainly a thresholds. Knowing how developing coun- reason for low transport costs, just as falling tries can get past these thresholds depends transport costs have encouraged countries to on where they are, what they produce, and move production to China. Scale economies the costs that traders must pay. In the devel- in transport mean that falling transport costs oped world, the total trade and transport and increasing trade reinforce one another. costs as a share of the value of goods can be The Northern Hemisphere is heavily split into 20 percent transport costs, 45 per- trafficked, with ever-strengthening trade cent border-related trade costs, and 55 per- links as intraindustry trade flourishes (see cent retail and wholesale distribution costs. map 6.1). But ships sail through or around These costs multiply, piling up to a 170 Central America, South Asia, and Sub- percent tax on the value of goods and ser- Saharan Africa, going only to countries vices traded.7 What they show is that lower that have natural resources such as oil. international transport costs have reduced Trade passages between South America distance but that trade costs due to inter- and the most prosperous parts of the world national division remain high. Meanwhile, are narrow roads, not the busy express- transport costs due to internal distance have ways between East Asia, North America, stayed high even in the developed world. Map 6.1 Busy seafaring in the North, little landfall in the South Intensity of shipping routes during one year beginning October 2004 Equator Shipping routes Container ports Sources: Data from the World Meteorological Organization (WMO) Voluntary Observing Ships' (VOS) scheme, processed by Halpern and others 2008. Note: Container ports shown are the 20 largest by TEU of total containers handled in 2005 (Heideloff and Zachcial 2006). Transport Costs and Specialization 173 Developing countries can learn how related aspects--the naturally monopo- countries have reduced transportation listic nature of transport--have been costs, including how trade has been stim- assumed away. Developing countries ulated and new technologies developed. should do more to address the negative From the analysis of the past two decades, effects of market structure in the trans- they can learn how spatial concentration of port sector. And for some aspects of the production may change as transport costs agenda, they will need international fall. What does this mean for latecomers to support. economic development? The main points: What has happened: two centuries · Better transport technologies devel- of experience oped over the past two centuries have Falling transport costs in the 100 years increased the volume of trade and radi- or so before World War II brought closer cally altered its nature. Before World economic integration within and between War I transport costs declined enough to countries. Then, as in the twentieth cen- make large-scale trade possible, but only tury, the fall was caused by large infra- between countries that were dissimilar. structure investments and breakthroughs They happened to be countries that were in transport technology. distant, because big differences in climate From the early nineteenth century to the and natural endowments usually meant beginning of World War I, the global econ- the countries were in different parts of omy went through what economic historians the world (Indonesia and the Nether- call the "first era of globalization." Domes- lands, for example). During the second tically, canals and then railways greatly wave following World War II, transport reduced transport costs, leading to larger costs fell enough for small differences in integrated home markets and to converg- products and tastes to fuel trade. This led ing prices for manufactured and agricul- to a rise in trade between countries that tural goods. The routing of these transport are similar (for instance, Argentina and links greatly influenced the rise and decline Brazil), which often happen to be neigh- of urban agglomerations. Internationally, bors.8 As transport costs fall, physical steamships lowered maritime transport geography matters less. But with econo- costs and increased the speed and reliability mies of scale in production, economic of service. The results were narrower inter- geography matters more. country price differences, expanding trade · A decline in transport costs--with on routes that the new shipping technology increasing returns to scale--generally could serve, and the emergence of large-scale means more spatial concentration of interindustry trade. production. Recent thinking in eco- Domestic transport. Inland waterways nomics has emphasized the importance and railways reduced intercity and inter- of transport costs in development. With area transport costs dramatically in the high transport costs, large economies of first half of the nineteenth century. Before scale will remain unexploited, and pro- the railway era, which started around 1830 duction inefficient. Efficient production in Europe, most transportation was on is more specialized. When transport roads or--50 to 75 percent cheaper--on costs fall, spatial differences in produc- water. In the United States massive invest- tion and economic growth will increase, ment in canal construction completely both within and between countries. changed interregional trade and shaped a · Developing countries should pay more new urbanization pattern. The construc- attention to transport and communi- tion of the Erie Canal between 1817 and cations regulations to reduce trans- 1825 reduced the cost of transport between port and trade costs. The new economic Buffalo and New York City by 85 percent, geography has inadvertently contributed cutting the journey time from 21 days to to an exclusive policy focus on "hard" 8. Productivity in the U.S. internal trans- infrastructure. The most critical policy- portation sector grew at an annual average 174 WORLD DEVELOPMENT REPORT 2009 of 4.7 percent in the four decades before India's rail expansion had even bigger the Civil War. British navigable waterways impacts. In the 1860s the prices in some quadrupled between 1780 and 1820. French districts were 8 to 10 times higher than in canal construction boomed similarly, and others, and famines were common. The rail continental European countries made a big system reduced transport costs by about 80 step toward overcoming division when the percent, and the coefficient of variation of Congress of Vienna recognized the free- wheat and rice prices fell from more than dom of navigation on the Rhine in 1815. 40 percent in 1870 to below 20 percent in Steamships appeared on important rivers the decade before World War I.14 Lower and lakes in the early nineteenth century, transport costs had little effect on indus- drastically reducing travel times. trial development, however. At the turn of The major nineteenth-century devel- the eighteenth century, modern industry opment in transport was the expansion of employed 2 to 3 percent of India's indus- railroads, which quickly surpassed inland trial workers (about 10 percent of the work- waterways and "performed the Smithian force). Modern factories were concentrated function of widening the market."9 Cities no in two maritime trading hubs, Bombay and longer just provided public services--they Bengal.15 attracted industries with increasing returns International transport. The invest- to scale, reaping productivity effects from the ments in domestic transport created large more specialized inputs and larger labor mar- and integrated home markets. Tariff bar- kets. In the United States the expansion of the riers remained low, and international trade railways had strong effects on the geographic benefited from technical and organizational distribution of economic activity. Illinois, progress, mostly in shipping. Ocean ship- Michigan, and Ohio had marked increases in ping rates differed substantially for routes population, construction, and manufactur- and commodities, reflecting cost differ- ing with the new rail lines within and across ences in harbor technologies, ship types, their borders. One canal after another was and stowage opportunities.16 But overall the abandoned. In 1850 boats carried six times trade costs for grain, the main internation- the freight of railroads; by 1890 railroads car- ally traded good, fell by 40 percent between ried five times the freight of boats. The drop 1880 and 1914 within Europe and between in transport costs narrowed price differences the United States and Europe. This substan- for agricultural goods between local markets tially reduced the price differences between dramatically. The spread in the wheat price exporting and importing countries. between New York City and Iowa fell from 69 Liverpool wheat prices exceeded Chi- percent to 19 percent from 1870 to 1910, and cago prices by 58 percent in 1870, 18 per- between New York City and Wisconsin from cent in 1895, and 16 percent in 1913.17 For 52 percent to 10 percent.10 nonagricultural products, the reduction Railways expanded less in Europe than in price differences was no less impres- in the United States, reflecting the national sive. The Boston-Manchester cotton textile scope of rail systems and the smaller size price gap fell from 14 percent in 1870 to ­4 of European countries.11 The higher freight percent in 1913, while the pig iron price transport intensity of U.S. rail propelled a gap between Philadelphia and London fell further productivity increase. In 1910 the from 85 percent to 19 percent.18 Interna- labor productivity in American railways tional prices also converged in European was 3.3 times that in Britain, a gap that trade. The steamboat initially shifted the had doubled since 1870.12 Russian railway relative importance of trade relations from construction took off after the mid-1860s, European and Asian routes to the North spreading wheat and rye production with Atlantic routes. Steamships could not serve the narrowing of regional price differences. Asia until the opening of the Suez Canal The export share of Russian agriculture because coal was not available on the long increased from 29 percent of the grain pro- route around Africa.19 duced in European Russia to more than 42 During this first era of globalization, percent between 1906 and 1910.13 increasing competition from abroad due to Transport Costs and Specialization 175 declining transport costs gave rise to pro- Empirical assessment therefore depends tectionist trade policies. In North America, on the estimation of transport costs.) One during the Civil War, tariffs reduced the study in France shows that truck transport financial burden on the federal govern- costs fell by 33 percent between 1978 and ment, and they remained high after the war 1998,23 with substantial regional varia- ended. Continental Europe shifted away tion due to the differences in the quality from liberal trade policies in the late 1870s of roads and the charges for road use. The in response to cheap American and Russian main contributors were the deregulation of grain. Tariffs were reintroduced on finished the trucking industry (a reduction of 21.8 manufactured and agricultural goods. percentage points) and the lower vehicle costs (­10.9 percentage points). Transport Increasing "transport intensity" and infrastructure (­3.2 percentage points) and intraindustry trade in the modern era declining fuel costs (­2.8 percentage points) Freight costs have about halved since the were much less important. mid-1970s,20 driven by investments in Rail freight costs. Rail costs fell much transport infrastructure, better capacity less than road costs. Technical progress use, and technological progress. Recent was uneven across rail submarkets, and trends differed from those in the first era of the monopoly power of large, mostly state- globalization:21 owned enterprises slowed cost reductions (see box 6.1). Obligations to serve regions · The major cost declines have been in road with small transport, for instance, have and air transport. Maritime transport went through the containerization revo- lution without reducing costs overall. · Biggest in the world: size and social obligations The surge in international trade has been BOX 6.1 within industries, not between them, as of Indian Railways in the first episode of falling trade costs. The railway industry exhibits increas- to the bundling with infrastructure · Reduced trade friction has been less a ing returns to scale in two ways. First, services. It owns and runs factories for consequence of falling transport costs network economies and economies locomotives, coaches, and even their than of a drop in freight costs as a share of density lead to size advantages at parts. Long transport distances on the Indian subcontinent should give the of the value of goods traded. Most of the the rm level. Second, rail transport operations are almost universally railways a stronger competitive edge increase in trade has been in easily sub- over roads. Indeed, Indian Railways stitutable goods.22 combined with the supply of infra- structure services, granting rail rms makes 70 percent of its revenues and · Transport reforms and falling trade bar- a natural monopoly, at least locally. most of its pro ts from freight, cross- riers have contributed substantially to Given the importance of the railways subsidizing the loss-making passenger the fall in transport costs. for economic development and the sector. The overpricing of freight ser- enormous market power of rail rms, vices is one reason it has lost business · Falling communication costs, interact- it is not surprising that many rail com- to roads in recent years. ing with falling transport costs, have been panies are state owned. Curtailing the potential to provide instrumental in fragmenting production The biggest of these mammoths low-cost freight transport over long processes and outsourcing intermedi- is Indian Railways. The Guinness distances are extensive social obliga- ate goods production. Relative wage dif- Book of World Records lists it as the tions. Net social service obligations in ferences have become more important world's largest commercial or utility 2005­06 were more than Rs 47 billion, employer, with more than 1.6 million plus welfare costs of Rs 9.6 billion. The because of the lower costs of managing employees. It moves more than 16 service obligations include shipping production processes over long distances. million passengers and more than essential commodities (sugar cane, Road transport costs. Road transport 1 million tons of freight each day. In livestock, paper) below cost, hav- 2002 it ran 14,444 trains daily, 8,702 ing freight subsidize passenger and costs have fallen substantially, by almost 40 of them for passengers, and owned other coaching services, and opening percent over the past three decades, despite 216,717 wagons, 39,263 coaches, and new unpro table lines. A major part higher energy and wage costs. (Compre- 7,739 locomotives. of the passenger transport de cits hensive statistics on prices for transport Founded in 1853 as a system of 42 covered by freight are urban and services do not exist, and the implementa- rail systems, it was nationalized as suburban losses in Chennai, Kolkata, tion of price indexes as part of the system one unit in 1951. Vertical integration and Mumbai. of national accounts is still in its infancy. of Indian Railways is not con ned Source: WDR 2009 team. 176 WORLD DEVELOPMENT REPORT 2009 motivated demands for public subsidies and quantities of bulk commodities on charter, cross-subsidies from profitable routes. with shipping prices set in spot markets. Rail costs are specific to the commod- There are no fixed schedules or routes, so ity shipped. For the United States, this has shipping is determined by current market been shown to depend on price discrimina- demand. Liner shipping is used for general tion by the freight railroads among shippers cargo on fixed trade routes and on a fixed of different commodities.24 There was no timetable. The liner trade is organized into uniform development of rail freight rates cartels, or conferences, which discuss and from 1981 to 2004. Markups for coal and coordinate prices and market shares. grain have increased significantly. Mark- Technical progress and institutional ups in intermodal traffic have been lower changes have reduced prices in both sub- because of competition from trucking and markets. The most important are the rail-to-rail competition between major growth of open registry shipping, the cities. Decreasing or flat rates had been scale effects from the enormous increase observed for shipping chemical products in maritime transport demand, the intro- and automobiles. This mainly indicates the duction of containers, and the resulting high value of these goods. Freight demand changes in port logistics. Open registry is a derived demand, and the prices shippers shipping is the practice of registering ships are willing to pay increase with the value of under flags of convenience (Liberia and the shipments. Panama) to circumvent higher regulatory Air transport costs. With the arrival and manning costs imposed by wealthier of the jet engine, air transport costs came nations. Open registry fleets did 5 percent down quickly from the mid-1950s to the of world shipping tonnage in 1950, 31 per- early 1970s. Jet engines were faster, more cent in 1980, and 48 percent in 2000.27 It reliable, and more fuel efficient than the is estimated that vessel expenses for open piston engines they replaced (see box 6.2). registry ships are 12 to 27 percent lower Quality-adjusted real prices of aircraft fell than those of traditional registry fleets, by 13 to 17 percent annually from 1957 to with most of the cost differences coming 1972.25 Technical progress slowed consider- from labor costs.28 ably after 1972, but prices were still falling Cost reductions because of scale effects by 2 to 4 percent a year from 1972 to 1983. come from greater vessel capacity and insti- Between 1955 and 2004, air freight prices tutional changes. The rapid expansion of fell from $3.87 per ton-kilometer to less maritime transport demand seems to have than $0.30, in 2000 U.S. dollars. Average accommodated these changes and reduced revenue per ton-kilometer fell 8.1 percent the danger of preemptive competition.29 a year in 1955­72 and 3.5 percent a year in The increase in vessel size seems to have 1972­2003. Despite this significant decline allowed for hub-and-spoke economies-- in nominal air freight rates, the trade fric- smaller vessels move cargo to a hub, where tion in air transport did not fall as dramati- shipments are aggregated into much larger cally. The price of air shipping in real U.S. and faster ships for longer hauls. Prime dollars per kilogram increased 2.9 percent examples are Hong Kong, China; Rotter- annually from 1973 to 1980, in part due to dam; and Singapore. oil price increases, and then declined by Vessels for bulk commodities, refriger- 2.5 percent annually from 1980 to 1993. ated produce, and automobiles are profitable The post-1980 decline varied substantially on individual routes. Since the mid-1980s, among routes, with longer routes and North dedicated "juice tankers" have cemented America showing the largest drops.26 After Brazil's dominant position in the global 2001 the real price of inbound air freight export market for orange juice, almost all to the United States rose sharply, possibly produced in São Paulo State. Standard- reflecting higher security costs. ized containers provide cost savings across Maritime transport costs. Two submar- transport modes--long-distance truck, kets have developed differently over the past inland waterways, rail, and short-distance decades. Tramp shipping is used for large truck--because goods do not need to be Transport Costs and Specialization 177 BOX 6.2 The jet engine An estimated 320 million people meet maritime shipping, but air transport in the morning arrive on Amsterdam's annually at professional and corporate lls an important niche in just-in-time markets by evening. Horticulture is now events after traveling by air. Of the world's production systems. While shipments among the top three export earners $12 trillion of merchandise trade, 35 per- by sea are routine, rms use air cargo (with tourism and tea). In 2007 the sec- cent by value was shipped by air in 2006.a to ne-tune intermediate input ows tor's free on-board (FOB) export value The estimated economic rate of return and to ship goods with high value-to- was 43 billion Kenyan shillings (about from investments in aviation infrastruc- weight ratios. Even for Brazil, known US$650 million), and the Kenya Flower ture and services is 56 percent in Kenya, for its primary goods exports, air cargo Council estimates that the livelihood of 28 percent in Jordan, and 19 percent in in 2000 accounted for 0.2 percent of 1.2 million people depends directly or Cambodia.b The reason for all this is the jet total export volume by weight, but indirectly on the industry. By contrast, engine, perhaps the most signi cant inno- almost 19 percent by value.c Inciden- Bangladesh's lack of cold storage facili- vation in long-distance transport ever. tally, Brazil is also home to the world's ties and refrigerated air cargo capacity The jet is safer, easier to maintain, better third-largest airplane maker, Embraer. has blunted its opportunities to export suited for longer distances, and more Prime examples of sectors bene ting high-value fruits and vegetables to the fuel e cient than the propeller. Since it from air transport are semiconductors Middle East.e revolutionized air travel in the 1960s, it has and fashion. Shipments of semicon- · Mass tourism in developing countries. become so closely identi ed with aircraft ductors are so highly correlated with In 2005 tourism receipts in low- and propulsion that one wonders how the air- air freight overall that they are con- middle-income countries were about craft industry managed to make so much sidered a key leading indicator for the $200 billion,f thanks mostly to inex- progress with pistons. sector's health. Product cycles in the pensive air travel. Charter ights But as with many path-breaking fashion industry have shortened so provide even larger cost reductions inventions in transport technology, the much that one Spanish clothing chain through packaging with other ser- gestation period between invention ships merchandise straight from fac- vices and high-capacity use. Airport and economic success was long. Frank tory to store, replacing designs twice construction in tourist areas gener- Whittle in Great Britain, in 1929, and Hans a week. The need to respond quickly ates clusters of development with von Ohain, a German physicist, in 1933, to changing customer tastes has led to a high density of complementary independently developed concepts for the relocation of some of its produc- services and thick and specialized jet propulsion. Jet engine technology tion from East Asia to Spain and nearer labor markets. Between 1990 and progressed quickly after World War II. The countries like Morocco and Turkey. 2005, tourist arrivals in Sub-Saharan breakthrough in commercial passenger From there, clothes are sent to stores Africa increased by 8 percent a year-- travel arrived with the Boeing 707 and elsewhere in the world: "Planes from from 6.8 million to 23.6 million--and Douglas DC-8. Earlier jet aircraft were Zaragoza, Spain, land in Bahrain with tourism receipts, from $4.1 billion to noisy and had higher operating costs goods for Inditex stores in the Middle $14.5 billion. Tourist arrivals in China than advanced piston-engine aircraft. East, y on to Asia, and return to Spain grew almost 10 percent annually. They could compete only on speed and with raw materials and half- nished Cambodia now receives more than greater seat capacity. But in the early clothes."d 2 million tourists a year, Vietnam 1960s, technology improvements (the so- · Enabling exports of perishable goods about 4 million--16 times as many as called by-pass engine) rang in the end of over long distances. Inexpensive and in 1990. propeller-powered long-distance travel. frequent air service has allowed coun- Within ve years, prices per ton-kilometer tries like Chile, Colombia, and Kenya Source: WDR 2009 team. fell by about 40 percent. to sell agricultural and horticultural a. International Air Transport Association Jet aircraft have a much higher power- products to markets in Europe, the 2007b. to-weight ratio, which enables longer Middle East, and North America. A b.International Air Transport Association range, faster travel, and bigger payloads. prime example is Kenya, which today 2007a. Higher quality and lower cost had a large has a third of the global market for cut c. Sanchez and others 2003. impact in many sectors. d. Rohwedder and Johnson 2008. owers. Naivasha in central Kenya hosts e. Dixie 2002. a highly e cient cluster of growers, f. World Tourism Organization (UNWTO) · Supporting buyer-supplier networks over showing that localization economies 2006. long distances. Most global trade is by also exist in agriculture. Flowers picked unloaded and reloaded (see box 6.3). Con- half to two-thirds of a ship's lifetime.30 And tainerization reduces direct port costs for it allows for larger and faster ships, which storage and stevedoring. It also reduces the reduce the costs per ton-mile while the ship indirect capital costs of idle capacity during is steaming. These cost reductions on the long port stops, which previously made up ocean leg have more than compensated for 178 WORLD DEVELOPMENT REPORT 2009 BOX 6.3 The big box About 90 percent of nonbulk cargo The Japanese government was the rst Container transport has continued to worldwide is transported in contain- to support the expansion of containeriza- increase at enormous rates. The boxes ers stacked on trucks, rail wagons, and tion. In 1966 the Shipping and Shipbuild- keep getting larger, with the standard freight ships. In 2007 more than 18 million ing Rationalization Council urged the FEU size giving way to 48-foot and containers made more than 200 million Ministry of Transport to eliminate exces- 53-foot boxes that allow trucks to haul trips. Containerization has even changed sive competition to bene t from the new more freight on each trip. The world's how port and ship capacity or maritime technology. It persuaded the government eet is expanding steadily, with the transport services are measured. Cargo to build container terminals in the Tokyo- capacity of pure containerships rising by shipped is now measured in TEU or Yokohama and Osaka-Kobe areas. The 10 percent annually between 2001 and 40-foot equivalent units (FEU). A TEU is rst container cranes began operation 2005. The size of the vessels has been the measure of a box 20 feet long and 8 in 1968. But highway regulations barred increasing, too. Dozens of vessels able to feet wide, with a maximum gross mass of full-size containers, and the Japanese carry 4,000 FEU joined the eet in 2006, 24 metric tons. National Railway was not equipped to and even larger ones were on order. The The revolution is popularly attributed carry containers longer than 20 feet. Emma Maersk (396 meters long), launched to Malcom McLean.a He owned a truck- In the United States, Matson Navigation in 2006, can carry more than 14,500 TEU. ing rm in New Jersey and had a simple won government approval to operate an Of all tra c, 26 percent now originates in insight: packages being shipped gener- unsubsidized container service between China. ally need to be opened only at origin and the U.S. West Coast and Hawaii and East Geography and topography limit the destination, but unloading and repacking Asia. The company had visions of unload- ever-increasing size of ships: Because costs a lot of money. In 1956 he inaugu- ing cargo at Oakland directly onto special the Panama Canal lost more tra c with rated the Sea-Land Service, with his con- trains that would carry it east. On the the old locks unable to accommodate verted tanker ship, the Ideal-X, setting sail return trip, the company planned to carry vessels larger than 5,000 TEU, it now is from Newark, New Jersey, for Houston, military cargo for the U.S. bases in Japan being expanded to allow ships up to Texas, carrying 58 aluminum truck bodies and the Republic of Korea. Business could 12,000 TEUs to pass. Most of the con- in frames installed atop its deck. not start before Matson entered a joint tainer ships are too large for the Suez The idea did not spread widely until venture with a Japanese partner, and the Canal as well. Container ships have more than a decade later, when the U.S. containership that completed its maiden an absolute size, limited by the depth armed forces needed e cient military voyage in 1968 from Japan to the United of the Straits of Malacca, linking the transport to Vietnam. Against consider- States was owned by Nippon Yusen Kai- Indian Ocean to the Paci c Ocean. This able resistance, McLean won contracts to sha Line. Six weeks later, McLean's Sea- "Malacca-max" size constrains a ship to build a container-port at Cam Ranh Bay Land Services started a regular service dimensions of 470 meters long and 60 and to run containerships from California between Yokohama and the U.S. west meters wide. to Vietnam. Without the containers, the coast. And what happened to Malcom U.S. military would have had a tough Once the infrastructure facilities were in McLean's company? Sea-Land grew and time feeding, housing, and supplying place, container tra c took o . By the end was the biggest shipping company in the 540,000 soldiers, sailors, marines, of 1968 the Japan­U.S. route was crowded 1995. The Danish company Maersk was and air force personnel in Vietnam in with containerships, seven companies second, followed by Evergreen. Four 1969. From almost nothing in 1965, Sea- competing for fewer than 7,000 tons of years later, Sea-Land was acquired by Land's Defense Department revenues eastbound freight each month. The speed Maersk. By 2000 Maersk-SeaLand had rose to $450 million between 1967 and of expansion was determined by port and a slot capacity of about 850,000 TEU. 1973. Routes to Okinawa and Subic Bay rail infrastructure. In the United States, rail McLean's big box is here to stay. in the Philippines were added later, but intermodal tra c tripled between 1980 McLean's business remained restricted to and 2002, from 3.1 million trailers and con- Source: WDR 2009 team. military logistics. tainers to 9.3 million. a. Levinson 2006. higher investment costs and higher costs port cost savings of containers are much for slack time in ports.31 lower. But containerization has also concen- How did this technological change affect trated shipping capacity in a few global shipping costs? When adjusted for inflation, ports. Most developing countries have been the real price for tramp services in 2004 was slow to containerize, because of their small about half the real price in 1960.32 But when trade volumes and different factor prices. deflated using the commodity price index, Where capital is scarce and labor abundant, there are large fluctuations but no down- the capital cost of specialized cranes, stor- ward trend. This means that the trade fric- age areas, and rail heads is higher, and the tion resulting from transport costs for bulk Transport Costs and Specialization 179 commodities typically shipped by tramp reduced weight-value ratio in all inter- trip charters has not gone down--the price national transport. A second is the lower of transporting a dollar's worth of iron ore price of air transport relative to maritime or wheat has not fallen. Liner prices show a transport. Goods that traditionally have steady rise before peaking in 1985, based on been transported by sea are now shipped by long time-series from the German Ministry air. After accounting for the changes of the of Transport. The price index for liner ship- weight-value ratios, the modal shift, and ping emphasizes general cargo, including the changes of routes, the ad valorem tax containerized shipping and manufactured equivalent of maritime transport fell more merchandise. It also covers loading and than that of air transport (see figure 6.2).35 unloading expenses, which are particularly Changes in the composition of goods and relevant because reductions in cargo han- trading partners reduced the market fric- dling costs are thought to be a major source tion of transport, not its costs. of gains from containerization. Measured Logistics, time, and international trade. relative to the German GDP deflator, liner Transport services are not a homogeneous prices declined until the early 1970s, rose good, and transport costs are not product- sharply from 1974 to reach their peak in the or place-neutral. The revenue figures and 1983­85 period, and declined afterward. price indexes do not indicate quality or These trends in shipping costs run coun- speed. Shipping containers from Europe to ter to public perceptions of continually fall- U.S. destinations still requires two or three ing trade costs. Two possible explanations: weeks--from Europe to Asia five weeks. But First, these price trends do not capture air shipping requires a day or less to almost the true cost savings of containerization, anywhere in the world. With the decline in Figure 6.2 Air freight costs are since they do not factor in the total cost air transport costs, the price of speed has down less than ocean freight costs of door-to-door transportation. In 1956 fallen dramatically. Air freight the loading of loose cargo cost $5.83 a ton. This matters for trade. Every day in ocean % of value shipped When containers were introduced in that travel that a country is distant from the 15 year, the loading cost was less than $0.16 importer reduces the probability of sourc- Fitted ad valorem rate a ton.33 So the main savings came from ing manufactured goods from that coun- lower intermodal transfer costs. Contain- try by 1 percent.36 And exporting firms are 10 erization allowed goods to be packed only willing to pay 1 percent of the value of the once and shipped over long distances using good per day to avoid time losses associated 5 maritime, rail, and road transport. Second, with maritime transport. With the recent the quality of transport and logistics ser- increase in the intensity of international Unadjusted ad valorem rate vices increased markedly, particularly their trade, the demand for speed has increased. 0 speed and reliability. The absence of a more Goods with the highest time sensitivity have 1974 1984 1994 2004 significant price decrease is thus explained, seen the fastest increase in trade. Examples Ocean freight at least in part, by a greater willingness to are perishable agricultural goods and those % of value shipped pay for higher quality services. with short product cycles such as fashion 10 Fitted ad valorem rate articles, where consumer preferences shift, Small declines in transport costs, or electronics, where the latest technology but a big easing in trade friction earns a premium. Such cycles are important 8 Cost information suggests that international not just for Europe, North America, and transport costs have not dropped as much Northeast Asia but also for China, India, as is commonly believed. Real prices of air and Southeast Asia. 6 and maritime transport have not fallen or Faster transport can speed the changes in Unadjusted ad valorem rate risen much since the 1970s and early 1980s. the geography of trade. Production locations 4 But the ad valorem transport charge--the for textiles and electronics were initially 1974 1984 1994 2004 cost of transport as a share of the value of driven by wage costs. But with short product Source: Hummels 2007. the traded good--has gone down. Explain- cycles, shorter transport times may outweigh Note: The unadjusted ad valorem rate is the ratio of freight costs to import ing the decline are changes in the composi- higher wage costs, leading to relocations. good value. The fitted ad valorem rate is tion of traded goods and the composition Some apparel production outsourced from derived from a regression and controls for changes in the mix of trade partners and of trading partners.34 One change is the the United States to Asia has relocated to products. 180 WORLD DEVELOPMENT REPORT 2009 higher wage locations in the Caribbean and networks.39 But there are two more direct Mexico.37 Short product cycles, and more effects. The first is to reduce search costs. generally uncertain demand, are forces Because knowledge about potential cus- for agglomeration as firms need to locate tomers or suppliers in foreign countries near suppliers.38 But with more predictable is imperfect, trade relations start with the demand, faster speed might contribute to search for trading partners. The search outsourcing stages of production (compo- depends on the quality of the communica- nent production, research and development tion infrastructure, which is largely a fixed [R&D], and assembly) to other countries cost and therefore increases the intensity of according to comparative advantage. international trade as it reduces the search Communication costs. The cost of a cost for trading partners.40 The second is three-minute phone call from New York to reduce variable trade costs. These costs to London fell from $293 dollars in 1931 arise from the need of consumers and pro- (in 1993 dollars) to around $1 in 2001 for a ducers to interact on product specifications, much better connection--and to just a few quality control, and timing.41 They are low cents today (see figure 6.3). The Internet for homogenous goods traded on organized and other telecommunication advances exchanges or with reference prices. But they have lowered communication costs, reduc- are high for differentiated goods.42 Since ing even more the trade friction for physi- these kinds of goods are most prominent in cal goods, especially intraindustry trade. trade within more disaggregated produc- But they have had an equal if not greater tion processes, the line for communication impact on the trade in services. Yet many costs played a big role in the recent surge in tasks that require intensive communica- intraindustry trade (see figure 6.1). tion hardly have been affected. Direct per- Low communication costs make it pos- sonal interaction and face-to-face contact sible to control production processes over remain an important agglomerative force, long distances by computer-aided control especially and paradoxically in the most systems and online communication, reduc- communication-intensive industries. ing the need to co-locate management and Lower communication costs facilitate the technical staff with unskilled workers. This coordination of international production allows vertically integrated companies to outsource production to low-wage coun- Figure 6.3 The costs of communicating have fallen tries. But it also facilitates the breakup of to a fraction of what they were a decade ago production processes into supply chains Average cost of a telephone call to the United States of different companies distributed across US$ per 3 minutes countries and continents.43 8 Low communication costs are particu- 7 larly important for offshoring tasks that 6 do not require the shipment of physical 5 products, such as business, professional, 4 and technical services, including account- 3 ing, bookkeeping, computer programming, 2 and information and data processing. U.S. imports of these services increased by more 1 than 66 percent in real terms between 1997 0 1997 1998 1999 2000 2001 2002 2003 2004 and 2004. The shift of jobs to foreign coun- Year tries has stirred fears among white-collar workers of massive labor market adjust- East Asia & Pacific Latin America & Caribbean ments and has reduced political support Europe & Central Asia Middle East for open trade regimes.44 Contrary to these & North Africa High-income, fears, business service imports in rich coun- OECD countries South Asia tries have remained fairly low. In the United Sub-Saharan Africa States, the import share of computing and Source: World Bank 2007j. business services reached only 0.4 percent Transport Costs and Specialization 181 in 2003, and exports of these services are it allows more efficient sharing of facilities even higher. and services. Recent research also explains The biggest outsourcers, in relation two somewhat unexpected consequences to local value added of these services, are of falling transport costs: (1) at the inter- small countries like Angola, Mozambique, national level, trade increases with nearby Papua New Guinea, and the Republic of countries, not with those farther away, and Congo. Among the advanced economies, (2) within countries, improving transport Germany, a country with high absolute infrastructure may lead to more concentra- imports in business services ($39 billion in tion of economic activity, not less. 2002), ranks 59th with a share of 2.9 per- Research has been far less successful in cent. The United States, with the highest showing why falling transport costs may absolute import value in business services make it more difficult for developing coun- ($41 billion in 2002), ranks 115th. But the tries and lagging regions to break into world United States was the biggest exporter of trade--indeed, increasing returns in the services ($58 billion in 2002) and so was transport sector have often been ignored in a substantial net exporter. Ireland shows formal models. But just as falling transport that trade in immaterial services tends to costs facilitate economies of scale in pro- be within industries rather than between duction, higher production and trade pro- them. It is the largest exporter of computer duce economies of scale in transport. and information services and the fourth- largest importer. Still its ratio of exports Falling transport costs create bumpy to the local value added of computer and economic landscapes information services was only 16 percent (9 Before the recent accelerated drop in trans- percent of GDP). port costs, natural or "first-nature" geo- So the relocation of back-office services graphic conditions (such as waterways) to foreign countries is not a large threat to largely determined the location of settle- employment in advanced countries. Trade ments and the spatial arrangement of pro- balances in business services in almost all duction and trade. Shared investments developed countries have been positive and then created increasing returns to scale that increasing from 1981 to 2001. India, seen shaped economic geography. Such invest- to attract many business services from rich ments could include local health and educa- countries, had a smaller increase in output tion facilities or markets and other services in this sector from 1995 to 2001 than did the that reduce trade and transaction costs-- United Kingdom.45 In short, the impression such as enforcing property rights, resolv- that services drive economic dispersion ing contract disputes, or identifying market across countries is not confirmed by the evi- opportunities. The more the people who use dence. Trade in these services has increased a facility or communal service, the lower the a lot, but for both imports and exports. For costs per user. The larger the settlement, the most countries, the share in local services more the people who share the fixed costs. remains small. And when business has To use the service, people and goods have been outsourced, much of it has remained to travel. So as transport costs fall, access concentrated in a few places. Low commu- increases, scale increases and the unit cost of nication costs have had little effect on cre- provision drops. This is how transport costs ative activities and high-value services that define the geographic size of markets and require frequent personal interaction. the reach and scale of communal services. As more facilities and services are pro- Transport costs and scale vided centrally in larger cities, smaller com- economies: two decades of analysis munities become less attractive and spatial The evolution of transport costs, a criti- disparities emerge--the size distribution cal factor in economic geography, helps of towns and cities changes. First-nature explain the experience in the previous sec- geography and the lumpiness of urban tion. A fall in transport costs increases the infrastructure investments result in irre- concentration of people and firms because versible dynamics that determine how the 182 WORLD DEVELOPMENT REPORT 2009 economic landscape first becomes rough, favoring trade between countries with differ- then bumpy. ent endowments. Countries traded because Economic historians had long recog- they could not produce the imported prod- nized that these processes, driven by chang- ucts themselves--bananas from Central ing transport costs, are critical for economic America to Europe for cars in return. But development.46 And geographers and plan- with differentiated goods, trade is within ners formalized the effect of indivisible classes of goods rather than between them. communal facilities in differentiating city Countries trade because they want slightly functions and sizes in the "central place different versions of similar goods--Japan theory."47 Economists went beyond first- and Sweden trade Toyotas for Volvos. In nature geography and public goods. They other words, in the old trade theory and realized that increasing returns to scale in with high transport costs, countries trade the production of manufactured goods and only what they need to. In the new trade ideas further influence the distribution of theory and with scale economies, a love of economic activities in geographic space. variety, and low transport costs, countries With urbanization, manufacturing and trade because they want to. allied services become the drivers of growth, International trade surged between as discussed in chapter 4. These forces inter- (often nearby) countries of the North- act with transport costs to determine the ern Hemisphere in the 1960s and 1970s, spatial evolution of economies, at all spatial even though these countries have essen- scales--international, national, and local. tially similar resource endowments. Trade between rich and poor countries was ini- Falling transport costs increase trade tially dwarfed by these developments. In the between neighbors beginning of the 1980s intraindustry trade The growing demand for varieties of simi- between medium- and high-income coun- lar goods helps explain the paradox that tries expanded--and later between other falling transport costs have led to more categories of countries (see figure 6.5). The trade between countries that are close by differentiation of demand--that is, the love and have similar characteristics. In fact, of variety--and intraindustry trade did not over the past 40 years, distance has become remain confined to rich countries. a larger deterrent to trade while divisions-- Accompanying the surge in intraindus- border effects--have become less of a deter- try trade was a large increase in trade in rent (see figure 6.4). intermediate goods relative to final goods. Traditional trade theory did not consider Intraindustry trade in intermediate goods the increasing returns to scale and the dif- requires an especially efficient transport ferentiation of demand. It predicted more sector. The ability to coordinate and con- intensive trade in goods that are different, trol production processes in real time by computerized systems has been central to Figure 6.4 Distance has become more of a deterrent, divisions less the vertical disintegration of production Coefficients for distance and national borders in trade models, 1960­2005 processes in the high-income countries and Distance Division the outsourcing to medium-income coun- Distance coefficient Border effect coefficient tries.48 So lower transport costs, changes in 1.4 400 goods traded, and lower communication costs reinforce each other. 1.2 One might expect that goods with low 200 value-to-weight ratios would be traded 1.0 mainly over short distances. But prod- uct cycles for knowledge-intensive intra- 100 0.8 industry goods and for consumer items such as electronic gadgets and fashion arti- 0.6 50 1960 1970 1980 1990 2000 1965 1970 1975 1980 1985 1990 1995 2000 2005 cles have become shorter. This greater time Year Year sensitivity helps explain why the distance- Source: Mayer 2008 for this Report. dependence of trade goes up rather than Transport Costs and Specialization 183 down. If countries want to benefit from Figure 6.5 Global intraindustry trade is no longer confined to rich countries current trends of globalization, regional Grubel-Lloyd Index, 5-digit coordination of infrastructure investment 0.6 and transport policies becomes even more important.49 0.5 How do increasing returns to scale in production, the love of variety for consumer 0.4 and intermediate inputs, and lower trans- port costs drive concentrations of economic 0.3 activities in geographic space? First, differ- entiated products and increasing returns 0.2 to scale will increase productivity more in larger areas or countries than in smaller 0.1 ones, even if they have identical per capita 0 resources and access to the same technol- 1962 1966 1970 1974 1978 1982 1986 1990 1994 1998 2002 2006 ogy. The important dimension of size is Year the volume of overall demand or economic High-income to high-income Low-income to high-income mass, not the size of the land area. When Medium-income to high-income Low-income to low-income such agglomeration forces are considered, Low-income to medium-income Medium-income to medium-income both Hong Kong, China, and Singapore are Source: Brülhart 2008 for this Report. viewed as "large." Note: The Grubel-Lloyd index is the fraction of total trade that is accounted for by intraindustry trade. Second, the larger a region, the more varieties or intermediates will be produced equal geographic distribution of economic locally. Compared with smaller regions, activities. High concentration is seen as a fewer goods have to be imported, saving on problem, and the spatial redistribution transport costs. People with equal nominal of economic mass is expected to promote incomes thus have a higher real income in overall development. Massive transport the larger regions, and firms realize cost infrastructure investments have been the savings. central policy instrument to induce firms Third, the higher real incomes will lead to move to lagging regions. But the out- to in-migration, putting pressure on local comes were usually the opposite--the tar- wages. Lower wages will attract more firms, get regions lost production and workers to making the larger market even larger and the leading regions (see box 6.4). leading to a new round of circular causa- Knowledge-sharing is another force tion of firm relocation, higher real incomes, shaping the economic geography of coun- and a larger market. Chapter 9 discusses in tries and areas. Technical know-how can more detail how developing countries can be used by more users at no or small extra address the challenging task of regional cost. It is embedded in an experienced integration, learning from the experiences workforce, and the accumulated stock of with institutional cooperation, regional knowledge leads to innovation. Larger infrastructure, and coordinated incentives local labor markets increase knowledge around the world. spillovers between workers and thus increase productivity nonlinearly.50 In big Falling transport costs lead to cities the benefits may not fully material- concentration within countries ize because congestion and fragmentation The productivity and income benefits of hinder interaction. But well-functioning agglomeration, driven largely by lower urban transport systems can increase transport costs, are often difficult for plan- the effectiveness of the labor market and ners and policy makers to accept. But they spread the results of learning on the job explain the second counterintuitive impli- (see box 6.5). cation of falling transport costs. There is Falling transport costs enhance localiza- a strong belief that an equal distribution tion economies in the production of knowl- of transport infrastructure will induce an edge and information--say, for business, 184 WORLD DEVELOPMENT REPORT 2009 BOX 6.4 Italy's intervento straordinario: an unexpected response to falling transport costs Regional disparities are caused by the trated on the areas where emigration was, of the mani puliti ("clean hands") of the unequal distribution of infrastructure, in fact, highest.b From the beginning of the country's judiciary. These scandals con- and infrastructure investment in lag- 1980s, when the original mandate of the tributed to the dissolution of the Christian ging areas will reduce these imbalances. Cassa ended, it was kept alive by 11 minis- Democratic Party and the Socialist Party That is the common assumption. But, terial decrees. In 1986 the "extraordinary and to the emergence of the Northern frequently, the industries intended to intervention" was re nanced up to 1993. League, which demanded the separation prosper from these investments move The total annual expenditures of the of the North from the South to end the elsewhere, accompanied by a mass out- Cassa rose to a peak of 3,750 billion lire waste in the Mezzogiorno. An intervention migration of workers. A prime example is (US$4.5 billion) in 1976, declining to 2,650 to make the country more uniform may the Italian regional policy to reduce the billion lire (US$2.1 billion) in 1991, and col- have increased internal divisions. development di erences between the lapsing afterward. The money had little Chapter 8, "Unity, Not Uniformity," North and the South. The Mezzogiorno e ect on economic indicators in the Mez- discusses how countries have promoted has become a generic term for a region zogiorno (see box table). national integration by using a calibrated that su ered from the good intentions of The unemployment rate fell until the blend of spatially blind institutions, con- regional policy.a beginning of the 1970s because of the out- nective infrastructure, and spatially tar- A short-term interventostraordinario was migration of millions of workers to Northern geted interventions. managed by a special agency, the Cassaper Italy and other countries. It then more than il Mezzogiorno, set up in 1950. It was sup- doubled up to the end of the 1980s, indicat- ported by the International Bank for Recon- ing a rapidly growing dependence of the Source: WDR 2009 team. a. Boltho, Carlin, and Scaramozzino 1997; struction and Development (IBRD), led by South on scal transfers from the North. Sinn and Frank 2001. Paul Rosenstein-Rodan, who developed the Scandals surrounding the Cassa per il b. By that time, some critics of the Mezzo- Big Push Model of economic development Mezzogiorno were disclosed as part of giorno policy demanded that funds in the 1940s. The development impact tangentopoli ("bribesville") by the e orts assist outmigration (Lutz 1962). was to come from massive infrastructure investment, with much emphasis on road Economic development of the Mezzogiorno building and railways. It soon became obvi- ous that short-term success would not be 1951­60 1961­70 1971­80 1981­90 1990 achievable, leading to repeated rede ni- The South's share of the national total (%) tions of the strategic directions. By the mid- Population 37.2 36.0 35.1 36.1 36.6 1950s, the Cassa shifted its focus to sup- porting industry investment, concentrating GDP per capita 54.5 56.6 58.6 58.2 56.7 on "nuclei" and priority areas. Fixed investment 26.0 29.0 31.2 29.0 26.9 The result was that through the 1950s about 2 million workers left the target Unemployment regions. By the end of the 1960s, the South 9.1 6.4 9.6 16.3 19.7 emigration was perceived to be the main Center-North 6.8 4.5 5.2 7.6 6.5 development problem, and infrastructure investment and subsidies were concen- Source: Faini, Giannini, and Galli 1993. professional, and technical services. Lower What to do: transport policies in communication costs might be expected the developing world to lead to a footloose tradable service sec- What do these events and insights mean tor. But most communication-intensive for developing countries? Trade costs have industries remain strongly agglomerat- fallen because of lower transport and com- ed.51 A main reason for the persistence of munication costs, higher quality, and faster agglomeration economies in knowledge speed. But all countries have not benefited production is that verifying the quality of equally. Transport costs have fallen faster information requires understanding and where the demand for transport services is relationships of trust. Informal networks greater. Increasing scale in traded produc- work as screening devices to build up trust tion has raised competitiveness and allowed in a group of knowledge producers.52 And scale economies in the transport sector. lower urban transport costs increase the The resulting lower trade and transport size of networks. costs encourage trade and allow greater Transport Costs and Specialization 185 BOX 6.5 Mobility with density in Hong Kong, China Hong Kong, China, in the second half of 16-kilometer tram in the North Side of The experience in Hong Kong, China, the 1970s, had real growth of about 10 Hong Kong Island. Five private bus com- provides these lessons for the rapidly percent a year, an in ux of immigrants, panies operate franchised services with growing cities in the developing world: and roaring demand for private cars. Car more than 6,000 buses. They are comple- · There is a limit to mobility and acces- registrations more than doubled in a mented by minibuses (public light buses) sibility by private car in megacities. Even decade. The results were huge time losses with xed fares and exclusive rights to without congestion charging and the for passengers and freight transporters provide service on certain routes. Entry to pricing of parking, strong scal disincen- and health costs of air pollution. The this submarket is strictly regulated, with tives can contain motorization in a phase Transport Department in Hong Kong, a maximum number of minibuses set for of rapid income growth and limit the China, reacted with draconian measures. the city's quarters.b Switching between share of private cars in urban transport. In 1979, it de ned a transport policy di erent modes or submodes does not · Buses, and particularly minibuses, can be to increase road capacity, expand and lead to big time losses. regulated to avoid congestion and high improve the mass transit system, and bet- Road pricing failed politically in 1985. travel costs. Even with regulation, almost ter manage the road capacity. One reason was the opening of the all public transport can be pro table. The government trebled the annual Island Route of the Mass Transit Railway, license fee for cars, doubled the rst which carried about a quarter of all · Along with policies to contain motor- registration fee (to 70 to 90 percent public transport boardings in 1988, and ization, new tra c management instru- of the import price of a vehicle), and the Island Eastern Corridor a year ear- ments can make more e ective use of doubled fuel taxes. Private and public lier. Both eased congestion. Today, road existing infrastructure. vehicle ownership fell quickly. In 1985 the charges in Hong Kong, China, are seen Chapter 7 discusses, for countries at share of private cars in registered cars as a device not to reduce congestion various stages of urbanization, the insti- had fallen to 50 percent, 10 percent of but to curtail air pollution and main- tutions, infrastructure, and incentives them taxis.a The public transport system tain the city's attractiveness. It ranks that can facilitate concentration without consists of a 74-kilometer underground fifth in the infrastructure index of the congestion. mass-transit railway, a 34-kilometer heavy global competitiveness report, with a rail line (linking Kowloon with China), score of 6.2 out of 7, and first in product Source: WDR 2009 team. a 32-kilometer light-rail system in the market efficiency and financial market a. Hau 1990. northwest of the New Territories, and a sophistication. b. Cullinane 2002. specialization and exchange. Some coun- has contributed to an exaggerated focus tries such as China and Chile have broken of transport policy on physical improve- into international markets and benefited ments. And by using techniques that essen- from lower transport costs. But most of tially assumed away the internal workings the others have not. In much of Africa, this of transport--the goods to be transported cumulative causation has hurt, not helped, are seen as an iceberg to be hauled from one because agglomeration economies in Afri- place to another, and transport costs are ca's divided neighborhoods remain small. the part of the iceberg that melts away-- By increasing local market interactions the most critical policy-related aspects also and reducing intercity and interarea dis- have been assumed away. The fundamental tances and international divisions, transport features that deserve the attention of policy policies in developing countries can get these makers are the scale economies in the trans- virtuous circles started. Improving physical port sector that tend to create monopolistic infrastructure is an indispensable part of behavior, and the circular causation between transport policy. Indeed, chapters 7, 8, and lower transport costs and greater trade and 9 discuss the need for spatially connective traffic. Another underemphasized aspect is infrastructure in the local, national, and the external cost of transport and commu- international contexts. But other important nications, notably the congestion, pollution, aspects of transport and communications and safety-related hazards. policies are often neglected. The two neglected policy priorities are The new economic geography has high- (1)reducing the negative effects of market lighted what transport costs do for eco- structure in the transport sector and (2) nomic growth. Inadvertently, though, it improving trade facilitation and regional 186 WORLD DEVELOPMENT REPORT 2009 coordination. Both will promote agglom- places--it increases the value of all other erative forces and will sometimes provide related connections by enhancing overall greater payoffs than more physical infra- connectivity. These effects can be large. One structure investments. A third policy pri- estimate of the infrastructure-productivity ority is to address the negative externalities link for India found a sizable externality of in transport. transport infrastructure. By providing a 5 percent rate of return on road infrastructure Regulating transport to get the benefits investment over and above the direct payoff, of scale economies the network-related benefits accounted for Markets for transport services rarely are per- almost a quarter of the overall increase in fectly competitive, with major differences infrastructure productivity.57 between the different modes. Competition The absence of effective regulation limits in the trucking industry increased because of competition in the transport sector and can deregulation,53 but there is a tendency toward reduce the construction of new infrastruc- consolidated ownership in many countries. ture. It may cause underinvestment in main- In railways and airlines, markets remain tenance of existing infrastructure. A number dominated by state-owned enterprises.54 In of studies have confirmed the tendency to the airline and maritime transport indus- underprovide transport and telecommuni- tries, market segmentation allows providers cation infrastructure in developing coun- to discriminate between different goods.55 tries.58 Underinvestment in infrastructure These observations suggest firm-level size maintenance can be even more severe. advantages in transport operations. Actual expenditures for road maintenance Transport providers consolidate power in Africa, for example, appear to have sys- by owning infrastructure. In 1980 the top tematically fallen short of planned figures.59 20 percent of the world's carriers controlled It was estimated that $45 billion was lost in just 26 percent of the global port slot capac- road stock value during the 1970s and 1980s, ity. By 1992 this had increased to 42 per- which could have been avoided by spending cent, and by 2003, to 58 percent. It may be $12 billion for preventive road maintenance. even higher today. Badly maintained roads increase transport Infrastructure services are not provided costs by increasing costs of maintaining in competitive markets because the indivis- vehicles and reducing their speed. The direct ibility of infrastructure facilities naturally costs of badly maintained roads are thus precludes competition. At early stages of higher than the losses in cost-based road development, the demand for ports, roads, asset values as recorded by the road admin- and telecommunication equipment does istrations. On top of this, higher transport not exhaust minimum capacities. As traffic costs slow the spatial transformation and increases, so does productivity. This is ulti- reduce gains from specialization. mately balanced by increasing time losses The monopolistic sector also encour- caused by congestion--as diseconomies ages corruption. In smaller markets, users of scale set in. Recent developments have often have no substitutes for the services of made the advantages of large ports and large ports and airports. The higher these airports even more pronounced--and the substitution costs, the higher the potential technological progress in shipping has rein- for high markups or bribes, depending on forced the cost advantages of large ports whether the infrastructure is private or (see box 6.3). Assessing the size of these public. How much rent-seeking increases scale effects is a daunting task, but stud- transport costs is difficult to estimate. But ies have confirmed economies of scale and a recent World Bank study that reviewed spotlighted the indivisibility of transport the main road corridors in all the regions infrastructure.56 of Sub-Saharan Africa reveals big gaps A second reason for limited competition between prices for transport services and arises from "network economies." Adding their costs (see table 6.1). The surplus is a link to a road or rail network does not shared among bribes, regulatory rents, and just provide the benefits of connecting two transport company profits. Transport Costs and Specialization 187 Table 6.1 Prices, costs, and profit margins are all high on Africa's transport corridors Route Pricea Variable cost Fixed cost Profit marginb Corridor (countries) (gateway­destination) (US$ per kilometer) (US$ per kilometer) (US$ per kilometer) (percent) West Africa Tema/Accra­Ouagadougou 3.53 1.54 0.66 80 (Burkina Faso, Mali, and (2.01) (0.59) (0.64) Ghana) Tema/Accra­Bamako 3.93 1.67 0.62 80 (1.53) (0.23) (0.36) Central Africa Douala­N'Djaména 3.19 1.31 0.57 73 (Cameroon, Central African (1.10) (0.32) (0.30) Republic, and Chad) Douala­Bangui 3.78 1.21 1.08 83 (1.30) (0.35) (0.81) Ngaoundéré­N'Djaména 5.37 1.83 0.73 118 (1.44) (0.25) (0.44) Ngaoundéré­Moundou 9.71 2.49 1.55 163 (2.58) (0.64) (0.43) East Africa Mombasa­Kampala 2.22 0.98 0.35 86 (Kenya and Uganda) (1.08) (0.47) (0.14) Mombasa­Nairobi 2.26 0.83 0.53 66 (1.36) (0.17) (0.19) Southern Africa Lusaka­Johannesburg 2.32 1.54 0.34 18 (South Africa, Zambia, and (1.59) (0.41) (0.40) Tanzania) Lusaka­Dar-es-Salaam 2.55 1.34 0.44 62 (0.08) (0.52) (0.51) Source: World Bank 2008d. a. Some indicative prices are set by some ministries of transportation in Africa but are not used. Prices set by freight allocation bureaus in Central Africa may be more respected. b. Data should be interpreted cautiously since some companies may omit some costs or, on the contrary, double count some costs. Transport infrastructure and service Map 6.2 Crossing borders or climbing walls? providers are not the only ones extracting Checkpoints on priority transport corridors bribes and enjoying extranormal profits. Ouahigouya N I G E R The Improved Road Transport Gover- Ségou Niamey B U R K I N A nance Initiative in West Africa monitors Bamako F A S O road practices harmful to trade on inter- Ouagadougou M A L I state trunk roads between Burkina Faso, Ghana, Mali, and Togo (see map 6.2). In Sikasso Mali, truckers face 4.6 checkpoints, pay Bobo $25, and waste 38 minutes for every 100 Dioulasso kilometers traveled. Internal or distance- B E N I N related costs are compounded by costs G H A N A imposed by the divisions of international GUINEA Korhogo borders. N I G E R I A C Ô T E D ' I V O I R E National efforts and regional Checkpoints TOGO Bouaké coordination to facilitate trade Police Ibadan Customs Trade facilitation has become the most Gendarmerie Other: Transport department, Kumasi important policy instrument to achieve trade unions, forestry department, Lagos Porto- gains from international trade--improv- health Lomé Novo Tema ing the efficiency of ports, harmonizing Abidjan Subject road network Accra standards, reducing bureaucratic burdens Sekondi- to cross borders, and coordinating behind Takoradi the border regulatory norms (see box 6.6). Source: Improved Road Transport Governance Project, USAID West African Trade Hub 2007. Since August 2004, trade facilitation has moved to the center of the Doha Round 188 WORLD DEVELOPMENT REPORT 2009 BOX 6.6 Neighborhoods matter, but so do trade and transport policies Proximity to prosperous places can be neighborhoods of Central America, Algeria and Indonesia have the same a blessing, and to poor places, a curse. North Africa, and Southeast Asia, Mexico, location as Tunisia and Malaysia, but they The box map illustrates the advantage Tunisia, and Malaysia have the highest do not do as well in business and trade of being in good neighborhoods. It market access. Their rankings in the policies. Their market access indicators shows the foreign market potential World Bank's Doing Business indicators-- are accordingly lower than those of their across the world, using an index that especially those related to trading across neighbors. Sri Lanka and Ghana also do combines geographic proximity (dis- borders--are among the highest in their not do well in market access; they have tance) and policies to reduce trade barri- regions. Unsurprisingly, their recent good business and trade policies, but are ers (divisions). growth performance has been impres- not fortunate in location. But good location is not enough. Even sive, and their living standards have within the geographically fortunate improved. Sources: Mayer 2008; World Bank 2007d. Being near prosperous places is important, but not enough Foreign market potential, 2003 Foreign market potential relative to U.S. Level < 0.760 0.760­1.110 1.111­1.670 1.671­4.570 > 4.571 No data Sources: Mayer 2008 for this Report. Note: To compute foreign market potential, each country is assigned a score for the size of international markets with which it can trade. This is computed by weighting the GDP of other countries by the inverse of a measure that combines physical distance, transport costs, and barriers to trade to show how difficult it is to access these markets. The measure, which is expressed relative to the foreign market potential of the United States, essentially combines the two spatial dimensions of distance and division into a composite of potential market access that does not include the effect of the home market (density). This map is a complement to map 9.2 showing Real Market Access. of the World Trade Organization. Recent or bribes connected with import and World Bank studies have identified several export permits measures of trade facilitation as the main · Information technology (IT) infrastruc- entry points for policy reform: ture--a measure of the speed and the cost of Internet access and the contribu- · Port efficiency--an average of the effi- tion of the Internet to the reduction of ciency of port, inland waterway, and air inventory costs transport facilities, based on data from the World Competitiveness Report Improving trade facilitation capacity in · Customs regimes--the hidden import 75 countries to half the global average could, barriers other than published tariffs as one study suggests, yield a $377 billion and quotas, and irregular side payments increase in world trade.60 Another study Transport Costs and Specialization 189 showed that new and old EU member coun- needed elsewhere. This is also true for other tries would benefit from trade facilitation hub infrastructure such as airports, which measures in what were then the accession are increasingly important for trade in low- member countries, Bulgaria, Romania, and weight, high-value goods and to support Turkey.61 If these countries reached half the booming export-oriented services that need trade facilitation standards of the EU-15 in efficient air travel. In 2007 passenger traffic terms of port efficiency, IT infrastructure, at Bangalore's airport jumped 35 percent. customs regimes, and harmonized regula- With the fall of effective rates in inter- tion, it would lead to $10 billion in overall national freight transport, time costs in trade gains. Among the four dimensions international transport have become more of trade facilitiation, improvement of IT important relative to the direct money infrastructure would result in the highest costs.63 International transport suffers from trade gains (40 percent) followed by port the extra time cost of border-crossing pro- efficiency improvements (30 percent). cedures. These time costs depend not just Port efficiency improvements require on the customs and fiscal rules of crossing both institutional and infrastructure the border but also on a host of behind-the- investments. Maritime transport accounts border elements concerning regulation and for 90 percent of world trade by volume.62 the supply of services.64 Access to a well-run port may not guar- Among the poorest performers: the time antee export-oriented agglomerations, costs of transport range from 46 days in the but these agglomerations certainly will Democratic Republic of Congo to 104 days in not emerge without it. Private participa- Uzbekistan, set against the Organisation for tion will be feasible where trade volumes Economic Co-operation and Development are sufficiently high, but public support is (OECD) average of 9.8 days (see table 6.2). Table 6.2 Time costs for crossing borders are highest in Central Asia, Central Africa, East Africa, and Southern Africa Country Documents for exports Days Country Documents for imports Days Iraq 10 102 Uzbekistan 11 104 Kazakhstan 12 89 Chad 9 102 Tajikistan 10 82 Iraq 10 101 Uzbekistan 7 80 Tajikistan 11 83 Chad 6 78 Kazakhstan 14 76 Afghanistan 12 67 Kyrgyz Republic 13 75 Angola 12 64 Afghanistan 11 71 Kyrgyz Republic 13 64 Burundi 10 71 Eritrea 9 59 Eritrea 13 69 Niger 8 59 Rwanda 9 69 Mongolia 10 58 Niger 10 68 Central African Republic 8 57 Zimbabwe 13 67 Azerbaijan 9 56 Central African Republic 18 66 Zambia 8 53 Congo, Dem. Rep. of 9 66 Haiti 8 52 Venezuela, R.B. de 9 65 Zimbabwe 9 52 Mali 11 65 Congo, Rep. of 11 50 Zambia 11 64 Lao PDR 9 50 Congo, Rep. of 12 62 Burundi 9 47 Mongolia 10 59 Rwanda 9 47 Angola 9 58 Congo, Dem. Rep. of 8 46 Azerbaijan 14 56 Sources: World Bank 2007d. 190 WORLD DEVELOPMENT REPORT 2009 Most of the slowest border crossings occur the case in many rapidly growing develop- in Sub-Saharan African or Central Asian ing countries. Estimating the costs of con- countries, many of which are landlocked. gestion is not straightforward, because it Having little control over other aspects of occurs mostly during certain times of the trade costs, such as transport over land to day, often caused by specific bottlenecks the nearest port, landlocked countries could in the network. One study of Washington, be more aggressive with the trade facilitation D.C., congestion put these costs at $0.065 policies that they do have control to improve. per mile.68 They could also benefit from a more explicit Emissions. With growing concerns about regional perspective. A variety of transit rules climate change, the transport sector--a vis- are recognized by international law and dec- ible consumer of fossil fuels--has been get- larations, such as the "Almaty Programme of ting more scrutiny. The largest share of these Action."65 Corridor facilitation and monitor- emissionsisgeneratedinindustrializedcoun- ing initiatives, such as those envisaged under tries. But with rising motorization in many the Sub-Saharan Africa Transport Program, developing countries, the world's vehicle fleet could reduce the risk of coordination failures, will rapidly grow, and so will emissions. Most but enforcement has been weak.66 estimates of greenhouse gas emissions from Even for fairly small coastal countries, transport are close to 13.5 percent of the total regional approaches can be beneficial. Since (see figure 6.6). One integrated assessment increases in trade produce scale economies study puts the population-weighted expected in transport, hub infrastructure is most global costs of a 2.5° C warming in 2100 at 2 beneficial if it is shared by as many market percent of world GDP.69 Half of this is caused participants as possible. Few countries in by abrupt climate change, including the pos- West or East Africa, if any, can support a sible spread of tropical disease, especially in medium-size, deep-water container port on Africa. Other costs are incurred in agricul- their own. But a shared port with a large ture (less than 10 percent) and from rising catchment area would be more likely to seas (6 percent). support agglomeration, if costs and access What would internalizing these costs are distributed among coastal countries mean for the overall costs of transport? and their landlocked neighbors. Sharing is Estimates vary. A meta-analysis of earlier not easy, however, because of the domestic estimates suggests a current upper bound bias of national infrastructure policies. of $50 per ton of carbon.70 The Stern Review (2007) puts the total damage from future Addressing the negative externalities warming at 5­20 percent of world GDP in of transport perpetuity and infers a current social cost Efficient transport provides external ben- equivalent of $311 per ton of carbon. With efits that go beyond simple time savings or a gallon of gasoline containing 0.0024 tons lower maintenance; these benefits are often of carbon, damage of $50 per ton of car- underappreciated. But transport also has bon would translate into $0.12 per gallon external costs that usually are not internal- of gasoline (or $0.03 per liter) and damage ized by transporters and traders.67 Conges- of $300 into $0.72 a gallon ($0.19 a liter). tion and greenhouse gas emissions affect Internalizing the carbon dioxide (CO2) both developed and developing countries, costs of transport would thus increase but the direct health-related costs of pollu- transport costs by an amount well within tion and poor safety are generally highest in historical gasoline price variations. Efforts developing countries. to increase fuel efficiency have been under Congestion. The lumpiness of trans- way for the past three decades, aided at least port infrastructure implies that there is no as much by fuel taxes and efficiency regula- smooth and immediate supply response tions as by the rising price of oil. when demand increases. With overcapacity, Pollution. Gasoline vehicles emit carbon the extra cost could be spread over a larger monoxide (CO), nitrogen oxides (NOx), number of users. With insufficient capacity, and hydrocarbons (HC). CO reduces oxy- congestion causes time and quality losses, gen in the bloodstream, causing breathing Transport Costs and Specialization 191 Figure 6.6 Transport accounts for about one-seventh of CO2 emissions Sources of greenhouse gas emissions, 2005 Sector End use/activity Gas Road 9.9% Transportation 13.5% Air 1.6% Rail, ship, & other transport 2.3% Residential buildings 9.9% Electricity & heat 24.6% Commercial buildings 5.4% Unallocated fuel combustion 3.5% Iron & steel 3.2% Aluminum/non-ferrous metals 1.4% Machinery 1.0% Carbon dioxide Pulp, paper & printing 1.0% Food & tobacco 1.0% (CO2) 77% ENERGY Other fuel 9.0% Chemicals 4.8% combustion Cement 3.8% Industry 10.4% Other industry 5.0% T&D losses 1.9% Coal mining 1.4% Fugitive emissions 3.9% Oil/gas extraction, refining 6.3% & processing Industrial processes 3.4% Deforestation 18.3% Afforestation -1.5% Land use change 18.2% Reforestation -0.5% Harvest/management 2.5% HFCs, PFCs, Other -0.6% SF6 1% Agricultural energy use 1.4% Methane Agriculture soils 6.0% (CH4) 14% Agriculture 13.5% Livestock & manure 5.1% Rice cultivation 1.5% Other agriculture 0.9% Nitrous oxide Landfills 2.0% Waste 3.6% (N2O) 8% Wastewater, other waste 1.6% Source: World Resources Institute; see Baumert, Herzog, and Pershing 2005. difficulty and cardiovascular damage. HC and injuries. But in developing and transi- and NOx combine to form ozone, making tion countries, these rates are increasing. breathing harder and reducing visibility. The rate of road fatalities in the Russian NOx and HC also react to form fine par- Federation, for example, is five times that ticulate matter (PM2.5), small enough to of the Netherlands. Some 1.2 million peo- enter lung tissue and increase mortality ple die in road accidents each year, and 90 risks. Vehicle emissions of all local pol- percent occur in low- and middle-income lutants have fallen in developed countries, countries. World Bank projections sug- but they remain high elsewhere. Diseases gest an increase by more than 80 percent related to air pollution contribute to the between 2000 and 2020 in these countries, premature death of more than half a mil- but a decrease of 30 percent in high-income lion people each year, imposing a cost of countries. For every death, there are many up to 2 percent of GDP in many developing cases of injury and disablement. Projected countries. Transport may be responsible for health losses from traffic accidents as a about a quarter of this impact, mainly from share of the total health losses are high- private and commercial vehicles.71 est in the Middle East and North Africa Accidents. Similar to local air pollution, (5 percent)--expected to rise to 8 per- developed countries with high but stable cent--followed by Latin America and the motorization have reduced road fatalities Caribbean and East Asia and the Pacific (3 192 WORLD DEVELOPMENT REPORT 2009 BOX 6.7 Unclogging Latin America's arteries: transport costs now matter more than tariffs For the last two decades, the trade policy A reduction of transport costs would reduction in tari s, the bene ts of a fall in agenda of Latin America has been domi- bring about substantial bene ts. A transport costs for intraregional exports nated by traditional market access and 10-percent decrease in trade costs would are almost ve times larger and lead to policy barriers issues. It has paid o . Tari s increase the region's imports by 50 per- an increase in the number of products have come down a lot. Most-favored-na- cent and intraregional exports by more exported to the region, which is nine times tion tari s fell from more than 40 percent than 60 percent. The bene ts of better bigger than a similar reduction in tari s. in the mid-1980s to close to 10 percent by transport policies seem to be much larger Source: Inter-American Development Bank, 2000. Still, trade agreements continue to than lower tari s. Compared with a similar forthcoming. dominate policy discussions in the region. But transport costs are now more important than tari s. Simple averages of import ad valorem freight range from Transport costs now matter more for trade Percentage change in transport costs by making port efficiency, tariff rates, and number of 6.5 percent in Argentina to 12 percent in shippers the same as U.S. levels, base year 2005 Colombia for intraregional freight, and from 7.5 percent in Uruguay to 25 percent Percent in landlocked Paraguay. Freight costs 50 in Latin America and the Caribbean for Port efficiency exports to the United States are--with 40 Tariff the exception of Bolivia, Mexico, and Number of shippers República Bolivariana de Venezuela-- 30 even higher than intraregional freight. Low port e ciency and weak compe- 20 tition in the maritime transport sector 10 seem to be the culprits (see gure to the right). On average, transport costs in Latin 0 America would decline by 20 percent if Peru Ecuador Latin America Uruguay Chile Brazil countries in the region had U.S. levels of & Caribbean port e ciency. Source: Inter-American Development Bank, forthcoming. percent), then Africa, Eastern Europe and What is needed for transport to continue to Central Asia, and South Asia (2 percent). contribute to development? Fatality risks are highest where motor- Poor countries become big producers ization rates are rapidly increasing, because before they become big consumers. Income of long lags in implementing road safety generation by importing intermediate goods measures. The transport sector can thus and raw materials and exporting processed impose a cost on development. To what goods will be important. The relocation of extent accident costs are "external" depends intermediate production processes to mid- on how well insurance markets can cover dle- and low-income countries indicates the external costs, but even where those the enormous potential benefits from inte- markets do not work well, road safety mea- gration into world markets even for these sures can protect pedestrians, cyclists, and countries, limited mainly by transport other drivers from the reckless. and communications costs. But achieving this raises difficult institutional questions. Transport: an increasingly The provision of access to foreign markets important sector implies that some of the benefits of trans- For most modes of transport, costs have port policies will accrue to foreign coun- declined in many markets. Still, those costs tries. Coordinating international transport are becoming a larger share of overall trade policies thus requires a growing confidence costs because of steeper declines in tariffs in reciprocal support for international in regions such as Latin America and the transport. Caribbean (see box 6.7). And with fuel costs The increasing returns to scale in trans- rising, transport's share will rise even more. port add two more coordination problems. Transport Costs and Specialization 193 The scale of least-cost port and airport their transport costs to North America investments calls for hub-and-spoke trans- andWestern Europe have fallen.They have port systems in which neighboring coun- joined the growing trade in intermediate tries share facilities. Because ownership of and final manufactured goods. Countries large infrastructure facilities provides mar- such as Tunisia can do the same. ket power, the sharing of facilities requires ·In South Asia, falling trade and commu- credible agreements. Increasing returns in nication costs have helped India enter transport operations--with maritime ship- western markets for intermediate ser- ping supplied by a small number of firms vices, eliminating some of the disadvan- and logistics services being consolidated tages of being distant. Countries such as in fewer hands--may require regulatory South Africa can do the same, exploiting regimes to realize the potential for lower their home market potential. transport costs. The mutual dependence ·In Central Asia--with economies that of transport policy and competition policy are small, landlocked, and dependent on implies a global effort, such as that started exports of primary products such as oil by some multilateral organizations. and gas--reducing transport costs will Transport and communication costs will be more difficult. It will also be difficult remain a principal influence on the speed for smaller countries in divided neigh- and efficiency of the spatial transformations borhoods, such as Burkina Faso, Malawi, needed for growth. Countries at different Niger, and Rwanda. These countries will stages of transformation will have to formu- need aggressive measures to lessen the late different policies for reducing transport trade friction, enforceable agreements costs. East, South, and Central Asia illustrate with neighbors to share expensive infra- the contrasts: structure, and selected investments to · Developing countries in East Asia are now encourage agglomeration and reduce closer to world markets, as Japan and the the transport costs for primary goods Republic of Korea have prospered, and exports. Geography in motion Distance and Division in East Asia When Admiral Zheng He brought a giraffe to Nanjing in 1415, it was believed to be a heavenly beast, associated with great peace and prosperity. It also marked the heyday of Chinese influence in East Asia and the region's wealth relative to the rest of the world. China at the time was probably the world's largest economy, enjoying the highest standard of material, living with flourishing art and education and advances in a range of technologies. Its naval skills had enabled voyages to places as far away as Africa. C hina alone may have accounted the Qing and Tokugawa periods. Others Shortly thereafter, most of Europe went for one-third of global manu- believe that it is more apt to character- through the Industrial Revolution, facturing. This was not to last. ize these societies as having stagnant and the "great divide" between Europe A hundred years later, a new emperor rather than declining economies. In any and Asia widened, with widespread destroyed Zheng He's navigation logs event, wage levels in Japan and China at advances in European wages and gross and slashed the navy to one-tenth its the start of the nineteenth century were domestic product (GDP). According size, believing that the costs of for- well below those in London or Amster- to Angus Maddison (2006), East Asia's eign expeditions outweighed the ben- dam, even in real terms, perhaps by as share of global GDP, constant at around efits. China entered centuries of self- much as 50 percent.1 Adam Smith had 40 percent between 1500 and 1800, fell imposed isolation, broken in infamous already recognized this:"The difference to less than 15 percent by 1950. and damaging fashion by the British between the money price of labour in during the Opium Wars of the nine- China and Europe is still greater than Fifty years of Asian integration teenth century. that between the money price of sub- Fast forward to today. East Asian econo- sistence; because the real recompence mies have become integrated through East Asia's age of isolationism of labour is higher in Europe than in a dense array of regional production China was not alone in trying to shut China."2 networks. These supply chains started out the rest of the world. In Japan, Smith was correct. Even before the with outsourcing by Japanese multina- Tokugawa Iemitsu issued the "Closed Industrial Revolution, parts of Europe tionals in the 1980s, as wages and land Country Edict of 1635" and the"Exclu- had advanced beyond Asia in their liv- costs in the dense production area of sion of the Portuguese, 1639,"effectively ing standards. He was also right in writ- Tokyo grew prohibitive for competitive shutting off Japan to external influences ing about China as a unified economy. manufacturing. In fact, economic con- for the next two centuries. The edicts The mandarins of China kept excep- gestion in Hong Kong, China; Japan; the not only prevented foreign entry into tional records of wages paid to armorers Republic of Korea; and Taiwan, China, Japan but also banned Japanese from and other craftsmen providing services has resulted in spillovers--first to mid- going abroad. The dislike of all things to the government. These show little dle-income countries in Southeast Asia, Western extended to technology. In regional difference despite the vast dis- and then to China,as the barriers of eco- an extraordinary attempt to preserve tances within imperial China. Only the nomic ideology were reduced. Recently, its culture and social hierarchy, Japan less dense, sparsely populated northern supply chains have centered on China gradually abolished the gun in favor of areas had somewhat higher wages. and the great assembly operations in the more elegant and symbolic samurai By the middle of the nineteenth cen- Guangdong and Shenzhen.As China has sword. tury, real wages in Canton and Tokyo, matured, it too has become an exporter These extreme examples show the the most advanced cities in Asia, were of intermediates and capital equipment. vast division between countries in East only as high as in small European cit- China is now the main trading partner Asia, especially after the seventeenth ies like Milan and Leipzig. Elsewhere in for Japan and the Republic of Korea and century. Scholars do not agree fully on East Asia living standards were lower sources more than half its rapidly grow- the economic effects of such division. still. The old Chinese tributary states ing imports from East Asia. Some have argued that reductions in had been colonized,andAsian countries Intraregional trade in East Asia living standards were significant during were further divided (see map G3.1). today approximates that within the Geography in Motion 195 Map G3.1 Asia divided: conflict in the middle of the nineteenth century intensification and rural industrial- ization, followed by urban expansion Chinese territory lost and manufactured exports. There has Chinese tributaries lost been learning from abroad--of new technologies and of new institutions. Taiping rebels, 1853­64 Sakhalin Exports have become more technolog- RUSSIA ically complex. Middle-income coun- MARITIME PROV. Hokkaido tries have specialized in component TO RUSSIA production, while rich Asian countries M A N C H U R I A TO JAPAN have added more value through inno- Vladivostok vation, branding, and greater techno- Honshu logical sophistication. MONGOLIA LIAOTUNG PENINSULA As the region has grown,it has devel- (Russian sphere: (Russia) Port Arthur JAPAN independant 1911) (Lüshun)(Russia) KOREA oped a dynamic that reinforces growth. S I N K I A N G Peking Dairen (Beijing) (Dalian) ASEAN, China, Japan, and the Republic Shikoku Tientsin Weihaiwei of Korea are an economic mass compa- (Tianjin) (Weihai) (United Kingdom) Kyushu Muslim rebellion rable to North America in the 1990s. Nien rebellion Kiaochow Bay 1862-77 1853-68 (Germany) As the center of gravity of the global CHINA economy shifts toward the Pacific Rim, Nanking Shanghai T I B E T Ningpo global market access for everyone in Okinawa Hankow East Asia has improved. SIKKIM Chungking Foochow (Chongqing) NEPAL (Fuzhou) The degree of intraregional trade in BHUTAN East Asia may be considered surprising Taiwan Canton given the history of divisive political (Guangzhou) Kowlong & relationships between many East Asian Hong Kong (United Kingdom) countries. In the Western Hemisphere, Kwang-Chow TO BRITAIN Luzon BURMA Bay (France) the economic effects of conflict between Hainan countries were overcome by formal INDO- CHINA Samar institutions of codified legal systems Mindoro SIAM PHILIPPINES and political agreements that governed TO FRANCE arm's-length commercial transactions Palawan Andman and that could be readily expanded to Islands Mindanao accommodate rapid growth in com- merce and finance. In East Asia these Source: www.fordham.edu/halsall. institutions have been slower to develop. Instead, a long history of social net- works, communities, and informal European Union (EU), growing con- in soft infrastructure, such as customs institutions--with roots in the migra- sistently faster than that between East reforms and visa exemptions within the tions over millennia of people from Asia and other regions in the world. Association of Southeast Asian Nations Southeastern China to Southeast Asia-- East Asian countries are the source for (ASEAN). provides the trust to support modern almost two-thirds of all foreign invest- The falling division between coun- international integration of goods and ment in the region. Even technology is tries in East Asia has coincided with money (see map G3.2). starting to originate within the region, rapid growth across a diverse spectrum The integration ahead-- especially in key export industries such ranging from Lao People's Democratic as electronics. East Asian countries are Republic with a per capita income the twin challenges of busily driving down divisions between of $500 in 2006 to Singapore with a distance and division each other in the form of trade barri- per capita income of almost $30,000. Peering into the future, the region ers and other border costs. They started Within East Asia, incomes are slowly faces challenges on its path of rapid with world-class logistics in ports and converging: poor countries are grow- integration. airports--albeit sometimes restricted ing faster than rich countries. Most The shift in economic density toward to special economic zones. And they East Asian countries have followed the north poses a special challenge for have continued with improvements similar paths, starting with agricultural Southeast Asia. How can it remain 196 WORLD DEVELOPMENT REPORT 2009 Geography in motion Map G3.2 Asia integrated: trade at the end of the twentieth century markets are developing where people are concentrated. Urbanization is large Total imports (US$ billions) and rapid in most countries, perhaps >250 adding 25 million city dwellers every 100­250 10­75 year for the next two decades. Most of 75­100 REP OF REREP. OF <10 these people will move to small and KOREA KOREA JAJAPANAN JAP medium-size cities of less than 1 mil- CHINACHINA lion people, not to major metropolitan areas. Managing these small cities effi- ciently and integrating them into the national economies will be a crucial task for reducing distance and sustain- TAIWAN,TAIWAN, HONGHONG CHINA CHINA ing growth. KONG,KONG, CHINACHINA Meanwhile, East Asia still faces strategic questions about how to bring down divisions between countries in THAILAND THAILAND VIETNAMVIETNAM the region. ASEAN's two-speed pro- PHILIPPINES PHILIPPINES cess shows how hard it is for countries with different incomes and economic MALMALAYSIASIA MALA structures to integrate deeply. No for- mal process of economic integration brings together all the economies of SINGAPORESINGAPORE the region. A first attempt to start a INDONESIAINDONESIA regional dialogue was at the East Asia Bilateral import flows (US$ billions) >150 Summit in Kuala Lumpur, Malaysia, in 100­150 25­50 10­25 December 2005. The summit called for 50­100 financial stability, energy security, pov- erty eradication, and narrowing gaps Source: Gill and Kharas 2007. between countries. It underscored the challenges that still divide the region: significant as an economic force in the EastAsia,with high poverty rates in areas cross-border migration, environmen- region? One answer perhaps lies in the like western China, southern and eastern tal spillovers, diversity of governance development of world-class cities. The Philippines, Thailand's northeast, and standards, and cultural understand- major metropolises of Southeast Asia Vietnam's central highlands. The gap ing. Other interesting experiments to need to develop themselves into "sticky between per capita incomes in the richest foster regional integration are under places," attracting and retaining global and poorest provinces of China--negli- way, such as within ASEAN+3, but talent. Meanwhile, the integration of gible under the imperial dynasties of the the institutional leadership to forge a Australia and India into the region past--has swelled to 13.1:1 (compared common future is fragmented. Even might alter the dynamics of place,offset- with 2.1:1 in the United States).Although so, leading scholars have noted that ting to some degree the northward drift many have moved closer to prosperous "the emergence of an integrated East of Asia's economic center of gravity. areas, overcoming the geographic dis- Asia is inevitable and necessary."3 The The problems encountered by coun- tances that isolate these populations is challenge is figuring out how to make triesdistantfromthemajormarketsof the still seen as a major challenge. this happen quickly. region are echoed in lagging areas within Within East Asian countries, peo- countries. Significant poverty remains in ple are moving to the markets, and Contributed by Homi Kharas. Part Three reframing the policy debates Can crowded cities in developing countries pull people in and power them out of poverty? Does migration help those who move and those left behind? How can trade help the world's wealthy and most destitute? What can policy makers do to address the three big challenges facing the developing world--a billion slum dwellers, a billion people living in remote and underserved areas, and the"bottom billion"? Part three of the Report provides the answer: economic integration. How? By using spatially blind institutions, spatially connective infrastructure, and spatially targeted incentives and calibrating the response to the di culty of integration. Chapter 7 explains what economic integration means for metropolises, cities, towns, and villages. Chapter 8 proposes how integration between economically leading and lagging areas can bene t everyone. Chapter 9 lays out the di cult steps needed to successfully integrate the world's most isolated countries. In doing so, the chapters in part three revisit and reframe long-standing policy debates on urbanization, territorial development, and international integration. CHAPTER 7 Concentration without Congestion Policies for an inclusive urbanization A team of urban experts, as part of Projections now suggest that cities in a routine exercise in 1974, forecast developing countries will double in three the size of the world's most popu- decades, adding another 2 billion people. lous cities in 2000. Kinshasa, the Democratic Indeed, large cities in developing countries Republic of Congo's capital, would grow to 9 will grow bigger to provide the urbanization million, more than London today. Pakistan's economies sought by entrepreneurs, workers, Karachi would expand to 16 million, almost and innovators. But this will happen mostly as large as New York City. The forecasts were in economies that are doing well. Medium- way off (see figure 7.1). Kinshasa's popula- size cities remain the backbone of urban tion is about half of London's today, Kara- systems, providing the localization econo- chi's about half of New York City's. Why mies that producers with more specialized were the experts, generally good at forecast- needs seek. But they will flourish only where ing national populations, so wrong in pre- economies are industrializing. Smaller cities dicting city sizes? and towns continue to serve and to depend The reason: forecasting the spatial dis- on surrounding rural settlements. But they tribution of people in a country is not the will grow rapidly only in areas where farms same thing as predicting the size of its and village economies are doing well. population. As shown in earlier chapters, The spatial transformations that lead spatial transformations--the growth of to the rise of cities and towns will not be cities and leading areas--are linked closely orderly. Informal settlements--slums and to changes in the economy, especially the shantytowns--may form and expand as the sectoral transformations that accompany rising demands of workers and firms out- growth and the opening of an economy to strip the capacities of governments to insti- foreign trade and investment. So predict- tute well-functioning land markets and to ing the size of a city is economic forecasting, invest in infrastructure and accommodation. a hazardous occupation. Take Guangzhou If today's developed countries are a guide, it in China. Its population in 2000 was more takes many decades to address within-city than a third larger than the 4.5 million disparities and to absorb informal settle- predicted in 1974. Beijing's was half the 19 ments into more organized city structures. million predicted. The experts could not Trying to restrict rural-urban migration can have foreseen China's economic liberal- be counterproductive. Why? Because limiting ization and growth, which quickly would density and diversity stifles innovation and change the country's spatial structure. Sim- productivity. ply extrapolating past trends, they should Policy makers, if they are not careful, can instead have examined the market forces end up harming these transformations. By of agglomeration, migration, and special- not instituting flexible regulation and versa- ization--and the government policies that tile land use conversion, they can make urban 198 help or hinder them. areas inhospitable to firms and investors. By Concentration without Congestion 199 not providing adequate water, sanitation, Figure 7.1 The growth of cities has been grossly overestimated schools, and health care in rural areas, they can prematurely push villagers to towns and Dhaka Guangzhou cities. By not investing enough in the infra- Tokyo structure of rapidly growing cities, they can Hong Kong, China Hyderabad encourage congestion. More generally, by Istanbul not responding appropriately to the needs of Bangalore Moscow spatial transformations for different types of Delhi places, they cannot reap the full benefits of Lagos density and diversity, which congestion and Tianjin Chicago division can undermine. Osaka-Kobe To help nations benefit from urbaniza- Leningrad Buenos Aires tion, this chapter proposes a policy frame- Mumbai work--informed by the stylized facts of New York­Newark Ahmedabad spatial transformations (chapter 1) and the Los Angeles insights about agglomeration economies Paris that drive these changes (chapter 4). It out- Manila Bogotá lines the policy priorities and their sequenc- Shanghai ing, using the experiences of countries that Kolkata São Paulo have urbanized well and of those that have Jakarta struggled. The main messages: London Cairo · Rural-urban transformations are best Karachi Chennai facilitated when policy makers recognize Mexico City the economic interdependence among Bangkok settlements. Within a country's hierar- Lima-Callao Rio de Janeiro chy of cities, towns, and villages, each Kinshasa specializes in a different function and has Seoul Beijing strong interrelationships with others. So Tehran the policy discussion should be framed Baghdad not at the extremes of the national level ­60 ­40 ­20 0 20 40 60 80 100 120 or the individual settlement. Instead, it Prediction error relative to population in 2000, % should be framed at the level of what is Source: Satterthwaite 2007. termed an "area," usually a state or prov- Note: Comparison of predictions in 1974 with estimates of city populations in 2000. Bar indicates the extent to which the city population was overpredicted in 1974 relative to its size in 2000. A negative number indicates ince. Policy makers should see themselves that a city size was greater in 2000 than predicted. as managers of the portfolio of places in such an area. An area approach can also urban shares, for example, traffic con- inform national urbanization strategies. gestion and slums may not be the major While area-specific urban shares in the policy problems. But in rapidly urban- population will determine priorities for izing areas, congestion can quickly set all levels of government (central, pro- in. And metropolitan areas may need, vincial, and municipal), a nation's urban in addition, to address within-city divi- share can be a good guide to the overall sions posed by shantytowns and slums. complexity of its challenges. · Prioritizing and sequencing of policies · Policy challenges become more com- can help governments facilitate inclu- plex with urbanization. Cities and sive urbanization even in the early towns provide firms and families the stages of development. Each dimension benefits of proximity, but the compact- of the integration challenge requires ness of activity produces congestion, a different family of instruments. For pollution, and social tension, which can areas of incipient urbanization, the pol- offset those benefits. Whether a policy icy challenge is one-dimensional: build is desirable depends on if it addresses density with spatially blind institutions. the market's failures and abets concen- For areas of intermediate urbanization, tration. In countries or areas with low it is two-dimensional: build density and 200 WORLD DEVELOPMENT REPORT 2009 reduce economic distance with spatially Principles for managing a portfolio connective infrastructure. And for areas of places of advanced urbanization, it is three- Debates about urbanization often evoke dimensional: build density, overcome images of overcrowded cities, visible con- distance, and address the economic centrations of poverty, and appalling envi- and social divisions--caused, say, by ronmental degradation. This can result in a slums--with spatially targeted inter- generalpolicystancetocontrolurbangrowth ventions. But at every stage, policy mak- and curb rural-urban migration. Geographi- ers should emphasize the spatially blind cally targeted interventions to clear or clean institutions that encourage density in up slums that proliferated during low- and the right places. middle-income stages of development can The chapter first summarizes how end up dominating the discussion. urbanization policies can help places facili- This chapter reframes the urbanization tate agglomeration economies. It next out- debate. Historical evidence suggests that lines a framework for economic integration urbanization in developing countries will to guide the management of a portfolio of continue to be rapid at early stages of eco- places, using the experience of places that nomic growth--much of the rise in urban have urbanized successfully. It then dis- shares takes place before nations get to cusses examples of the framework in action upper-middle incomes. But the rising den- in today's developing countries. sity is to be welcomed if it produces agglom- eration economies. The debate should not be mainly about the pace of urbanization, the amount of rural-urban migration, or BOX 7.1 Are the policy messages of this Report antirural? the ways to eradicate slums with targeted No. interventions. Instead, it should be about the efficiency and inclusiveness of the pro- The economic geography of nations · Sound macroeconomic policies to cesses that transform a rural economy into does not conform to a simple reduce market distortions, elimi- an urban one. And it should be about how urban-rural split. A continuum of nate biases against agriculture, policy can best address the coordination density gives rise to a portfolio of improve the business climate, failures that arise at each stage of urban- interrelated places. Symbiosis is the stimulate competitiveness, and rule. At the head is a country's lead- promote investment and adopt ization. "The poor are gravitating to towns ing city, and below it, a spectrum new technologies. and cities, but more rapid poverty reduc- of settlements--secondary cities, Such "aspatial" policies promote rural tion will probably require a faster pace of small urban centers, towns, and development--both in agriculture and urbanization, not a slower one--and devel- villages. the nonfarm economy--so that every opment policy makers will need to facilitate A low-income country's portfolio place becomes better equipped to par- this process, not hinder it."1 And because a of places consists of primarily rural ticipate in industry and services. rural-urban transformation involves both areas. At this phase of incipient These policies will dispropor- urbanization, the mainstay of a strat- the urban and the rural, urbanization strat- tionately bene t rural households egy to facilitate spatial transforma- egies must include measures to improve because the rural nonfarm economy tions necessary for economic growth typically accounts for 30­50 percent rural lives and livelihoods (see box 7.1). is a set of spatially blind policies: of rural employment. Likewise, rural · Versatile and well-implemented households engage in diverse eco- The principle: maximize agglomeration regulations governing land mar- nomic activities, with nonagricultural economies across the portfolio of places kets to enforce property rights, sources contributing 35­42 percent Concentration, associated with rising density, safeguard land tenure, improve of household income. Growth in the brings potential benefits from "thick" mar- land transfers, ease land use con- nonfarm sector will stimulate growth kets. But it also brings congestion and squa- version to re ect market needs, in agriculture as inputs become lor. The main aim of urban policy is to help and bolster land taxation cheaper, pro ts are reinvested in agri- settlements deliver agglomeration economies · Basic and social service provision culture, and technological change to improve education and health, allows better farming methods.a while reducing the grime, crime, and time increase productivity, and encour- Source: WDR 2009 team. costs that come with rising concentration. age mobility a. Feder and Lanjouw 2001. At different stages of urbanization, the bind- ing constraints to promoting concentration Concentration without Congestion 201 while controlling congestion differ, as do the in each of these types of place. In predomi- priorities at each stage of urbanization. nantly rural neighborhoods, the policy challenge is one dimensional and corre- · Incipient. Areas of incipient urbaniza- sponds to the need to build density. In areas tion--with urban shares of about 25 per- where urbanization has gathered momen- cent--are predominantly agricultural or tum, the challenge is two dimensional. It resource based, with low economic den- incorporates the need to promote density sity. The priority is simply to facilitate and overcome problems of distance caused agglomeration forces and to encourage by congestion. In areas of advanced urban- internal economies of scale for plants, ization, the challenge is three dimensional. mills, and factories in towns. Because it is For metropolises, again, there is a need to not yet clear which places will be favored encourage density and overcome distance. by markets and for what purposes, neu- To this should be added the need to elimi- trality between places should be the nate divisions within cities, which segregate watchword for policy makers. the poor in informal slums from the rest, in · Intermediate. As urbanization pro- formally settled parts (see figure 7.2). gresses, economic alliances strengthen within and between urbanized areas. The unit for deliberating government Many firms and plants in the same sec- action: an area tor colocate to take advantage of sharing Different parts of a country urbanize at inputs and knowledge spillovers. In such different speeds. Unevenness is the rule, areas--with urban population shares of not an exception. And there are synergies about 50 percent--the promotion of and economic interdependencies among localization economies is the highest settlements of different sizes. Refram- priority. Efficiency in production and ing urbanization policies to better meet transport is the watchword. the economic imperatives at all stages of · Advanced. For highly urbanized areas, the rural-urban transformation requires productivity and consumption bene- rethinking the spatial scale for deciding fits arise from urbanization economies policy priorities and design. This Report associated with the diversity and inten- makes the case for considering policies at sity of economic activity. While func- an appropriate geographic scale: an "area," tionality is the goal for industrial towns or state or province, generally the middle and cities, the watchword for postindus- tier of government between the central and trial metropolises, with urban shares of municipal. The scale should be big enough about 75 percent, is livability. to permit both rural-urban and interurban linkages. The experience of Beijing, Shang- The policy rule: sequence and calibrate hai municipality, and Guangdong prov- The spatial dimensions of density, distance, ince supports a deliberately designed area and division spotlight the policy challenge approach to urban strategy. Two other areas Figure 7.2 The dimensions increase with the level of urbanization One-dimensional area Two-dimensional area Three-dimensional area (density) (density + distance) (density + distance + division) Incipient urbanization Intermediate urbanization Advanced urbanization Metropolis DENSITY DENSITY Large city Towns Villages DENSITY DISTANCE DIVISION Core Towns Secondary city cities DISTANCE DENSITY Source: WDR 2009 team. 202 WORLD DEVELOPMENT REPORT 2009 in western China--Chengdu and Chongq- geographic transformations for the push ing--are now taking the same approach to from middle to high incomes. urbanization, with some success. An area approach does not rule out the A framework for integration aggregation of urbanization strategies to a national level. High-density areas tend to As urbanization advances, the policy impera- have populations concentrated in metro- tives change, with the instruments spanning politan cities, intermediate-density areas in the spectrum from spatially blind to spatially medium-size cities, and low-density areas in targeted.Whilethepolicydebatesoverempha- small towns and villages. In the same way, size the most spatially explicit of government more urbanized countries have more of their actions, such as slum-upgrading programs, people in high-density areas, and less urban- successful urbanization aimed at integrating ized countries have some high-density areas, every nation's portfolio of places requires the but most people are in low-density areas. use of the full range of instruments--institu- Urbanization policies should incorporate tions, infrastructure, and incentives. this unevenness of economic development. A country's aggregate urban share can Spatially blind "institutions" to be a good indicator of the complexity of facilitate economic density the urbanization challenges it faces. In the The responsibility for building institutions simplest case, one area may characterize an that will be the bedrock for urbanization in entire country, as for Singapore. For larger all parts of the country lies mainly with the countries, a careful aggregation can help central government. Chief among them are determine the priorities at different levels those governing the management of land. In of government. this Report, "institutions" encompass three broad sets of measures: law and order (espe- · In countries where urbanization is incipi- cially the definition and enforcement of ent,such as Ethiopia,with three-fourths of property rights), the universal provision of its population in rural areas, the integra- basic services, and macroeconomic stability tion challenge is unidimensional: facilitat- (see "Navigating This Report" for details). ing density.To be sure,the capital cities and These are core mandates for the central gov- a handful of other cities even in predomi- ernment, and delivering them--or failing nantly rural nations face multidimensional to--will alter the geographic distribution challenges in their spatial transformation. of economic activities forever. But the top priority is the set of aspatial The institutions governing land markets policy instruments that apply universally include a comprehensive land registry; cred- to all places--establishing market institu- ible mechanisms for contract enforcement tions to regulate land use and transactions, and conflict resolution; flexible zoning laws; and delivering such basic services as secu- and versatile subdivision regulations that rity, schools, streets, and sanitation. So in help, rather than hinder, the conversion of the countries of Sub-Saharan Africa and land for different uses. The transformation Central Asia, the role of national govern- of the agricultural sector from one based on ments is pivotal in laying the foundations communal land rights to individual prop- of inclusive urbanization. erty rights is the sine qua non for urbaniza- · Where urbanization is intermediate--as tion. The evidence clearly shows that once in many parts of the countries of East property rights have been established and and South Asia such as India and Chi- density is increasing, land regulation and na--central and provincial adminis- planning can ensure the efficient coordi- trations must also build transport and nation of different land uses. But if regu- communication infrastructure. lations are overzealous, they can hinder · Where urbanization is more advanced, the benefits of density and agglomeration as in the countries of Latin America, economies. Similarly, rigid land use conver- North Africa, and Eastern Europe, cen- sion rules, which may be a consequence of tral, state, and municipal governments inflexible regulations, can be detrimental to must synchronize efforts to facilitate the density, as are overly restrictive minimum Concentration without Congestion 203 building standards. The fourth institution similar to that of other parts of the New is adequate housing finance. World (see "Geography in Motion: Over- Institutions for fluid land markets coming Distance in North America"). After remain important. The property rights winning the Anglo-French Seven Years' embodied in land titles are essential for War of 1756­63, the British vigorously converting assets into usable wealth.2 The debated whether to claim the small Carib- practical problems of titling, not least the bean islands of Guadeloupe (1,628 square cost of implementation, should not deter kilometers) or Canada (9.8 million square strengthening the legal framework for indi- kilometers) as the spoils of victory.8 vidual property ownership. Indeed, formal The development trajectories of North titles are necessary for functioning land and America and Latin America would diverge property markets. Although customary sys- radically. In Latin America, the Spanish tems of tenure still permit informal trans- colonialists gave large tracts of land to a actions, the absence of formal titles hinders handful of individuals, along with the right the conversion of land to areas of higher to tax the local populace. Customary com- economic return. Informality is a brake on munal property rights determined land land development, constraining an efficient use, making people less willing to move. In spatial transformation. North America, by contrast, there were few Consider preindustrial Europe. With barriers to the acquisition of land, creating more secure individual property rights to land, English cities grew rapidly.3 Indeed, England may have been the first to indus- trialize because it introduced such rights BOX 7.2 Land reform to jump-start urbanization: before other European countries. The aiding villeins in Denmark Nobel prize­winning economist Douglass North uses this to spotlight what land insti- In central Copenhagen a "pillar In 1788 the abolition of villeinage tutions can do for long-run growth and of freedom" commemorates land further improved the bargaining development.4 reform. The monument honors the power of tenants. Policy interven- nal abolition, in 1788, of "villenage," tions in the credit market also helped. The "enclosure" movement made indi- a form of serfdom common in West- In 1786 two public credit institutions vidual private property rights possible. ern Europe in the Middle Ages. At the were established to provide loans for Starting around 1500 open commons were beginning of the eighteenth century land purchases, complementing an fenced, hedged, or otherwise closed off and less than 1 percent of agricultural already rather active and e cient pri- deeded or titled to individuals. By 1545 land was farmed by land-owning vate credit market. The result: credit around 40 percent of England's surface area peasants, and large amounts of land did not constrain many prospective were common property. Required to buyers.b belonged to private individuals. The Enclo- work for landlords, "villeins" could Market-based transactions, facili- sure Act of 1604 fostered the conversion of not move without their landlord's tated by government land policies, open commons into private plots, which consent. But starting in 1760, most formed a large class of owner- continued until the early twentieth centu- communal land was transformed into occupier farmers, later a driving force ry.5 Most researchers agree that enclosures in private holdings. Between 1788 and behind Denmark's 1849 Constitution England increased agricultural productivity, 1807, landlords sold around half of and emerging democracy. By pro- which released labor from the land, and pro- their land to tenants. By 1835 almost moting education for the poor and 65 percent of the land was owner for rural residents, they also spread vided the food surplus to support the rapidly basic services and gave a strong increasing urban population.6 This allowed occupied.a A few decades later, Den- mark experienced a "take-o " into impetus to industrialization and small England to become, for a time, the "work- industrialization and urbanization. towns.c Denmark's population rose shop of the world."7 More evidence on how This urbanization did not mean sharply--in preindustrial societies, aspatial institutions initiate urbanization rural squalor. Structural, techno- a growing population is a sign of comes from Denmark (see box 7.2). logical, and institutional changes prospering agriculture. After 1890 Another example of what widespread reduced the value of tenancy to agriculture fed a surge of industrial landlords, increasing the economic growth in small townsd instrumental private property rights can do for growth, leverage of tenants. In a 1784 decree in Denmark's industrialization.e and for density, comes from North America the century-old obligation of land- and the countries of Latin America and the Contributed by Thomas Markussen. lords to collect taxes on behalf of a. Henriksen 2003. b. Henriksen 2003. Caribbean. In the early period of European the state was waived on land the c. Henriksen 2003. d. Pedersen 1990. settlement, Canada and the United States landlord sold to their former tenants. e. Christensen 2004, p. 1. were seen as having economic potential 204 WORLD DEVELOPMENT REPORT 2009 land markets and a predominantly owner- Overly stringent regulations undermine operated agricultural sector on the western investor confidence and unnecessarily dis- frontier.9 The U.S. Homestead Act of 1862 tort housing markets. Consider Mumbai.12 gave individuals the right to 160 acres of Overly restrictive land and building regula- unoccupied public land, the early founda- tions have put unnecessary upward pressure tion of a strong property rights system. on land and property prices, hampering the Contemporary research confirms the city's competitiveness. Height regulations role of well-enforced individual property hold Mumbai's buildings to only between rights. A study of 80 countries finds that a fifth and a tenth of the number of floors institutional quality does more for long- allowed in major cities in other countries. run growth than either geographic factors The city's topography should exhibit a high- or a country's openness to trade.10 Another density pattern similar to that in Hong Kong, study of 75 countries finds that security of China, but it is instead mostly a low-rise city property rights, as measured by an index (see box 7.3). Half of all poor workers com- of expropriation risks, aids development. mute less than 2 kilometers to work. And when such institutional effects are Stringent restrictions on land use con- accounted for, physical geography has weak version produce shortages of affordable effects on a country's average income.11 housing, hurting migrants to a city. For In incipient urbanization areas or in this reason, the average ratio of the median nations where the rural share is high, the house price to the median annual house- institutions governing property rights hold income in many African and Asian may be both the base and the mainstay cities is twice that in many large U.S. cit- of the policies for a rapid and sustainable ies.13 Bangladesh has a per capita income of urbanization. By contrast, when there is no $1,230, less than 3 percent of the U.S. per secure individual land and property rights, capita income of $44,070.14 But in Dhaka land transactions and urbanization may prime land prices are similar to those in become divisive. Consider China, where New York City. As much as 20 percent of land is collectively owned in rural areas and the city's inner area is underserviced. Tracts farmers do not enjoy clearly defined or fully of centrally located, publicly owned land protected land property rights. There, the remain idle and underdeveloped, while conversion of land to industrial use gener- the rest is allocated for low value-added ates social conflict because farmers perceive uses--a cantonment, public housing, and land confiscation as inappropriate without residential areas for government workers. fair compensation. The problems are simi- Stringent land development parame- lar in many Sub-Saharan African countries, ters--including minimum plot sizes and where 90 percent of the land is communal. road widths, setbacks, and land for commu- Land use and building regulations nal facilities--exclude a majority of house- become important as urbanization holds from formal land ownership. Indeed, advances. Governments regulate land mar- although the underlying plot might be titled, kets for two reasons. First, regulation can the dwelling may be rendered illegal because ensure the appropriate separation of land of the failure to meet official construction between different uses, such as preventing standards. Without a downward revision the location of heavily polluting industries of standards, the benefits of legal title are in residential areas. Second, it can ensure lost. Such legal codes also contribute to red the integration of private and public uses of tape and excessive housing costs.15 In Addis land, such as providing space for transport Ababa high construction standards have rel- infrastructure in densely populated areas. egated many low-income households, which But land use regulation can be overzealous, might otherwise possess a tradable title, to upsetting the delicate balance between the the ownership of "illegal property."16 public interest and private opportunity. In 1979, the federal government in Bra- A city's future depends on investor confi- zil, by passing a national land use regulation dence in its prospects and its responsiveness setting a minimum lot size of 125 square to future changes in the demand for land. meters and frontage of 5 meters, effectively Concentration without Congestion 205 BOX 7.3 Bombay fights the markets, and more than half of Mumbai's residents live in slums The city of Mumbai, once known as Bom- of Greater Mumbai went the other way, is curtailed. The government relies on bay, provides sobering lessons. In the lowering the permitted FSI to 1.33 in 1991. property taxes and on in ated real estate 1960s and 1970s, city planners decided Almost all buildings in Mumbai with an prices for revenue, so it has little incentive that Bombay's population should be FSI exceeding 4.5 were built before 1964. to ght the groups that resist relaxation controlled at about 7 million. Land regu- Under the rules that existed until recently, of building height restrictions. lations and infrastructure policies were new buildings, including those in the cen- The result is a vicious circle of supply designed accordingly. But people ooded tral business district, were subject to the shortages and high land prices. Mumbai into the city anyway, and today the city is FSI of 1.33. As a consequence, space con- slipped from 25th place to 40th in the more than twice the intended size, with sumption in Mumbai averages 4 square league table of "best cities for business" the highest population density of any meters, much less than the 12 square between 1995 and 1999. It remains metropolitan area in the world. Estimates meters in Shanghai and the more than India's premier business city--it topped indicate that 54 percent of Mumbai's 16 20 square meters in Moscow. And about Chennai and Bangalore in investment million people now live in slums, and half of its residents are huddled within 2 in 2007 and was the top destination for another quarter in degraded apartments. kilometers of the city center (see the gure domestic migrants. But how quickly it The Floor Space Index (FSI)a regula- below). reforms its regulations and builds infra- tions in Mumbai were introduced in 1964, Meanwhile, high housing costs account structure will decide how long it will stipulating the maximum building space for as much as 15­20 percent of the keep this position. for every square meter of the plot of land. income of a low-income family. Rent In Mumbai it was set at 4.5. The standard control regulations freeze 30 percent of Source: WDR 2009 team; Bertaud 2003. practice in cities with limited land is to Mumbai's housing stock, leaving it dilapi- a. The FSI is the ratio of the total oor space in a building to the area of the plot on which raise the permitted FSI over time to accom- dated because landlords see little point it is built. For example, suppose a building modate urban growth, as in Manhattan; in investing. Weak property rights imply covers half of a plot that is 1,000 square Singapore; Hong Kong, China; and Shang- that only 10 percent of the housing stock meters in size. If this building has 10 oors, it hai. Instead, the Municipal Corporation has legal title, so land redevelopment exhibits an FSI of 5. Land should be used better in Mumbai, and people should not live so close to work km2 30 % of households 25 35 Sea 30 20 25 All households 15 20 Poor households 15 10 10 5 5 0 0 0 5 10 15 20 25 30 35 40 45 50 0­1 1­2 2­3 3­5 5­10 10­15 15­20 20­30 >30 Distance from central business district (km) Distance to work (km) Source: Bertaud 2003. Source: Baker and others 2005. Undeveloped land outside Airport municipal limits Port Undeveloped land inside Industries municipal limits Slums Transport Built-up area 206 WORLD DEVELOPMENT REPORT 2009 excluded many low-income residents from Occupant-owned housing, usually a access to land in the formal sector. Many household's largest single asset by far, is suburban low-income developments could important in wealth creation, social secu- not meet the 125-square-meter requirement, rity, and politics. People who own their but were built on land owned by developers. house or have secure tenure have a larger Outside the formal sector, such develop- stake in their community and thus are ments could not be legally supplied public more likely to lobby for less crime, stronger services. So more than half the population governance, and better local environmental in regional capitals, such as Recife and Sal- conditions.21 vador, lives in slums or informal areas.17,18 After democratization in 1988, cities were Spatially connective infrastructure encouraged to service the informal sector. to reduce distance to density Although the national minimum lot size Policies to unify land markets and facilitate requirement remained in place, areas could labor mobility remain important for the be designated as Special Zones of Social buildup of economic density at all stages of Interest and exempted from the require- urbanization. But they are not enough for ment as part of a package to secure tenure dealing with the more complex challenges and improve services. of advancing urbanization. In Seoul and Land market restrictions can have Shanghai, downtown traffic averages 8 kilo- adverse spillovers on the urban economy meters an hour; in Bangkok, Manila, and and blunt the instruments intended to aid Mexico City, it averages 10 kilometers an urbanization. Unless basic institutions for hour or slower; in Kuala Lumpur and São land markets and social services are in Paulo, it averages 15 kilometers an hour or place, infrastructure development will be slower. Workers in Jakarta, Kinshasa, Lagos, hindered, and spatially targeted interven- and Manila spend on average 75 minutes tions will likely be ineffective. commuting to work.22 For such areas, con- Regulations for housing finance affect gestion can eat away the benefits of rising urbanization. Since the deregulation of density. Spatially connective infrastructure financial systems in the second half of the must join the spatially blind institutions as 1980s, market-based housing financing has priorities for inclusive urbanization. expanded rapidly. Residential mortgage Connective infrastructure needs insti- markets are now equivalent to more that 40 tutions. Successful cities react to growing percent of gross domestic product (GDP) in traffic congestion with spatially connective developed countries. But those in develop- infrastructure. But preceding such infra- ing countries are much smaller, averaging structure in all successful cities (or accom- less than 10 percent of GDP.19 The public panying it in the fastest urbanizers) is a fluid role should be to stimulate well-regulated land market and an empowered local gov- private involvement. For example, private ernment. The sequencing of policies should initiatives of the type developed by the be spatially blind measures to create condi- Grameen Bank in Bangladesh, Bancosol tions suitable for economic concentration, in Bolivia, and the Housing Development followed by connective policies to deal with and Finance Corporation in India show congestion. that uncollateralized lending can nour- The United Kingdom in the nineteenth ish housing finance even in countries with century is illustrative. With systems of budding financial systems and weak legal governance varying widely across towns, and regulatory structures. Establishing the the Reform Act of 1832 and the Municipal legal foundation for simple, enforceable, Corporations Act of 1835 brought about and prudent mortgage contracts is a good regularization of municipal government.23 start. When a country's financial system Municipal authorities could take over pri- is more developed and mature, the public vately owned sewerage, water, and gas sector can encourage a secondary mortgage systems. By the 1880s they had started pur- market, develop financial innovations, and chasing land to compete with private utili- expand the securitization of mortgages.20 ties, transport, and other services. In doing Concentration without Congestion 207 so, they unified the hodgepodge of preexist- 2 million in 1930.30 To address ever more ing private systems, separating sewage and complex urbanization, the City Planning drainage systems from the water systems, Commission was created in 1938. After and extending the reach of basic services to studies and public debate, the 1916 Zoning poor areas.24 The Land Enquiry Commis- Resolution was replaced in 1961. The new sion recognized that "municipal land own- resolution incorporated parking require- ership, town planning and the building up of ments and emphasized open space. the system of transit will go hand in hand and Although based on the leading plan- each will help the other."25 By the end of the ning theories of the day, aspects of those nineteenth century, institutions governing zoning policies have revealed shortcom- land markets were maturing and adapting ings over the years. The emphasis on open to the changing urban requirements.26 space has sometimes resulted in buildings With this as background, the United that overwhelm their surroundings. Since Kingdom's urbanization was rapid. In 1830 then, new approaches have been developed the average GDP per capita was $1,749 (in to make land use conversion more respon- 1990 international prices), roughly equiva- sive to changing needs. A more flexible lent to Honduras, Mozambique, or Paki- approach at the Department of City Plan- stan in 2003.27 The urban share rose from ning encourages a mix of uses that creates 28 percent in 1830 to 69 percent in 1910.28 lively urban streetscapes that can sustain At the top of the urban hierarchy: London, increased density.31 whose population grew from 2 million in New York City provides an example of 1830 to 6.6 million in 1900.29 the changing spatially blind institutions Institutions and infrastructure must necessary for spatially connective policies. evolve continually. As areas urbanize and Indeed, their interaction enabled the den- nations develop, the networks for public sity of Manhattan, the Bronx, Brooklyn, transit become more complex, and institu- and Queens to increase from 230 people per tions such as legislation governing land use square kilometer in 1820 to more than 5,000 must also adapt. Building a new transport in 1900 and about 12,000 today. network requires the purchase of contigu- Successful urbanization requires con- ous plots of land, and holdouts can extract necting ever wider areas. Inevitably, density huge rents or thwart the project entirely. brings crowding. New York shows the enor- Compulsory purchases ("eminent domain" mous benefits of an efficient metro system in in the United States) may be necessary, with reducing congestion while encouraging den- the safeguard of just compensation for sity. The key is an integrated system of mass land owners. Another safeguard is that the transport (see box 7.4). Dense city centers acquired land be for "public use," although and skyscrapers are feasible only when thou- how widely this should be interpreted can sands of office workers can be transported be contentious. efficiently to downtown offices. The United States, by the mid-nineteenth Long-term success does not rule out century, had a reasonably well-defined occasional bouts of congestion, but it does system of property rights. As New York's require flexible institutions. A British pam- transportation network expanded, and phlet published in 1860 observed that the needs of the city changed over the past century, its institutions evolved. The 1916 [F]rom day to day, and from year-to-year, the Zoning Resolution has been amended to streets of London become more and more crowded, and must . . . come to a dead-lock respond to shifts in population and land between Westminster and the City, unless use. Waves of immigration helped swell the some more efficient remedy can be provided. city's population from 5 million in 1916 to The great City lies as it were handcuffed, almost 8 million in 1960. New mass transit panting and exhausted under the weight of routes and growth corridors were created. its own wealth.32 And with the rise of the mass-produced automobile, car registrations in New York London's congestion appeared no closer State exploded from 93,000 in 1915 to about to resolution in 1939, with traffic averaging 208 WORLD DEVELOPMENT REPORT 2009 of the substitution of one scarce resource, travel time, for another even more scarce BOX 7.4 Widening the reach of New York City resource, land. It is not efficient to devote New York's subway system has more than 100,000 passengers on so much urban land to roads that conges- become one of the busiest and most the day of the opening ceremony. tion is completely eliminated. Some urban extensive in the world, serving nearly Subway trains, at close to 40 miles an congestion is likely to be optimal, if for no 5 million passengers every day with hour, were much faster than trolleys other reason than to prompt policy mak- 26 train lines operating on 800 miles (6 miles an hour) and elevated trains ers to review and update institutions and of track. As New York City spread into (12 miles an hour). More people could a wider metropolitan area, commuter now be moved at faster speeds. infrastructure. bus networks and rail lines grew. It has been a never-ending struggle As cities specialize, intercity infrastruc- New York City's commuter rail system to expand the transport system fast ture becomes a priority in the most dynamic is the most extensive in the United enough to accommodate population areas. There is a symbiosis between cities States, with about 250 stations and growth. Most of the subway system and their peripheries, but economic rela- 20 rail lines serving more than 150 in use today was built between 1913 million commuters annually.a tionships also join other cities in an urban and 1931; the number of annual rail hierarchy. Transport links between cities Public transportation in New York passengers jumped from 500 million City began in the late 1820s with in 1901 to 2.5 billion in 1929.b In 1940 reinforce agglomeration economies and horse-drawn omnibuses. The rst the city uni ed the three indepen- generate complementary and specialized steam-driven cable car line opened dent subway lines under public own- functions. In the United States the mega- in 1883. In 1909 electric trolleybuses ership, allowing for a more integrated lopolis stretching from Boston, through replaced them, and for 70 years trol- approach to transport development. New York City, Philadelphia, and Baltimore leybuses ran in all ve boroughs of The payo is inclusive and sustain- to Washington along the northeastern coast New York City. The rst elevated line able urbanization. New York's Metro- is linked by highways and rails (the first ("el") opened in 1868. By 1880 most politan Transportation Authority has Manhattan residents were within a served a 5,000-square-mile region freight rail link was between Baltimore and 19-minute walk of an "el," which took since 1968. According to the 2000 Washington in 1827).35 Japan also invested passengers above the congested U.S. census, New York City is the only in spatially connective infrastructure to streets. locality in the United States where link its two largest agglomerations Tokyo- The mid-1880s saw rapid immigra- fewer than half of all households own Yokohama and Osaka-Kobe (see box 7.5). tion. Overcrowding was rife. As in a car--the gure is even lower in In the Republic of Korea, Seoul-Incheon London, an underground rail network Manhattan at fewer than a quarter-- (in the northwest) and Pusan (the second was seen as necessary. But it took a compared with 92 percent nationally. blizzard in March 1888, completely One in every three users of mass tran- largest city at the southeast tip) were linked paralyzing the streets, to provide sit in the United States and two-thirds as early as 1905 by the Gyeongbu Rail, and the impetus for an underground rail of the nation's rail riders live in New in 1970 by the 400-plus-kilometer Gyeo- system. The subway was designed York City and its suburbs.c ngbu Highway. In 2004 the bullet train--the both to move people about within Source: WDR 2009 team. Korea Train Express--connected the two Manhattan, and to connect tracts of Note: a. The New York City Transit cities in two hours' travel time. In the Pearl undeveloped land. Museum Teacher Resource Center, and River Delta--a region of China studded After years of political wrangling, the Port Authority of New York and New Jersey O cial Web site. b. The New York a plan for a subway was approved in with factories and interwoven by freeways-- City Transit Museum Teacher Resource 1894. In 1904, the Interborough Rapid the economies of several cities are linked Center. c. The New York City Transit Transit Company opened and carried Museum Teacher Resource Center. together so effectively that, according to the chief executive officer of a large electronics manufacturer, "In practice, we are a single 8 miles an hour,33 not so different from vast factory scattered across the territory." Demand management and public trans- the 10­11 miles an hour in central London today.34 This may suggest an "equilibrium port encourage development of higher den- sity. There are many instruments to increase level" of congestion. The city's economic connectivity, among them--36 density consistent with this equilibrium depends on the quality of spatially connec- · Improving transport options--say, tive infrastructure. In this sense, the value through better transit management that of additional investment in such infra- increases the use of, or gives preference structure is not as much in the long-run to, high-occupancy vehicles abatement of congestion as in the continu- · Managing land use--through transit- ally rising economic density for any given oriented development or smart growth level of congestion. Congestion is the result that gives preference to new develop- Concentration without Congestion 209 BOX 7.5 Promoting concentration in Japan between 1860 and 1980: spatially connective policies for Tokyo-Yokohama and Osaka-Kobe Japan's manufacturing industries are Industries that left the traditional promote concentration while preventing spatially concentrated--a trend that can industrial cores were mostly exporters the grime and time costs of rising density. be traced to the Meiji era starting in the of machinery and electronic appliance These e orts did not interfere with the 1860s. In Tokyo both state-owned fac- plants. They continued to enjoy localiza- pro t motives of enterprises, but instead tories and private industrial complexes tion economies from producing similar strengthened agglomeration economies. were concentrated along the main river. and related products in new clusters. Government policies and market forces Gradually small machinery workshops Their locations alongside the Tomei high- reinforced each other spatially to sustain conglomerated, and industries expanded way connecting Tokyo and Nagoya gave economic growth. toward the south along the new Tokaido easy access to markets and high-tech Contributed by Keijiro Otsuka and Megumi railway connecting Tokyo, Yokohama, and enterprises in the urban centers. Muto. areas farther south. The geographic distribution of indus- Sources: Fujita and Tabuchi 1997; Sonobe and After World War II, when exports to the tries over several decades of rapid growth Otsuka 2006; Whittaker 1997; Overseas Eco- United States began to accelerate, indus- re ected the government's e orts to nomic Cooperation Fund 1995. trial production became concentrated in the Keihin industrial zone around Tokyo and Yokohama, and in the Hanshin Connected cities in Japan facilitate agglomeration economies in Tokyo-Yokohama (Keihin) and industrial zone around Osaka and Kobe. Osaka-Kobe (Hanshin) by road and rail This led to heavy tra c congestion, water shortages, and air and water pol- lution. In 1962 the Japanese government responded by instituting Zenso--the Inte- grated Spatial Development Plan--which aggressively developed the Paci c Ocean Industrial Belt by linking the core agglom- erated areas between Tokyo and Osaka Sapporo and establishing new industrial zones in HOKKAIDO between. The investments included the bullet train (Shinkansen) and other trunk railways, expressways, and ports (see the map to the right). Despite heavy infrastructure invest- ments in new industrial clusters in more remote regions, they could not attract industries out of the Paci c Ocean Belt. During the miraculous growth era of the 1950s through 1970s, industries remained spatially concentrated, thanks HONSHU to the mobility of workers, even though there has been massive relocation of industries from the congested core to surrounding new industrial areas. Tokyo Enterprises that remained in core urban Osaka clusters upgraded from standardized products to high-tech products and new models by taking advantage of urbanization economies accruing from SHIKOKU diverse economic activity and a large pool of skill and talent. Other enterprises KYUSHU retained their central management functions in the core agglomerations to Main railroads bene t from the convenience of face- Main highways to-face communications with banks, government o ces, and major industrial organizations. Source: WDR 2009 team. 210 WORLD DEVELOPMENT REPORT 2009 ments along established public transport buses operating on unsegregrated and con- routes gested roads. But suburban rail and subways · Launching general policies and pro- require huge investments in fixed capital, so grams--such as freight transport man- dedicated busways (plus their more sophis- agement and market reforms. ticated relation, "bus rapid transit") have been gaining in popularity. Pricing mechanisms can be most effec- Busways, most common in Latin tive in optimizing private car use. Extreme American cities, cost about $10 million a but effective are Singapore's auction per- kilometer to install. Operating in Bogotá, mits to purchase cars. Add the car taxes, Colombia; Curitiba and São Paulo, Brazil; and the cost of a car in Singapore is four and Quito, Ecuador, they are being planned to five times the world price. Amsterdam, or built in many other cities. A more expen- London, and Stockholm also have schemes sive alternative, ranging in cost from $10 that price road use according to the time of million to $30 million a kilometer, is light day and level of congestion--substantially rail, a modern form of tram covering short reducing peak-time car traffic and car emis- distances. It usually feeds a larger system sions. The revenue streams from congestion of heavy metro rail. Cities with light rail charging could be earmarked for reinvest- include Hong Kong, China; Kuala Lumpur; ment in public transportation. But such Singapore; Sydney; and Tunis. schemes require substantial investment in The most costly mass rapid transit option technology to ensure the efficient collection is the metropolitan subway system, which of fees. has the largest capacity. Building costs aver- Easier to implement are simple regula- age more than $100 million a kilometer, tions or traffic plans that reduce the num- explaining why there are fewer than 200 sys- ber of vehicles in specific parts of a city tems in the world, mostly in industrialized or overall. In Tehran entry to parts of the countries.39 But their number is growing: city center is restricted to essential traffic. China, India, and the República Bolivariana Budapest and Buenos Aires have pedes- de Venezuela have built subways. When cit- trian-only zones in the city center, easily ies reach a certain size and density, a subway reached by public transport. Gothenburg is the only transport mode capable of mov- (Sweden) and Bremen (Germany) restrict ing large numbers of people to concentrated private car links between different zones job centers. The benefits that come from ("cells"), encouraging public transit. India's enabling such density include efficiency Chandigarh built some 160 kilometers of and productivity gains--traditionally in wide-cycle paths to ease traffic on arterial industry, increasingly in services (see chap- roads.37 The most popular restraint limits ter 4)--but also lower energy consumption, the use of vehicles on specific days accord- less pollution, and greater compactness, ing to their registration plate number, as which increase interaction and encourage in Athens, Bogotá, Lagos, Manila, Mexico nonmotorized transport for short intracity City, Santiago, São Paulo, Seoul, and Sin- trips. Compact and densely packed cities gapore. Such measures have been proven might also meet the imperatives of climate easier to enforce than expected, with wide- change (see box 7.6). spread public acceptance.38 Public transport, successful on its own, Demand management is the most cost- has encouraged new developments at higher effective means of increasing mobility. But densities, which in turn permit more suc- traffic will increase even with the best poli- cessful public transport while reducing the cies, especially in rapidly growing cities. economic distance between places. Manag- Investments in public transport infrastruc- ing all this takes patience and the discipline ture can connect different parts of a city to build from the bottom. The establish- and guide land use and urban expansion. ment and strengthening of land and prop- Mass rapid transit includes subways, subur- erty market institutions--including secure ban rail, and dedicated busways, all having property rights, flexible land use regula- a capacity and performance far superior to tions, and ease of land conversion--is not Concentration without Congestion 211 BOX 7.6 Climate change calls for a different urban form, not slower urbanization Urbanization is associated with industrial- correlated with carbon emissions.d At the network is 99 kilometers, and 60 percent ization, which increases emissions of car- national level, Sweden and Japan have of its population lives within 600 meters bon dioxide (CO2) and other greenhouse used incentives and regulation to greatly of a metro station. Only 4.5 percent of gases. And increasing wealth tends to be reduce the emissions intensity of their trips are by mass transit in Atlanta, a associated with higher energy consump- economies. At the urban level, an empha- fraction of the 30 percent in Barcelona. tion, for instance through motorization. sis on density and smart choices that For Atlanta to achieve Barcelona's metro But to be concerned about the climate reduce distance can help do the same. accessibility would require building an does not mean that urbanization should This requires land use policies that favor additional 3,400 kilometers of metro be slowed. If anything, economic density compactness and transport policies that tracks and about 2,800 new metro sta- may need to be encouraged even more. guide urban form and provide convenient tions. This would allow the Atlanta metro Historical data going back to the nine- and e cient public transit.e to transport the same number of people teenth century show that today's rich Atlanta and Barcelona illustrate alter- that Barcelona does with only 99 kilome- countries experienced rising per capita native urban growth scenarios. They ters of tracks and 136 stations. carbon emissions as they urbanized and had similar populations of 2.5 million Density makes the di erence. industrialized through the twentieth to 2.8 million, but Atlanta had a density century.a Industrialization, motorization, of six people per hectare in 1990, and Source: WDR 2009 team. a. Marland, Boden, and Andres 2007. b. and consequently carbon emissions in Barcelona had 176.f In Atlanta the longest World Bank 2002, gure 2.1; Lanne and Liski developing countries follow the trajecto- possible distance between two points 2003, gures 1, 4, and 5; and data in devel- ries of developed countries in their earlier within the built-up area is 137 kilome- oping countries from http://cdiac.esd.ornl. stages of development.b For instance, per ters; in Barcelona, the distance is only gov/ftp/ndp030/nation.1751_2004.ems. capita carbon emissions in Germany dou- 37 kilometers. Per capita CO2emission c. Bento and others 2003. d. Scholz 2006; bled from 0.8 metric tons of carbon in 1880 was 400 metric tons in Atlanta, 38 tons Vance and Hedel 2006; Golob and Brown- stone 2005; Ingram 1997; International to 1.6 in 1900. In the United States and the in Barcelona.g Atlanta's metro network Union of Public Transport. e. Bento and United Kingdom, carbon emissions were is 74 kilometers long. but only 4 percent others 2003; Scholz 2006; Vance and Hedel about 2.5 in 1900. Today's developing of its population is within 800 meters 2006; Golob and Brownstone 2005. f. Ber- countries have lower average emissions of a metro station. Barcelona's metro taud 2004. g. Kenworthy 2005. at the equivalent GDPs per capita of Ger- many, the United Kingdom, and the United Countries can change their energy trajectories States in 1880 and 1900. Botswana's car- Per capita carbon emissions (tons, metric) bon emissions were 0.36 per capita in 1987 20 and 0.57 in 1996 (see gure at right). The trend in most developing countries 18 suggests continuing growth in carbon emissions both in total and per capita. 16 The policy response to the projected increases in urbanization and carbon 14 emissions in developing countries should not be to try to prevent the growth of 12 cities. This would not be feasible or desir- able in light of the evidence on growth 10 and poverty reduction. Instead, growth Germany, France, 8 in cities--many of which might double in Canada, the United Kingdom, size over the next few decades--should and the United States 6 1880­1940 be managed to create urban areas far more carbon e cient than many of 4 today's mature cities. Monocentric structures and high popu- 2 lation densities tend to reduce the length and number of motorized trips.c Compact 0 cities use less energy for transport, con- 0 5 10 15 20 25 30 sume less land for housing, and use less GDP per capita (1990 international Geary Khamis $, thousands) energy for heating. Several studies nd Sources: Carbon emission data worldwide can be found in the Oak Ridge National Laboratory, U.S. Department that high population density is negatively of Energy at http://cdiac.esd.ornl.gov/ftp/ndp030/nation.1751_2004.ems; Lanne and Liski 2003. 212 WORLD DEVELOPMENT REPORT 2009 easy. But without the commitment to such authorities powers to retain the houses built institutions, and without investment in under slum-clearance schemes, paving the connective infrastructure, targeted inter- way for future public housing schemes. ventions to deal with slums are unlikely to Better transport was part of the solution, work. evidenced by conferences in 1901 on the subject of "Improved Means of Locomo- Spatially targeted interventions to tion as a First Step towards the Cure of reduce social and economic divisions the Housing Difficulties of London." The For the three-dimensional problem faced conferences passed a resolution that "a by advanced urbanization areas, spatially complete system of transportation radiat- blind and connective instruments must be ing from urban centers, and which shall be supplemented with spatially targeted inter- cheap, rapid, and under municipal owner- ventions to address the social and economic ship, is a primary step towards dealing with division within a city's boundaries--most the housing problem."40 Londoners obvi- visibly, slums. The lesson of experience ously understood that connective policies is that spatially targeted efforts succeed must precede targeted interventions. when they are applied where land markets Indeed, the link between improvements work reasonably well, basic social services in spatially connective transport infra- are widely accessible, and a connective structure and the solution of London's slum infrastructure links the city's core to its problem was made clearly in the policy periphery. discourse of the time. In 1890 the Cheap Institutions and infrastructure are Trains for London Workers Bill proposed prerequisites for successful interventions. extending the provision and further regu- Successful programs to integrate slums lating the fares on "workmen's trains." The have been built on a foundation of spatially private railway companies already had been blind and spatially connective policies. This obliged by law to introduce these trains in integration included the establishment of 1883, to provide an affordable means of institutions to effectively govern the work- commuting to working-class workers who ing of the housing market, a spatially blind lived in the suburbs but earned their living provision of social and basic services to all in central London. By lowering the cost of settlements, and transport infrastructure commuting, the suburbs could be devel- investments to connect the newly created oped, decongesting London's central areas. housing areas. Likewise, during the late-nineteenth Slum clearance requires a legislative basis and early twentieth century, governments to empower local authorities to take action across North America and Western Europe and institutions to facilitate an orderly con- implemented large-scale slum clearance versionoflandfromagriculturaltoresidential and re-housing. Landowners were com- or other purposes. At the same time, subur- pensated, and the cleared land was sold for ban development is made feasible through redevelopment. These programs would not the provision of basic amenities and social have succeeded without a rapid expansion services alongside ongoing improvements in of transport infrastructure. New transport the transport connectivity of cities with their systems helped "open up" the outskirts, or suburbs and surrounding counties. suburbs, of cities, making periurban hous- In London social outcry at the dreadful ing attractive for both real estate develop- conditions of Victorian slums provided the ers and urban workers. The resulting flight impetus for slum clearance and improve- to the suburbs was also both a cause and ment. But the efforts were preceded by a consequence of the relocation of many steps to improve housing markets and industries to the peripheries. And it coin- transport systems. The Housing of the cided with the spread of basic social ser- Working Classes Act 1890 provided local vices and recreational amenities. authorities with the power to build houses Policies to integrate slums into cities for the working classes and to clear areas have worked where institutions and infra- of unfit housing. An amendment gave local structure were adequate. After World Concentration without Congestion 213 War II, Sweden urbanized rapidly and the new housing developments to the local Stockholm's population grew swiftly, from city economies. As infrastructure improved 741,000 in 1950 to 1.39 million in 1980.41 even more, particularly after the passage of Stockholm had an inadequate and dilapi- the Housing and Urban Development Act dated housing stock, while rents were high 1965, the more prosperous residents left the relative to most other European cities. 42 In city centers for the suburbs.48 reaction, the Swedish government formed Better connective infrastructure is a pre- the Royal Housing Commission in 1945. A condition for applying targeted policies to plan was formulated to demolish slums in deal with slum housing. This takes time, Stockholm and other cities, and re-house but both Hong Kong, China, and Singapore the displaced slum dwellers in publicly show that it can be done over decades rather provided rental housing in well-designed than centuries. Regardless of differences in high-rise buildings on the city's periphery. the speed, the sequencing of policies appears The first generation of high-rise residential to be the same: targeted policies to integrate buildings was integrated with the provi- slums cannot come before the application of sion of schools, health clinics, and recre- geographically blind and connective policies ational and shopping facilities, as well as (see box 7.7). service centers. Spatial connectivity to city centers was ensured through easy access to transport.43 BOX 7.7 Speeded up, but still in sequence: spatial integration Swedish authorities managed to con- in twentieth-century Hong Kong, China tinually upgrade urban living conditions throughout the 1960s and 1970s. With the Before World War II, Hong Kong, spatially blind institutions had to Million Homes Programme the government China, developed its administrative be adapted over time as the city developed and the urbanization set itself the aim of ending innercity squalor structure and legislative framework to govern land markets. In 1935, progressed. The 1939 Ordinance was and overcrowding by building 100,000 new mounting awareness of poor living amended in 1958, 1969, and 1974.a dwellings a year from 1965 to 1974, adding conditions in urban slums led to the Having established the necessary one-third to Sweden's aggregate housing formation of a Housing Commission. planning framework, Hong Kong, stock of 3 million units.44 The new settle- This was followed by the Town Plan- China, was better placed to imple- ments provided basic amenities, including ning Ordinance of 1939, which estab- ment spatially connective policies in schools and clinics, and were linked to urban lished a Town Planning Board. the 1970s. These policies were a nec- essary response to the doubling of employment centers through well-planned Nevertheless, the proper imple- car registration within a decade and traffic systems.45 mentation of the ordinance and town planning in Hong Kong, China, had the concomitant increase in conges- Similar lessons come from the United to wait until after the passage of the tion that was a product of the city's States. By the end of the nineteenth cen- Town Planning Regulation in 1954. It rapid economic growth of around 10 tury, American philanthropists had raised was only after the famous Shek Kip percent a year. awareness of the hardships facing slum Mei slum re of 1953 that e orts to Institutions and infrastructure went dwellers. They urged building regulations develop public housing programs hand in hand. With e ective plan- ning laws in place, the government to ensure minimum standards in the con- went into full swing. In 1965 the was able to introduce the Temporary struction of new tenements.46 But it was not Working Party on Slum Clearance was formed. It took Hong Kong, China (a Restriction of Building Development until the 1930s that the government became city in a hurry), more than 30 years Ordinance of 1973 in the Pok Fu Lam active in the provision of housing.47 Follow- before it began e ectively address- and Mid-Levels areas of Hong Kong, ing the Great Depression, the United States ing the problem through spatially China. This in turn paved the way for Housing Authority was established by the targeted interventions. Hong Kong, building the Mass Transit Railway, Wagner-Steagall Housing Act of 1937, which China, rst had to develop and modifying building height restric- tions in the area around the Kai Tak allowed for subsidized loans to be made strengthen the spatially blind institu- tions governing the operation of land Airport, and accelerating relief for an to local housing authorities for clearing and housing markets, and connective overloaded transport network. dilapidated areas and building replacement infrastructure to improve the use of The result: the city now ranks homes. A sound legal framework enabled land. among the world's top ve in infra- national and state authorities, civil society The rst land use strategy and structure e ciency, and the slums are organizations, and private developers to zoning plan--the "Colony Outline gone.b deal concertedly with slums. Good intra- Plan and Outline Zoning Plans"-- Source: WDR 2009 team. urban public transport systems connected were only drawn up in 1963. The a. Bristow 1984. b. Cullinane 2002. 214 WORLD DEVELOPMENT REPORT 2009 Targeted interventions may have to wait South Africa's experience is more sober- until institutions and infrastructure have ing. When the first post-apartheid South been improved. The experience of developed African government came to power in 1994, countries remains relevant today. Costa Rica, it faced a housing crisis with an estimated South Africa, and Singapore show why. deficit of 1.5 million housing units and During the 1980s, the combination of an additional requirement of 170,000 new rapid in situ population growth, migration, units a year. Some 18 percent of households, and an influx of refugees from war-torn or 7.4 million people, lived in slums.53 The neighboring countries made Costa Rica's new housing policy that emerged from already acute urban housing shortage even multiparty negotiations was implemented worse. This led to the creation of a National through the Housing Act of 1997, provid- Housing Finance System in 1986. The aim ing housing-related subsidies to as many was to provide subsidies to low-income people as possible for renting, purchasing, households for house purchases or construc- constructing, and improving homes. tion. Households were able to supplement But the requisite institutions and con- the funds that they received with loans from nective infrastructure were not yet in place. private institutions, including commer- A range of reforms--including changes to cial banks, savings and loans institutions, regional government boundaries, financial and cooperatives. Minimum wage house- support mechanisms, and housing-related holds were entitled to a full subsidy, while legislation--were introduced simultane- households earning more than four times ously. Indeed, the Housing Act repealed, the minimum had access to smaller subsi- incorporated, or amended 35 separate dies and loans at near-market conditions.49 pieces of legislation.54 The first wave of low- Costa Rica's housing subsidy succeeded cost housing developments, on the periph- because the necessary spatially blind insti- eries of major cities, lacked basic amenities tutions and spatially connective infrastruc- and transport links to city (and job) cen- ture were in place to facilitate its targeted ters. These developments failed to amelio- interventions. As early as 1869 the govern- rate intraurban divisions. Recent shifts in ment decreed that primary education was policy are more encouraging. The Breaking a basic universal right. The 1949 constitu- New Ground national housing program has tion guaranteed free access to secondary focused on integrating low-income com- education as well. Costa Rica's government munities through improvements in access invests more than 20 percent of its budget to public transportation and basic social in education each year, and has a literacy and commercial services.55 rate of about 95 percent to show for it. Improving institutions and infrastruc- Similar investment in the public medical ture and intervening at the same time is system reduced infant mortality rates, with a tall order for any government, but Sin- an average life expectancy at birth of 79 gapore shows it can be done (see box 7.8). years.50 Although incomes are much higher Perhaps the most successful example of how in Costa Rica's central regions, social indi- slums can be eradicated, Singapore is to cators are similar across the country.51 some extent an anomaly. It is one of the few Costa Rica has well-functioning institu- countries that have managed to implement tions governing the land market. An effec- all three sets of integration policies simulta- tive property registry system is in place. neously. Singapore was successful because it Indeed, more than 80 percent of property had exceptionally rapid economic growth owners possess registered titles, and there is and a focused government in power since a high degree of legal security.52 The capital 1965. The fact that it is a city-state helped city, San José, is linked with the major pro- greatly, indirectly providing a general les- vincial cities by an efficient and affordable son: successful urbanization takes coordi- bus system, and private bus companies con- nated action at all levels of government. nect San José and its outlying suburbs. In Most countries will not be able to rep- many ways, Costa Rica is a model for other licate Singapore's efforts--aligning priori- developing countries. ties and the effort of central, state, and city Concentration without Congestion 215 governments is difficult for any country that is not a city-state. More likely, they will BOX 7.8 Singapore: from slums to world city have to sequence their policy efforts along the path followed by Costa Rica and earlier At independence in 1965, 70 percent ernment became a major provider developers such as Great Britain, Sweden, of Singapore's households lived in of infrastructure and services. The scarcity of land made good plan- and the United States. South Africa shows badly overcrowded conditions, and a third of its people squatted on the ning an imperative. Multiyear plans the difficulties of trying to implement all city fringes. Unemployment aver- were produced, implemented, and three sets of policies simultaneously. aged 14 percent, GDP per capita updated. Finally, the housing author- was less than $2,700, and half of the ity (HDB) was mandated to undertake "An I for a D"--a policy instrument population was illiterate. Falling a massive program of slum clearance, for each dimension of urbanization's mortality rates and migration from housing construction, and urban challenge the Malay Peninsula implied rapid renewal. Public housing has been an integral part of all development The sequence of policies corresponds to population growth, further increas- ing the pressure on both housing plans. At the height of the program, different levels of urbanization. Incipient and employment: 600,000 additional HDB was building a new at every urbanization requires mainly the applica- units of housing were needed, and eight minutes. Of Singapore's popu- tion of spatially blind policies. Intermediate private supply was less than 60,000. lation, 86 percent now lives in pub- urbanization requires the addition of spa- An account of this time comes from a licly built units. Most own their ats, tially connective policies. And advanced contemporary visitor to Singapore:a encouraged by special housing funds urbanization requires these and spatially nanced from the Employees Provi- The undercover walkways are usually dent Fund, a mandatory retirement targeted efforts. So the success of the new taken over by hawker stalls and junk. scheme. Serviced land was made policy is predicated on the successful imple- Laundry hangs from poles thrust out of windows above--just like in old Shang- available. Through the Land Amalga- mentation of the ones introduced before it. hai. This is Singapore, in the early 1970s. mation act, the government acquired For a predominantly rural country whose We were all devastated at the time--we almost one-third of city land. Slum who didn't live here. From 1871 to 1931 urban share is less than a quarter or so, the the city's Chinese population rose from dwellers were relocated to public portfolio of places is faced with what might 100,000 to 500,000. By 1960 it is estimated housing. be termed a one-dimensional challenge--to that more than 500,000 Chinese were For a city-state in a poor region, it living in slum-like conditions--indoors. is not an exaggeration to assert that facilitate density (see table 7.1). It is not Equipped with only one kitchen and one e ective urbanization was respon- obvious where this density will increase bathroom, the shophouses were designed for two extended families at most. After sible for delivering growth rates that first, and governments are best advised to extensive partitioning many of them averaged 8 percent a year through- allow market forces to play themselves out. housed up to 50 individuals. out the 1970s and 1980s. It required Neutrality between places is the rule, and its a combination of market institutions Today, less than 40 years later, urbanization strategy should consist mainly and social service provision, strategic Singapore's slums are gone. In their of spatially blind institutions. They include investment in infrastructure, and place is one of the cleanest and most improved housing for slum dwellers. the provision of basic and social services, the welcoming cities in the world. The establishment of market institutions and law secret? First, institutional reforms Sources: Yuen 2004, Yusuf and and order, the security of property rights, made the government known for Nabeshima 2006. its accountability. Then, the gov- a. Cockrem 2007. the efficient operation of the land market, and sound macroeconomic policies. Regu- lations must be versatile enough to facilitate efficient land use conversion, and standards a two-pronged response: continuing the for building must be enforceable without task of building spatially blind institutions, being overly restrictive. This is a tall order and investing in spatially connective infra- for governments in countries at the low lev- structure to offset the congestion that might els of income with which low urbanization otherwise offset the efficiency gains from rates are associated. They should not make "localization economies" (see chapter 4). it harder by attempting spatially explicit For a highly urbanized country with policies. urban shares above 75 percent, urbaniza- For a rapidly urbanizing country with tion should emphasize livability, creativity, urban shares between one fourth and three and urban social integration--the delivery fourths, managing its portfolio of places is of "urbanization economies." These coun- mainly a two-dimensional challenge--to tries face a three-dimensional challenge--to build density and reduce distance to den- build density, reduce distance, and diminish sity. A two-dimensional challenge requires divisions. To be sure, these countries have 216 WORLD DEVELOPMENT REPORT 2009 Table 7.1 An instrument per dimension--a simple framework for urbanization policies Area Incipient urbanization Intermediate urbanization Advanced urbanization Urban shares Less than 25 percent About 50 percent More than 75 percent Examples Kampong Speu, Cambodia; Lindi, Chengdu, China; Hyderabad, India Greater Cairo, the Arab Republic of Egypt; Rio Tanzania de Janeiro, Brazil Dimensions of policy 1-D: Build density 2-D: Build density, reduce distance 3-D: Build density, reduce distance, eliminate challenge division Instruments for integration Institutions Land rights; basic education, Land use regulations; universal Land use regulation and land taxation; health and water and sanitation provision of basic and social services universal provision of basic services Infrastructure Transport infrastructure Transport infrastructure; demand management Interventions Slum area development; targeted programs to reduce crime and environmental degradation Source: WDR 2009 team. a varied economic geography: their port- level roughly the same as Benin, Cambodia, folio of places consists of a handful of one- or Tajikistan. Since then it has transformed dimensional areas, a good proportion face itself into a leading industrial country.56 two-dimensional challenges, and some face Consistent with the stylized facts of chapter three-dimensionalchallenges.Spatiallyblind 1, the Republic of Korea's sectoral transfor- and connective policies continue to facilitate mation has been accompanied by an equally agglomeration economies, but now they also radical spatial transformation. In 1960 are prerequisites for successful interventions about 75 percent of all Korean citizens lived to reduce within-city divisions. in rural areas. By 1990 the country was 75 percent urban, and today the urban share of the population exceeds 80 percent.57 The framework in action Institutions to ensure the universal Low-density areas should build economic availability of basic social services helped density through rural-urban transforma- the nation lay the foundations of rapid and tions and stronger links between villages successful urbanization. In 1960 the pro- and towns. Rapidly urbanizing areas should portion of the overall population age 15 and ensure that the productivity gains from eco- over with no schooling was 36 percent, and nomic density are not offset by congestion by 1980, when it had entered the intermedi- costs. Highly urbanized areas should focus ate urbanization stage, this proportion had on livability by promoting social integra- fallen to less than 15 percent. By 2000, some tion and the gains from economic concen- years after it entered the advanced stage, tration. The priorities at the national level the proportion was less than 5 percent. The correspond with the predominance of one years of schooling of the average member or more of these types of areas. of the labor force had increased from five years in 1960 to nine years in 1980, rising Incipient urbanizers (one-dimensional to more than 12 years by 2000.58 A simi- areas): institutions for more efficient lar story unfolded for health-related ser- rural-urban transformations vices. In 1980 only 4 percent of children The recent developers show that success were immunized against measles. By 1989, does not require explicitly spatial policies 95 percent were. In 2006 only one child in to lay the groundwork for successful urban- every 100 was not immunized. ization. For areas of incipient urbanization, Matchingtheuniversalprovisionofeduca- the policy priorities remain the provision of tion and health services has been the nation- basic social services and the improvement wide flexibility in land use conversion.59 In of land markets. particular, the Republic of Korea's govern- Emphasize social services. In 1960 the ment has been willing to convert agricultural Republic of Korea had a GDP per capita land for industrial purposes. The central Concentration without Congestion 217 government also encouraged local govern- coastal provinces (see map 7.2). Its GDP per ments to promote the conversion of agricul- capita in 2005 was only 34 percent of the Chi- tural land through the formation of smaller, nese average,61 with almost 75 percent of its more localized industrial complexes. population classified as rural. The challenge While some areas have inevitably been Guizhou faces is building density to facilitate left behind in the Republic of Korea's urban- agglomeration economies. Its 11th Five-Year ization process, none has been left disadvan- Plan (2006­10) aims to deliver an urban share taged. Take Eumseong county, a largely rural of 35 percent by focusing on the area's larg- area in Chungcheongbukdo province (see est city, Guiyang. With the improvements map 7.1). As the Republic of Korea indus- in spatially blind institutions in China, this trialized and urbanized, the county experi- enced a continual outflow of people. In 1968 Map 7.1 The Republic of Korea--three areas at different stages of urbanization the population exceeded 120,000, but by 1990 it had fallen to just under 75,000. But even as the people of Eumseong were seeing their neighbors move closer to Korea's major cities, they got better education and health services and improved streets and sanitation. Between 1969 and 1990, middle and high school teachers tripled in Eumseong county 3D Area: Seoul Metropolitan Area from 1,000 to around 3,000. And the num- KANGWON ber of hospitals per million population in Chungcheongbukdo province doubled from KYONGGI around 400 in 1980 to 800 in 1990, while the water supply coverage increased from less 1D Area: Eumseoung than 30 percent to almost 60 percent. People County left Eumseong, but the Korean government NORTH did not abandon the county--instead, it CHUNGCHONG continued to emphasize the universal provi- SOUTH NORTH sion of basic and social services. CHUNGCHONG KYONGSANG The Republic of Korea is not the only successful economy to provide evidence of the framework in action. Over the last two 2D Area: Daegu decades, China has been gradually putting in Metropolitan Area NORTH place the institutions to improve its urban- CHOLLA ization processes. An urban land market has been created, and regulations standardizing SOUTH KYONGSANG the assignment of land use rights have been PUSAN established.60Inthe1980stheurbanplanning SOUTH law was aimed at controlling the size of large CHOLLA cities, but the 10th Five-Year Plan (2001­05) instead chose to emphasize the synergistic development of China's large, medium, and small cities. The household registration sys- tem, which for years had imposed restric- tions on rural migrants looking to move to urban areas, has been reformed, and the 11th Five-Year Plan (2006­10) aims to further strengthen land market institutions. Against this backdrop the urbanization prospects of China's one-dimensional areas CHEJU have improved. Take Guizhou province. In southwest China, and home to almost 40 million people, Guizhou lags far behind the Source: WDR 2009 team. 218 WORLD DEVELOPMENT REPORT 2009 Map 7.2 China--three areas at different stages of urbanization system. The results are more security of ten- ure--leading to easier transactions, higher land values, and more land investments-- and greater mobility to urban areas. Kampong Speu illustrates the critical issues for areas of incipient urbanizers. Spatially blind policies to encourage rural- HEILONGJIANG urban integration should be the mainstay of a government's strategy: improve land markets and property rights, improve rural and urban social services, and encourage JILIN inclusive governance in towns and small NEI MONGOL cities (box 7.9 discusses the importance of LIAONING land titling in Cambodia and Vietnam). BEIJING Secure tenure promotes greater invest- Beijing ment in land and shelter, improves the abil- TIANJIN ity to transfer land, and enhances access to HEBEI GANSU credit. China, Pakistan, and Vietnam con- NINGXIA SHANXI SHANDONG firm its importance for investment in low- QINGHAI density areas. Farmers use more labor and inputs on owned plots than on leased land. SHAANXI HENAN JIANGSU They also use land as collateral for new ANHUI activities and benefit from increases in land HUBEI SHANGHAI prices. In India the prices for titled land are, SICHUAN on average, 15 percent higher than those for ZHEJIANG CHONGQING JIANGXI untitled land. In the Philippines secure hous- 1D Area: Guizhou Province ing commands prices 58 percent higher than HUNAN 2D Area: Changsha­Zhuzhou­ Xiangfan Cluster housing without title, and in Jakarta prices GUIZHOU FUJIAN for secure housing are 73 percent higher.62 TAIWAN Despite these obvious advantages, more YUNNAN GUANGXI GUANGDONG than 50 percent of the periurban population HONG KONG in Africa and more than 40 percent in Asia MACAO lives under informal tenure. In many coun- 3D Area: Pearl River Delta tries improving land tenure (and registra- tion) is hindered by political and customary HAINAN tenure arrangements. In Africa, where cus- tomary institutions cover between 90 and 98 Source: WDR 2009 team. percent of the land, policies to formalize land appears to be more realistic now than would tenure must start with customary systems have been the case two decades ago. and gradually add features of modern land Define and enforce land rights. Kam- registration. Once community-recognized pong Speu province in Cambodia, about 100 rights are obtained in Benin, Ghana, kilometers southwest of Phnom Penh, covers Mozambique, and Namibia, individuals can about 7,000 squares kilometers. It has about apply for land certification and full registra- 700,000 people, and is 10 percent urban. A tion, and both can be used for credit. few market towns serve its farming com- Improve land administration. Central munes of Mohasaing, Ou, Traeng Trayeung, Asia and Eastern Europe have the most sys- and others. Many villages are beneficiaries tematic experience in tackling land admin- of the land registration and titling begun in istration, from Central European countries 2000. The aims are to strengthen land tenure with old traditions of land markets, to Cen- security and land markets, formulate policies tral Asian countries where no land markets for land administration and management, existed. During the transition from plan to develop mechanisms for dispute resolution, market, countries tried to reestablish equity and establish a national land registration in land and property rights, deepen land Concentration without Congestion 219 BOX 7.9 Titling land for a sustainable rural-urban transformation "Application receipts" This view was backed by the 1992 Red books increase land market in Cambodia land law, which ruled that an application activities, and these activities increase Cambodia is formalizing ownership of receipt is a valid claim on the plot. In 2004 agricultural productivity by transferring land through the distribution of land the government initiated a comprehensive land to the most productive users, reduc- titles, with immediate bene ts in higher land management and administration pro- ing ine ciencies.f The liberalization of productivity and land values.a Although gram, and one of its central components the land market has been followed by the courts and other formal institutions includes a systematic titling scheme. By increased mobility as households sell to resolve land con icts are slow, formal the end of 2005 about 457,000 plots had land to take up new economic oppor- land rights documents carry authority been registered under this program, and tunities in the wage labor sector.g So a in most rural communities. If landhold- 166,000 titles had been distributed.e more uid land market has facilitated the ers hold a formal title to a disputed plot ongoing shift in Vietnam from a predomi- of land, their position is considerably "Red books" in Vietnam nantly agricultural economy to a more strengthened, even if the con ict is In 1981 Vietnam moved to a system simi- diversi ed and urbanized economy. It is a resolved through informal means. lar to the Chinese "household responsibil- big part of a strategy that has yielded per- In 1989 individual land use in agricul- ity system," with land users entitled to haps the most impressive poverty reduc- ture was codi ed by law.b Rural residents keep surplus production above a xed tion in any country in recent history.h were encouraged to submit applications quota. In 1988 some individual property Contributed by Thomas Markussen. for land ownership certi cates, and appli- rights to agricultural land were trans- a. Markussen forthcoming; World Bank 2003b. cations came in for 4 million plots. Due to ferred to farmer households. In the 1993 b. Boreak 2000. the limited administrative capacity and land law, the distribution of land use c. Sophal, Saravi, and Acharya 2001. reach of the government, only about half certi cates--known as "red books"-- d. Sovannarith and others 2001. a million titles were actually distributed.c e. Deutsch 2006. was mandated. Red books come with f. Deininger and Jin 2003; Ravallion and van But people who applied got an "applica- the rights to sell, rent, mortgage, and de Walle 2006b. tion receipt," and this document often bequeath land. So the idea of a land mar- g. Ravallion and van de Walle 2006a. worked as a formal title.d ket was formally sanctioned. h. World Bank 2003b. and capital markets, and improve public functions such as land taxation, planning, BOX 7.10 Land markets in transition and asset management. Now they have to improve mechanisms to enable registrations, In the rst phase, Eastern Europe and cadastres. More e cient, a com- valuations, and transactions (see box 7.10). Central Asia restored property rights, bined system is easier to make self- nancing. Instituting mechanisms to govern land privatized state-owned assets, and promoted equity in housing. Next, use and conversion can be difficult. Some · A local champion is needed, pref- they rebuilt the land administration erably not a surveyor or a lawyer. nations fear that land conversion would hurt systems for cadastre and registration Enthusiastic change managers grain production and food security (China followed. Proper records were needed were instrumental in Moldova and and the Arab Republic of Egypt). Many oth- to stimulate real estate markets and the Czech Republic. Competent ers are constrained by traditional forms of take care of land allocation and con- o cials in the Russian Federation, land tenure, such as the communal systems solidation. Information infrastructure, Serbia, and Turkey were needed in Africa and the ejido in Mexico. In cases of institutional capacity, and databases to design and implement the new were the areas of focus. Now some are unclear property title, land conversion tends systems. Systematic registration entering the third phase of collecting was not necessary because good to benefit the state and the developers at the property taxes, managing public land, land records predated socialism, cost of the farmers or rural households that and issuing building permits. and there were few transactions traditionally held or cultivated this land. Some lessons: during the socialist period. In Mexico the traditional communal land · Reforming dysfunctional legal and · A solid system with Web-based system has evolved to enable land transac- institutional systems such as those applications to reduce user tions. After the 1917 revolution, Mexico dis- in Latin America, may be more di - transaction costs and the oppor- tributed more than 100 million hectares, or cult than starting anew (the Kyrgyz tunities for corruption can be 50 percent of its arable area, from large farms Republic, Georgia). updated on its own. to ejidos, rural communities organized along · A single agency should be respon- Contributed by Cora Shaw and Gavin P. the precolonial indigenous social structures. sible for both registration and Adlington. But the redistribution undermined property rights, and the requirement that land be used for self-cultivation precluded rental markets. 220 WORLD DEVELOPMENT REPORT 2009 In 1991 the system gave more freedom to eji- congestion. The priorities include pro- dos to sell and rent land. Of 150,000 hectares viding social services for rural and urban used for urban development between 1995 residents, ensuring fluid land markets, and and 2000, more than two-thirds were from investing in infrastructure in and around ejidos. The off-farm income of the farm- the growing city centers. ers increased 45 percent. (Box 7.11 presents Expand administrative jurisdictions promising examples.) to coordinate infrastructure investments. Among the many cities that have absorbed Intermediate urbanization rural Koreans are Seoul and Daegu. Both (two-dimensional areas): institutions of these cities were initially able to urbanize and infrastructure for increasing against a backdrop of spatially blind poli- density and reducing congestion cies, but they soon began to face congestion, Rapidly urbanizing areas expect a con- which required spatially connective policies. tinuing influx of migrants and increasing Indeed, housing congestion in Seoul became BOX 7.11 Strengthening land market institutions for rural-urban integration Land management on Douala's secondary rights allow small-scale farm- for better land use planning in urban urban fringe--Cameroon's ers to hire migrant workers in exchange areas. Areas for improvement include the Mbanga-Japoma Project for temporary rights to cultivate plots. equity of service provision, since rural and Mbanga-Japoma, a land development proj- The system is highly structured, with urban residents have di erent needs, and ect in Douala, Cameroon, provides serviced speci c days of the week designated for the weak rural voice, since urbanites are land at a reasonable price and reconciles laborers and others who are working on believed to be more organized. formal and customary development prac- the borrowed land. Land tenure in the vil- Land consolidation in Indonesia tices. The rst phase, covering 160 hectares lage is almost exclusively under the cus- The Land Consolidation Program imple- 30 kilometers from the city center, started tomary system, controlled by the village mented in Indonesia in the 1990s shows as a partnership among public institutions, council, allowing for secondary rights how to facilitate the orderly develop- formal private investors, and customary allocations. Strong demand from nearby ment of fast-growing areas and to plan owners. The partnership develops the site urban markets for horticultural produce the development of vacant areas on the with primary and secondary infrastructure makes cultivating of even a small plot urban fringe. services (roads, water, sewerage, drainage, pro table and thus attractive to migrants. The mayor has the authority to deter- electricity), delineating blocks of land of Inclusive administration--the mine the location of consolidation areas between 1 and 8 hectares. The developer Republic of Korea's integrated cities and to manage and supervise the pro- gives back 45 percent of the land to cus- The Republic of Korea developed the rural- cess. But the key actors are the private tomary landowners, keeping 55 percent. urban integrated city to overcome the landowners and the occupiers of (state- Blocks are then subdivided and sold, either shortcomings of earlier rural development owned) land. by the developer or by customary owners. initiatives. The integrated city policy incor- The minimum requirement for land The nal cost of a serviced plot is much porates rural counties with cities in a uni- consolidation is to have at least 85 per- lower than one provided by the formal ed spatial framework. It aims to improve cent of landowners representing at least private sector. local public services and local administra- 85 percent of the land area give their Although there are questions about eli- tion and reduce rural-urban disparities. agreement. All participants contribute gibility for purchasing the serviced plots, Starting in 1994 the government by providing land for infrastructure and the approach provides a new perspective selected 49 cities and 43 counties as can- services. The amount of land a partici- for partnerships in managing rural-urban didates. The selection criteria included pant is required to give up is determined land use in and near Sub-Saharan cities. historical homogeneity, natural topo- by consensus. Small landowners who Secondary land rights and farming graphical conditions, and the potential cannot contribute land can contribute in central Mali for balanced development within the money or labor. The contributions fund Secondary land rights--including share- integrated city. The selected cities and infrastructure and utilities--and build cropping, tenancy, and borrowing land counties held public hearings and citizen a pool of "cost-equivalent land," to under customary tenure--are often seen surveys. After this screening, 41 cities and be used only by small landowners to as exploitative because they do not give 39 counties were amalgamated into 40 enlarge their parcels. permanent tenure to users. But in some rural-urban integrated cities. circumstances they can bene t both sec- Attitude surveys suggest that residents Sources: WDR 2009 team; Groupe Recher- ondary and permanent rights holders. In and local councils see the bene ts. Every- che/Actions pour le Développement 2001; the village of Baguinéda, in central Mali, one agrees that the integrated city makes and Kim 1998. Concentration without Congestion 221 a severe problem in 1960 as the area grew Map 7.3 Decentralization and localization through the intermediate stage of urbaniza- economies in Daegu tion and absorbed a large influx of people Daegu urban area from abroad after the country gained inde- Daegu city center pendence from Japan and from rural areas Highways KTX High-speed train in the Republic of Korea. To help tackle this Industrial centers problem, the area of Gangnam, south of the 90 km Han River, was absorbed into Seoul's terri- Pohang tory in 1963, and the Gangnam Development Gumi 60 km Machinery & DAEGU auto: Program initiated. This program involved a Electronics 80 million 31 million 30 km series of spatially connective infrastructure mobile phones cars per year per year, projects spread over 30 years, including 1.4 million several bridges across the Han River and a LCDs per year 54.2-kilometer circular subway line to link Ulsan Automotive: Gangnam with central Seoul. 1.8 million cars per year Daegu's story is similar. Between 1950 and Busan 1990, Daegu's population swelled by a factor Container terminal: 1,500 kton of six, from 355,000 to nearly 2 million,63 as Changwon Machinery & auto: 240,000 cars per year its thriving textile industry pulled in rural Source: Daegu Metropolitan City. migrants seeking a better life. The policy response was to integrate Daegu and its hin- terland by expanding its administrative zone in 1987 and again in 1995, followed by build- forefront of the Republic Korea's export- ing a subway system and expanding the city led industrialization. bus system. The city also experienced con- Other examples are Chongqing and tinual building and upgrading of local roads. Chengdu, both rapidly urbanizing areas In 1980 just over 40 percent of Daegu's local in southwestern China. In line with an roads were paved, and by 1995, virtually all unabashedly urbanization-based economic had been paved. growth strategy, China is piloting the "area Combining universal access to the most approach" in western China. At about 43 basic services and reasonable land markets percent, they have similar urban shares to with investments to improve spatial con- the average for China. The objective is to nectivity with other areas of the country, increase those shares to 70 percent in 2020, Daegu has thrived. Manufacturing has but in a manner that speeds the concentra- deconcentrated from Daegu into the sur- tion of economic activities while reducing rounding Gyeongsangbukdo province, rural-urban disparities in living standards. and the local economy has diversified, The mainstays of the strategy are institu- reducing its reliance on the textile indus- tions and infrastructure (see box 5 in the try and moving into sectors with higher overview). value added production. Daegu now sits at If markets favor the two places as much the center of a vibrant urban system sur- as the central and provincial governments rounded by five cities, all with easy trans- have, the two cities will improve the lives port access to the central city, and each of millions in the Chinese hinterland. The having evolved to provide localization initiatives have already had a local impact. economies (see map 7.3). Gumi has been In Chongqing rural incomes in the first half dubbed "Korea's Silicon Valley" for its of 2007 increased faster than those of urban specialization in electronics, while coastal residents. Foreign investment is about the Pohang and Ulsan, respectively, provide same as it was in Shanghai a decade ago. homes for the Pohang Steel Company and Industries are attracted to the low wages Hyundai. Ulsan also houses one of the big- and low cost of land. According to govern- gest shipbuilding industries in the world, ment statistics, average wages, at $2 to $3 a and both these cities have been on the day, are much lower than those in Beijing 222 WORLD DEVELOPMENT REPORT 2009 or Shanghai. In Chengdu farmer concen- intensification in central city areas. The trations are believed to have increased pro- plan sets out a series of spatially connective ductivity by 80 percent. Industrialization policies aiming to promote connectivity has been absorbing about 100,000 farmers between the three cities in the cluster. These a year, with some narrowing of rural-urban include highway and rail-based expressway income differences. projects to connect Xiangtan with Zhuzhou Several areas at the intermediate stage of and ring roads around each of the three cit- urbanization in China appear to conform ies. The plan is a good illustration of how to the principle of an additional instrument the principle of "an I for a D" can be made for the dimensional challenge of conges- operational using an area approach. tion. An example is the tri-city area in the Invest in transport connectivity while northeast of Hunan province, in central continuing institutional reform. Devel- China (see map 7.4). The cluster of three oped metropolitan areas rarely leave city cities--Changsha, Zhuzhou, and Xiang- growth unplanned. Land and housing tan--has a combined population of around markets help allocate residential and office 13 million and accounts for about one-fifth space. Rapidly expanding systems need of the province. With an income exceeding clear property rights to provide incentives the national average by 17 percent and the for land transactions and correct land valu- Hunan average by 61 percent, this cluster is ation to avoid an urban bias and too much an intermediate urbanization area with an migration to the city. Singapore develops urban share of half. land and housing plans every 10 years and An area plan--the first of its kind in lets the market function once public and inland China--was formulated in 2005.64 private sectors agree on which economic The plan specifies a regulatory land use activities to develop and which residential planning framework within which market patterns are needed to accommodate firms prices will be allowed to allocate land for and workers. different uses. It also provides planning Urban transport, along with urban land guidelines including, for example, enforc- management, determines the shape of the ing land use rights and promoting land city and its ecological footprint. Urban mobility is particularly important for the poor. In Buenos Aires 87 percent of the jobs Map 7.4 Changsha, Zhuzhou, and Xiangtan--spatially in the metropolitan area are accessible in connective infrastructure in a two dimensional area 45 minutes. In Mexico City 20 percent of workers spend more than three hours trav- eling to and from work each day. The urban poor in Beijing and Shanghai spend less CHANGSHA than 5 percent of their income on transport Changsha because they walk or cycle. If they chose to Zhuzhou travel by bus, the costs would be 40 percent Xiangtan Loudi Pingxiang of their income.65 Brazil's vale de transporte XIANGTAN JIANGXI is an effective way to subsidize poor workers in the absence of good urban transport-- ZHUZHOU financed by the central government and by HUNAN the employer in equal parts. Several large Hengyang cities have public transport networks that are used extensively--metro in Delhi, Kol- Main connective artery kata, and Mumbai, buses and metro in São Paulo--but the network quickly becomes Secondary transport axis inadequate and congested because of the Tertiary transport axis rapid population growth. Regular mainte- Source: China Urban Planning and Design Institute and Hunan nance and new investments in infrastruc- Development and Reform Commission 2005. ture are needed to sustain density in urban Note: The main artery is formed by the Beijing-Guangzhou railway, Beijing-Zhuhai highway, and State Highway 107. areas (see box 7.12). Concentration without Congestion 223 Put infrastructure in the most promis- ing places. Several countries have created BOX 7.12 Retrofitting transport infrastructure in Bangkok new cities to move a capital city (Brasília), to decongest the capital (Seoul), or for eco- In the 1990s it was estimated that the Thailand, or to suburban districts of nomic reasons. Creating new cities with average car in Bangkok spent 44 days Bangkok, because of the enduring attraction of Bangkok's agglomera- the sole purpose of diverting population each year stationary in tra c. How did this situation come to pass? And tion economies, sociocultural ameni- from the capital is often risky, evidenced how is it being remedied? ties, and key export infrastructure, by Brazil, France, República Bolivariana de A city of around 7 million people, including its port. Venezuela, the United Kingdom, and more Bangkok is the product of hundreds In recent years several yovers and recently Egypt and Nigeria.66 New cities of years of incremental growth along elevated expressways have been built, become attractive to private investors only traditional land con gurations. The along with an elevated railway system after they reach a threshold, but there is no result is a city woven with narrow (Skytrain), dedicated bus lanes, and way to know that threshold.67 And when lanes, many of them culs-de-sac two peripheral ring roads. But car (called soi), but with few arterial ownership has shot up, too, adding to cities are created far from main transport roads. Indeed, the arterial roads can the tra c and diminishing the impact networks and business centers, they are be as far as 7 kilometers or more of remedial investments. Looking to unlikely to be economic successes. apart. According to a recent estimate, the future, congestion pricing and When markets identify promising cities, roads account for only about 6.1 increased parking fees appear to be strategic investments in infrastructure and percent of Bangkok's land area in promising policy options. Reducing public goods can accelerate their potential the inner city, and only 1.7 percent in fares on the Skytrain and extending rapid transit to more of Bangkok, for economic growth (see box 7.13). Second- its peripheral areas. In high-income countries it is usual for 20 to 30 per- perhaps using bus rapid transit, pose ary cities that promote access to markets, cent of urban land area to be devoted greater challenges. improve city management, and build human to roads. Even with this extreme con- Contributed by Austin Kilroy. capital seem a better alternative. And if polit- gestion, economic activity has been Sources: Angel 2008, Bae and Suthira- ical concerns dictate the creation of new cit- slow to decentralize to other cities in nart 2003, and Gakenheimer 1999. ies, efficiency concerns will guide locations to be close to growing markets and to have access to infrastructure. Working with exist- of 12 residents lives in a slum, and about ing cities is preferable to creating new cities one-third of new citizens in recent years are from scratch. But, if new cities are created, rural migrants. The city has taken steps to they should be constructed on an appropri- make urbanization inclusive. It built better ate scale, close to markets, and planned to schools, renewed parks, started community generate demand-side links. centers, and improved main networks for water and sanitation. Since 2000 a pub- Advanced urbanizers lic-private bus rapid transit system, the (three-dimensional areas): TransMilenio, has improved citywide acces- institutions and infrastructure for sibility. Travel times have fallen by an aver- higher density and shorter distance, age of 15 minutes, with larger reductions and targeted incentives to for households in poorer parts of the city address divisions (see map 7.5). Aided by these infrastruc- Successful metropolitan areas in both ture improvements, Bogotá's internation- developed and developing countries have ally recognized Programa de Mejoramiento well-functioning land markets, representa- Integral de Barrios has assisted the poorest tive management, state-of-the art transport neighborhoods to integrate with the city. infrastructure, and social policies to inte- Begun in 2003 it has already helped 930,000 grate low-income residents. people. The program is believed to have Use an inclusive mix of institutions, contributed increases of up to 11 percent in infrastructure, and incentives. Colom- house values.68 bia's capital, Bogotá, shows the resolve and The Republic of Korea also provides resources needed for inclusive urbaniza- lessons. In the 1950s it had an estimated tion in a metropolitan region. Although 136,650 unregistered slum districts, more the area's income places it in upper-mid- than 2,200 in the core of Seoul.69 Spatially dle-income levels, 43 percent of its popula- targeted policies to redevelop Seoul's slum tion of 6.7 million is considered poor. One areas started as early as the mid-1960s. 224 WORLD DEVELOPMENT REPORT 2009 BOX 7.13 New cities: escapes from urban jungles, or cathedrals in the desert? New cities were attempted in Europe with- record at best. Cities closer to Cairo ness within the area and with other cities out much success. In the United Kingdom have attracted businesses and people, in China; and its urban form: the Barlow Commission Report of 1940 though much fewer than anticipated. · Access to foreign markets. Locating the stimulated interest in new towns. Between Cities distant from Cairo (including Sadat SEZ close to the city of Hong Kong, 1947 and 1968, Britain created 26 new City, supposedly the new capital) remain China, facilitated foreign investment, towns to control the growth of London unattractive for skilled labor due to lack technical assistance, and access to for- and stimulate development in Scotland of amenities and transport links. The new eign markets. and Wales. In 1965 France followed a simi- cities have no more than 1 million inhabit- · Connectedness within the area. To lar program--nine towns, ve in the Paris ants (1 percent of Egypt's population), spread the fruits of development, the area and four in lagging areas, were con- compared with the 5 million target set boundaries of the municipality were structed. These programs soon were inter- for 2005. The program was also costly: 22 expanded to extend the bene ts of rupted and put aside as unsustainable. The percent of the Ministry of Infrastructure's the city to all workers. The rural hukou new towns never reached their targeted investment under the Fourth Plan (1997­ was abolished in the municipality, and population, nor did they forestall the 2001) was spent in these new towns. This all urban services became accessible growth of London or Paris. The experience will increase if the government continues to all residents. Placing the Shenzhen in developing countries has been mixed. its policy of developing the urban fringes. city-area in the Pearl River Delta area The emphasis on attracting investment Failure in the Arab Republic of Egypt ensured the best possible links to its was not balanced by the need to make Egypt's program of new cities is the hinterland and other urban nodes in cities attractive for skilled labor and acces- world's largest. In 20 years Egypt has the Delta regions. Complementary sible from the established urban centers. built 20 new cities and is preparing for decisions to ease the mobility and Eventually, the creation of the new cities 45 more. The rst set of 24 cities was integration include investments in had little impact on decongesting greater launched in 1974­75 as a manifestation of transport infrastructure and a shift Cairo. the political commitment to conquer the from a road-based to a rail-based desert and ensure sustainable growth. Success in China system. Large industrial zones were created and China's approach recognizes the need to · Functional urban form. The compre- generous tax incentives were given to the create cities with access to major markets hensive plan for Shenzhen envisions a private sector. Land was virtually free. The and transportation networks. Shenzhen polycentric metropolis that connects " rst generation" of new towns included was the rst special economic zone the SEZ to urban nodes through e - six towns, each with its own industrial (SEZ) to be approved by Deng Xiaoping cient transport. base and large target populations.a Ten in 1980. From a small town with 30,000 years later--by the mid-1980s--the next inhabitants, it grew to 800,000 in 1988 program based on satellite settlements and 7 million in 2000. The new residents was launched, and nine second-genera- include the best-trained professionals in Sources: WDR 2009 team; Stewart 1996; and tion settlements were launched around the country, attracted by high salaries, World Bank 2007k. Greater Cairo. A third generation included better housing, and education opportu- a. For example, Sixth of October had an original target population of 500,000, twin towns close to provincial capitals, nities for their children. GDP per capita which was raised in the late 1980s to 1 mil- such as New Thebes. increased more than 60 times. lion, and currently the target is 2.5 million. The performance of the six cities cre- Shenzhen owes its success to its near- The actual population is probably less than ated 30 years ago suggests a mixed ness to Hong Kong, China; its connected- 200,000. But without the requisite investments in was closing in on the advanced stage of spatially blind institutions and spatially urbanization, and Seoul's population level connective infrastructure in place, these had started to stabilize (see map 7.6). Fur- policies succeeded only in relocating the thermore, enough time had elapsed for the slums to new slum areas such as Mokdang, earlier spatially blind and connective poli- Nangok, and Shillim near the Guro Indus- cies to take full effect. trial Complex, whose textile and other Between 1984 and 1990, expanding industries provided low-skill employment road capacity struggled to keep up with the opportunities for rural migrants. growth of vehicle ownership. More exten- The Republic of Korea's government sive plans of city-periphery integration were would have more luck when it tried again required, and in 1989 the New City Devel- with an expanded set of policies in the opment plan was launched. Five new cities 1980s. By this time, the Republic of Korea were encouraged around 25­30 kilometers Concentration without Congestion 225 Map 7.5 Bogotá's TransMilenio has helped to integrate the poor TransMilenio Minutes spent during rush hour within transport system routes 80 Feeder routes 70 Average minutes saved, weighted by number of 60 passengers 18 50 15 40 10 10 13 10 30 20 Poverty: Low 10 0 1 2 3 4 5 6 Poorest Richest Neighborhood stratification Before the TransMilenio High After the TransMilenio Source: Yepes 2008. from Seoul. Central to the success of this In 1960 Istanbul's population was about integration were investments in connec- 1.5 million, a modern-day Kansas City in tive infrastructure. The subway system was the United States. With a population more extended, and a beltway was constructed, than 10 million today, Istanbul is now one easing traffic congestion. of Europe's largest cities, about the same During this period, 93 slum districts covering an area of 427 square kilometers Map 7.6 Economic density in Seoul with good connections to other cities were modernized, including the Wolgoksa- dong and Mok-dong slum areas.70 Dwellers in the former benefited from successful in Seoul situ upgrading; those in the latter were relo- Ansan cated as the government cleared the area and replaced it with a modern apartment complex. Squatters benefited from moving subsidies of about $2,000 per person and Chongju the right to purchase a new apartment at a Taejon Daegu Ulsan discount.71 Turkey has also transformed itself from a Chonju predominantly rural society to a primarily urban one over the past half-century. Since Pusan Kwangju becoming a member of the Organisation for Economic Co-operation and Develop- Changwon ment (OECD) in 1961, Turkey's urban share increased from around one-third to two- Mokpo thirds,72 as GDP per capita more than tri- pled to about $6,600.73 Driving this increase in density was the rapid growth of Turkey's Source: WDR 2009 team and World Bank Development Research Group, based on subnational GDP estimates cities, foremost among them, Istanbul. for 2005. See also Nordhaus 2006. 226 WORLD DEVELOPMENT REPORT 2009 size as Chicago.74 Much of this growth has population. Almost half of the city's resi- been accommodated on the Asian side of dents--some 5­6 million people--live the city, home to successive waves of rural in dwellings that are or were gecekondu. immigrants. It is now the origin of a daily Although settlements such as Sultanbeyli tidal wave of commuters who make their have integrated themselves into the city, way across the Bosporus to work on the others like Pas¸aköy have not.77 Istanbul European side of the city. still needs targeted programs to deal with Underpinning Turkey's transformation the divisions associated with the continu- to an urban economy are the spatially blind ing existence of poorly serviced and under- reforms that accompanied the creation of integrated informal communities.78 the modern republic of Turkey. The Turkish China's Pearl River Delta faces similar constitution of 1924 included the adaptation challenges. The area consists of nine cities of European laws to the needs of the new and has a population roughly equal to that republic, endorsing private property rights. of Spain, representing one of China's most The reforms introduced under Mustafa advanced urban agglomerations. With an Kemal Ataturk unified the country's edu- urban share of almost 75 percent in 2006, cation system, providing the backdrop for it is a three-dimensional area with the better access to education over the last half- triple challenge of building density, reduc- century. In 1960 the proportion of the adult ing distance, and overcoming divisions. population more than 15 years old with no The divisions are manifest in "urban vil- schooling was 67 percent, and in 2000, 18.6 lages," many of which would be known as percent.75 And better health care services slums elsewhere. They lack access to basic helped increase life expectancy from 51 years sanitation services and are subject to envi- in 1960 to 71 years in 2005.76 ronmental degradation. The cities of the The reforms of the 1920s helped to lay Pearl River Delta area have been introduc- the foundation for Istanbul's rapid expan- ing spatially targeted policies to deal with sion, but the city has found itself grappling urban villages. Guangdong province, where with congestion. In response, the city has the area is located, began rebuilding urban improved its connective infrastructure, with villages in June 2000. The city of Zhuhai, 1973 marking the opening of the eight-lane for example, aims to rebuild 26 adminis- Bosporus Bridge connecting the European trative villages. An incentive scheme allows and Asian parts of the city. This was followed, the village administration, residents, and in 1988, by the completion of the second Bos- developers to share any land appreciation. porus Bridge. In 1989 the first light metro rail Rural-urban integration has been part line opened between the areas of Aksaray and of post-1978 liberalization in China. Under Kartaltepe. Meanwhile, a second light metro the township model, the urban core has line opened in 2007, supplementing the 2005 responsibility for the surrounding rural construction of an 11-kilometer metrobus hinterland. As the city grows and its area line. More are in the pipeline. of influence expands, the administra- With this prosperity has come division. tive borders of the township also expand. Much of Istanbul's rapid growth over the Large cities promote the active inclusion last several decades has occurred through of their surrounding hinterland by financ- the growth of informal settlements, such ing investment in infrastructure and social as Sultanbeyli, Sarigazi, and Pas¸aköy on services in the small cities and rural areas the Asian side. These settlements formed under their influence (see box 7.14). as rural migrants took advantage of an Get regulations right. Shanghai, a metro- ancient legal precept, which survived the politan area with special status as a province, Ataturk reforms: no matter who owns the has a population of 13 million registered land, if people are able to get their houses and 4 million permanent residents, spread built overnight and are moved in by morn- out over 6,300 square kilometers. The ing, they cannot be evicted without being urban share is almost 80 percent, with 18 taken to court. Such settlements, called urban districts and the Chongming rural gecekondu, house a large share of Istanbul's county. Urban land markets function well Concentration without Congestion 227 in allocating the urban land available under the rural land conversion limits. Floor-area BOX 7.14 Rural-urban integration in Beijing, Guangzhou, ratios have adapted to changing market and Shanghai needs and increased the space per person from 3 to 12 square meters over the past 20 Beijing, Guangzhou, and Shanghai-- meters in 2003 to 550 kilometers years. Land leases are a source of revenue. all thriving areas--have plans to link in 2005, and will be extended Shanghai's built-up area has expanded rural and urban areas: to provide again to 750 kilometers in 2010. In from 300 to 500 square kilometers in the education and health services, invest Guangzhou the provision of roads, in infrastructure and transport net- electricity, and water to all rural past decade. Passengers on the metro have works, and construct townships. settlements with more than 100 increased tenfold, from 178,000 to 1.6 mil- residents was completed in 2007. lion over the past decade, but its share of all · Social services. The governments provide vocational training and · Integrating surrounding areas. trips is just 2.5 percent. Many of the 4 mil- other services to support non- The three cities have encouraged lion transient workers in Shanghai live in old agricultural employment and help traditional industries to move urban villages, affordable because they are farmers transition from agriculture from the central business district not subject to regulations for density, height, to nonagriculture. They also o er (where rents are quite high) to the and public space. Given the difficulties in incentives for rms that will train periphery (using scal incentives) and allow high value-added indus- converting rural land, these urban villages people and recruit the trainees after try to move in to the core area. A are attractive to developers, but developing training. And they provide social services such as medical insurance township construction program them would likely make the housing condi- and pensions to rural residents. was launched to have a city system tions worse for transient workers. Beijing subsidized rural coopera- centered on an inner city of 10 Less encouraging is Mumbai. Between tive medical insurance. Shanghai million inhabitants, surrounded the 1970s and 1990s, the city resisted the increased public spending on rural by secondary cities, central towns, influx of migrants by instituting land use social services, including education and villages. In 2003 Guangzhou initiated the building of 10 central and building regulations that favored and health to cover farmers (100 towns nanced by the city govern- incumbents and prevented efficient use of percent covered by a rural collective ment. Shanghai has begun imple- land.79 The result has been an evenly spread medical insurance plan). Guangzhou will establish a pension scheme to menting the "1966 plan," which by development, but with congested streets cover all the local residents. 2020 aims to have one main city, and the proliferation of slums. nine secondary cities (traditional · Infrastructure investments. In 2005 Integrate slums into cities, using all historical centers), 60 new town- Beijing built 304 kilometers of three instruments--institutions, infra- ships, and 600 central villages with roads and linked all administrative 1,500­3,000 residents each. structure, and incentives. Cities without villages. In Shanghai expressways slums is not a realistic vision for develop- were extended from 200 kilo- Source: WDR 2009 team. ing countries, as recognized in the midterm appraisal of India's 10th Five-Year plan: developing countries. But there is little con- There has, over the years, been a paradigm shift sensus on the choice of policies required to in government's slum policy prescriptions. improve living conditions and livelihoods Originally, a "slum free cities" policy was pre- of slum dwellers, while not compromising scribed. However, looking at the social dimen- sions of the whole problem and the various the economic potential of metropolitan economic activities carried out by the slum areas. Two questions have to be answered. dwellers, this concept has given way to reha- When should slums be improved? And bilitation of slum dwellers. The rehabilitation what should be done to develop slums? This involves either relocation or in-situ develop- Report proposes that the right time to sys- ment of the slum areas. In the initial years of tematically address the problem of slums is slum development, the focus was on provision when the institutional and infrastructure of infrastructure in slums through the National requirements are in place. And the correct Slum Development Program (NSDP) and now approach is integration of slums into the there is renewed stress on provision of shelter broader urban economy. to urban slum dwellers through the Valmiki Ambedkar Awas Yojana (VAMBAY).80 If the problem is crime and squalor, the better strategy would be to upgrade the Identifying and implementing poli- neighborhood. But if the problem is spa- cies for managing slum formation is a tial inefficiency, steps to improve land use major concern for policy makers in most efficiency and compensate slum dwellers 228 WORLD DEVELOPMENT REPORT 2009 for disruptions to their livelihoods probably improving slums are unlikely to be enough should take precedence. Interventions to for social integration, unless accompanied by improve living conditions in slums include infrastructure, institutions, and complemen- prevention measures, such as sites and ser- tary reform (see box 7.15). vices programs and remedial schemes (with Land use and zoning policies have often slum upgrading being the most common), excluded the poor from being physically packages of basic services, paving, shelter, and integrated into dynamic labor markets, social integration. The Kampong Improve- while deficient transport infrastructure low- ment Program in Indonesia is probably the ers the possibility of connecting distant resi- oldest, largest, and best-known urban-up- dents to urban jobs. South African zoning grading initiative in the world. It combines policies under apartheid segregated white low investment costs of $23 to $118 per per- and black people in cities. City structure son, benefits 15 million people, and uses a can exacerbate social divisions and hinder participatory approach. The Orangi Project efforts to reduce inequality and discrimi- in Pakistan and the Accra District Reha- nation. The abolition of apartheid was not bilitation Project in Ghana are also prom- enough to reduce the disparities. To offset ising (see table 7.2). But the experiences all spatial income inequalities, local govern- show that spatially focused interventions for ments can subsidize transport costs of poor Table 7.2 Interventions to integrate the urban poor Country (city) Focus and objectives Key features and lessons Tunisia (National) Agence Tenure security. Regularizes tenure, provides Triggered dynamic process of housing improvement; helped de Rehabilitation et de infrastructure, house improvement support, plots for explain low proportion of urban population in slums. Renovation Urbaine displaced households. Brazil (Goiania) The Goiania Tenure security. Public land occupied and tenure secured Covers 100,000 former tenants. Local grassroots organization Federation for Tenants and by appealing to rights of citizens to occupy unused and successfully supported efforts to get tenure security and Posseiros untitled land. access to infrastructure and services. Peru (Ilo) Municipal Land provision. Makes serviced plots available for 6,000 lots serviced for housing by 2005. Despite a fivefold Government construction by low-income households. increase in local population in 1960­2000, no pressure for land invasions. Argentina (Buenos Aires, Land provision. Serviced land donated to facilitate Program is the result of a series of actions and initiatives San Fernando, and San resettlement and density reduction; plots allocated in over the past 20 years, supported by an Argentine NGO that Jorge) IIED-America Latina community-managed lottery. lobbies provincial and municipal authorities. Namibia (Windhoek) Flexible zoning laws. Group purchases and leases of land Demonstrates how constraints in the form of urban land Shack Dwellers Federation with communal services; plot sizes below official national use standards and regulations can be overcome to make of Namibia and City minimum. serviced sites more affordable to low-income households. Government Malawi (Lilongwe, Blantyre, Land provision; flexible regulation. Lobbied government Since 2003, approximately 760 plots for housing have been Mzuzu) Malawi Homeless for land; demonstrated capacity of members to build good provided and housing construction loans made available to People's Federation quality housing at low cost. Changing official standards savings groups; slum accommodation containment and land important for cost. use improvement. Pakistan (Orangi) Research Amenities provision. Community development of drainage 96,994 households in Orangi and 300 locations in Pakistan. and Training Institute and sewerage systems, financed by local communities All costs can be covered by eliminating contractors and and government. modifying engineering standards. 15 countries (South Asia, Amenities provision. National federations formed by slum Savings groups (mostly women) and their collective East Asia, and Africa) Slum dwellers; initiatives to build and improve homes and basic management of money allow groups to increase capacity Dwellers International services. for cooperative action; negotiation of partnerships with governments. Thailand (National) Amenities provision. Infrastructure subsidies and housing 495 projects in 957 communities covering 52,776 households. Community Organizations loans to community organizations formed by low-income Activities identified by each community organization in Development Institute slum households. partnership with local actors; funding sources include community contributions. Nicaragua (National) Local Amenities provision. Cofinance small infrastructure 484 projects benefiting some 60,000 households. Funds Development Program projects (water, sanitation, drainage); house improvement provided to local governments, NGOs, community (PRODEL) and microenterprise loans and support. organizations, and households. Source: Sattherthwaite 2008, for this Report. Note: IIED = International Institute for Environment and Development; NGO = nongovernmental organization. Concentration without Congestion 229 children, provide private-school vouchers, and increase public-school spending.81 BOX 7.15 Slum upgrading and prevention: what works? Evidence from policy experience com- performance monitoring mecha- A strategy for inclusive urbanization piled by UN-HABITAT and the Cities nisms that require the involve- The Tinbergen principle proposes that one Alliance shows that successful initia- ment of all levels of government. policy instrument is needed to address each tives share several attributes. Among Cambodia, China, and Vietnam, policy objective.82 Applying the principle to them, institutional strengthening and for example, have strict upward the policy issues addressed in this Report coordination across government lev- accountability regarding municipal els seem to be the most important. implementation on infrastructure. implies that as many integration instru- Stronger institutions. Countries Brazil and Indonesia, on the other ments are needed as there are dimensions that have been successful in inte- hand, have bottom-up perfor- to a problem. As the integration challenges grating slums into their cities have mance monitoring, which enhances increase with the stage of urbanization, the strengthened their institutions and citizen participation in planning number of policy instruments required carried out complementary reforms, and decision making.b Coordination increases as well. Fortunately for develop- which include a broader urban pov- across government levels and with ing nations, the capacity of markets and erty reduction agenda (Indonesia, the private sector is also critical Islamic Republic of Iran, Mexico, South for successful scaling up of slum governments grows as they urbanize. But Africa, and Turkey). Some have imple- upgrading projects. One example these policies must be introduced in the mented policies to integrate the urban is Indonesia's Kampong Improve- right sequence. poor into the legal and social fabric ment Program, and there are oth- The foundations for an inclusive urban- of cities (Brazil, Chile, and Colombia), ers (for example, Brazil, Colombia, ization have to be instituted early in the others have carried out reforms in land Mexico, South Africa, Thailand, and development process. To do this, govern- and housing provision (India). Tunisia) whose programs began on ments must be selective. This chapter sug- Coordination across govern- a modest scale, and were success- ment levels and with private fully scaled up to the national level gest how they can prioritize and sequence: agents. Countries that performed because of the involvement of all · In areas of incipient urbanization, the well also made an effort to coordi- levels of the government and the objective should be to facilitate a natu- nate among central, regional, and private sector. ral rural-urban transformation. The core local authorities and the private sector (Chile, Egypt, Sri Lanka, Thai- policy instruments are spatially blind Based on a contribution by Eduardo land, and Tunisia). But cities and institutions that facilitate density in some López Moreno, chief, Global Urban countries that were successful in Observatory, UN-HABITAT. locations. These instruments include the delivery of basic services and a. Garau 2008. secure land tenure and property rights, housing improvements had clear b. Bazoglu 2008. basic and social services, and macro- economic policies that do not favor one productive activity (large industry) over another (small agriculture). Policy mak- ers should aim for neutrality between Efficiency should be the watchword of rural and urban areas. policy makers. · In areas of intermediate urbanization, ·In areas with advanced urbanization, the rapid growth of some cities cre- divisions within cities caused by formal ates congestion. In addition to spatially settlements and slums and by grime and blind policies to facilitate density, con- crime add to the challenges of density nective policies to tackle congestion and distance. In addition to spatially and economic distance become nec- blind and spatially connective policies, essary. They include investments in spatially focused policies for address- transport infrastructure (to enhance ing intracity divisions are necessary to connectivity both within and between target the difficulties of slums, crime, cities) and encouragement of socially and the environment--and to improve efficient location decisions by firms. livability. CHAPTER 8 Unity, Not Uniformity Effective approaches to territorial development G lobalization and liberalization When lagging areas have few people and may rearrange production within a small share of the country's poor, mea- countries, leaving people concen- sures to enhance migration should be the trated in places no longer favored by mar- mainstay of development policy. kets. In Brazil, China, India, and the Russian ·In countries where lagging areas have Federation, changes in both markets and large numbers of the poor, but few policies over the last century have altered the impediments to their mobility, insti- fortunes of places. Geographic differences tutions that promote mobility should in economic activities encourage migration be augmented by spatially connec- from lagging areas, concentrating people-- tive infrastructure. Some countries including the poor--in leading areas. But have high population densities in lag- geographic unevenness in living standards, ging areas--and large numbers of the by creating or deepening divisions within poor--but few cultural, linguistic, or countries, can also lead to conflict, slowing political impediments to labor and capi- social and economic development. tal flows. Investments in infrastructure Building on the findings and analysis in that increase the flow of goods, people, earlier chapters, this chapter discusses the and information would aid economic policy responses to widening or persistent concentration and spatial convergence differences in living standards between in living standards. areas of a country that markets favor with ·In countries fragmented by linguistic, greater economic mass and those that they political, religious, or ethnic divisions, do not. As in the rest of this report, the term spatially targeted interventions may be "area" is synonymous with a subnational needed. When lagging areas face the triple region or territory, so this chapter deals with challenge of long distances to economic the "regional development" (also referred opportunities in leading areas, large to as the "territorial development") debate. population densities, and large numbers Logic and experience indicate that policy of poor people, as well as domestic divi- makers should calibrate their responses to sions that limit the movement of labor the severity of the challenge. and capital, institutions and infrastruc- · In countries where labor and capital ture investments could be supplemented are mobile, economic distance between by targeted incentives to encourage eco- lagging and leading areas should be nomic production in lagging areas. But addressed mainly with spatially blind these incentives should not run coun- or universal policies, for which the term ter to the integration objectives pursued "institutions" is used as shorthand. through institutions that bring people These policies should make it easier for together and infrastructure that connects 230 people to move toward opportunity. lagging and leading areas of a nation. Unity, Not Uniformity 231 Institutions, infrastructure, and incen- In many countries, the decentralization tives--these are the three parts of a success- of administrative and fiscal responsibili- ful policy approach to domestic integration. ties has increased the role of subnational In deciding among the integration options, governments in the design and delivery of governments have to consider the fiscal and policies. Resources allocated to subnational opportunity costs of these instruments. This governments should come with agreements chapter provides an organizing frame for to ensure that local initiatives improve governments to think through these options national welfare along with local welfare and find the best combination of policies. (see box 8.1). Anewapproach. Policy discussions about In addressing these policy issues, the how to improve welfare in lagging areas chapter provides an answer to a question often begin with a focus on lagging areas-- of considerable concern to policy makers: and an emphasis on targeted interventions Should countries invest in people or in or policy "incentives" to move production places? The answer is to invest in activi- to these places. Instead, territorial develop- ties that produce the highest economic and ment policies should integrate lagging with social returns nationally. In leading areas, leading areas, and the discussion of spatially emphasize investment in places--durable targeted incentives should come last--after investments that increase national eco- considering spatially blind policies such as nomic growth. In lagging areas, emphasize national revenue-sharing and social expen- investment in people--portable invest- diture arrangements and spatially con- ments that stimulate mobility and acceler- necting initiatives, such as transport and ate poverty reduction. communication systems. The experience of developed and developing countries shows People seek opportunities that without these supporting institutions Throughout history, people have moved and infrastructure, incentives have been from places with harsh geography to those unsuccessful and expensive. offering a more pleasant climate and better BOX 8.1 Are the policy messages of this Report "anti-decentralization"? No. Spatially blind institutions are the bed- policy makers may be inclined to tilt on public services) but should move as rock of economic integration policies expenditures toward politically popular soon as feasible to outcomes (improve- seeking spatial e ciency and equity. activities. In a highly decentralized coun- ments in the national Millennium Regardless of where people live, they try such as Brazil, progress on national Development Indicators). should have a ordable access to basic priorities of eliminating illiteracy and · Infrastructure--to maximize synergies services such as primary health care, universalizing primary education is moni- from infrastructure investments and education, sanitation, and security. How tored using expenditure allocations, with to regulate interstate commerce e - these services are delivered depends on the constitution determining that 25 ciently, design and planning decisions country circumstances. percent of state and municipal revenues should be jointly made by subnational Decentralization in many countries has from taxes and transfers be earmarked governments a ected by these invest- made subnational governments more to nance primary education. But closer ments and regulations. responsible for improving local welfare inspection shows that around 10 percent · Incentives--while subnational govern- outcomes. In Vietnam subnational gov- of municipalities spend less than the con- ments may be well suited to assess ernments were responsible for almost 50 stitutionally recommended amounts.b local economic potential, the decision percent of public spending in 2002, up Large and visible investments are politi- of where to target incentives should from about 25 percent in 1992. In China cally expedient signals to voters that their lie with a national authority that can the ratio climbed from 67 to 72 percent representatives are hard at work. So how prioritize resources to accelerate overall between 1990 and 2004.a In the Philip- can decentralization be consistent with growth. How these programs are imple- pines the ratio was about 25 percent in the spatially integrative policies discussed mented can be decided locally. 2002, up from 11 percent in 1990. here? Problems arise when decentralization Source: WDR 2009 team. gets in the way of delivering spatially · Institutions--resource allocations to a. National Bureau of Statistics, China, 2005. blind policies. Rather than allocating subnational governments could be b. Instituto Brasileiro de Geogra a e Estatís- resources to social services, subnational based initially on inputs (expenditures tica (IBGE) 2004. 232 WORLD DEVELOPMENT REPORT 2009 Map 8.1 The poverty rate is high in China's western interior, but most poor people live closer to economic density in the East Poverty rate: proportion of poor (%) Poverty density < 17.5 17.6­35.9 Each dot represents 36.0­51.6 50,000 poor persons 51.7­70.3 70.4­81.1 Source: WDR 2009 team. Note: Poverty rates and counts are estimated for a $2/day poverty line in 2002. economic opportunities. The concentration country's poor. The numbers on the other of people in areas with hospitable natural islands are much lower: 1.3 million in Kali- environments attracted economic activities mantan, 2.6 million in Sulawesi, 2.7 million to these places, helping many to prosper even in Nusa Tenggara, and 1 million in Papua. when the initial conditions that made these Chapter 2 showed that Honduras and Viet- settlements economically attractive became nam have similar (overlapping) geographic less important. Mobility was not just for the distributions of economic production and well-off. Poor people also moved to econom- poor people. ically dense areas--to seek better lives. Using a sharper geographic resolution, a Consider the current distribution of similar distribution of the population can be poor people in China (see map 8.1). The discerned within lagging areas. The north- percentage of people living below $2 a day is east of Brazil is the country's poorest area. high in the country's lagging western areas. Per capita incomes in the southeast were But in absolute terms, many more poor 2.9 times that of the northeast in 1939, and people live in the dynamic coastal south- 2.8 times in 1992. Eight of the 10 poorest eastern areas--the leaders in China's rapid states are in the northeast, two in the north.1 integration with the global economy. Even The poverty rate is clearly high in the rural before this integration happened, people northeast and Amazon areas (see map 8.2). did not concentrate in places with inhospi- But even in the northeast, the mass and table geography, such as the Qinghai-Tibet concentration of poverty--the number of Plateau with an elevation of 4,000 meters poor people per square kilometer--is much above sea level, or the highlands of the cen- higher in urbanized agglomerations near the tral region with elevations of 2,000 meters. coast, from the lagging northeast all the way The flat lands and warmer climates along to the dynamic regions of Rio de Janeiro and the coast provided better conditions for São Paulo in the southeast.2 farming and trade. In some countries the market forces Maps of many countries would also prompting factor mobility are not quite as show that the poverty mass--the numbers strong. Their economic mass and poverty of poor people--and economic mass coin- mass do not coincide nearly as much as in cide. Java, the economically leading area in countries such as China and Honduras. Indonesia, is also the island in the archi- Consider India, where more than 400 mil- pelago where most poor people live. The lion people live in "lagging states" in the islands of Java and Bali are home to 21 mil- north-central part of the country, which lion poor people, about 58 percent of the includes--using the country's poverty Unity, Not Uniformity 233 Map 8.2 The poverty rate is high in Brazil's Northeast and Amazon areas, but the poor are massed in areas along the coast Poverty rate: Poverty density proportion of poor (%) < 12 Each dot represents 12­25 5,000 poor persons 25­35 35­45 >45 Source: State-level poverty rates computed by Phillippe G. Leite, based on the Brazil Household Budget Survey (Pesquisa de Orcamentos Familiares) 2002­03 (see World Bank 2007c). line--60 percent of the nation's poor (see better access to domestic and international map 8.3).3 Labor mobility from these areas markets.6 has been limited due to ethnolinguistic and Should today's policy makers try to class-based divisions, perhaps inclining correct historical accidents by reviving people to stay in their own "enclaves." investments in lagging areas and helping People are in these lagging areas for a them regain their past glory? Or should reason. With rich soils and good internal they accelerate India's integration with the connections, the fertile Indo-Gangetic global economy and help people in lagging plain attracted people. These areas were areas take advantage of new opportunities historically among India's most dynamic in places with greater economic density? locations when the country accounted for India's national policy discussion in a quarter of the world's gross domestic the mid-1970s focused on promoting spa- product (GDP), between 1600 and 1700.4 tially balanced growth to revive the lagging But historical "accidents," such as making areas with subsidized finance, investment landlords responsible for paying land taxes subsidies, industrial infrastructure, and to the British crown in these areas, eroded preferential industrial licensing. India's lat- agricultural productivity and infrastruc- est five-year plan recognizes the failures of ture investments.5 India is now reinte- industrial licensing and its inconsistency grating into the world economy, resulting with growth. The discussion today is more in economic dynamism in its coastal and about integration and the shift toward metropolitan areas. These areas offer good promoting better health and education in access to intermediate inputs and domes- lagging areas, along with strategic inter- tic and international markets, provide reli- regional infrastructure investments that able and high-quality local public services, connect the remote northeast with markets and have a business environment condu- in the rest of the country.7 Similar regional cive to entrepreneurship. About half of development policies have been imple- manufacturing investments in 2005 were mented in other countries with internal concentrated in only 10 of the country's divisions along religious or ethnic lines, more than 3,000 cities, the ones offering such as Nigeria and Ethiopia. 234 WORLD DEVELOPMENT REPORT 2009 Map 8.3 Both the poverty rate and poverty mass are high in some of India's lagging states Poverty rate: proportion of poor (%) Poverty density 6.4­9.7 9.8­16.7 Each dot represents 16.8­24.8 50,000 poor persons 24.9­35.4 35.5­46.6 No data Source: Based on poverty estimates for 2004­05 from the Planning Commission, Government of India. Note: State-level poverty rates and counts are based on the sum of the number of urban poor and rural poor in 2004­05, which are calculated using different poverty lines. Policies aimed at reducing such divi- leading areas, and also because of divisions sions have reinforced a natural tendency of from political, ethnic, religious, and linguis- people to seek places that offer better eco- tic differences. And all countries seek unity, nomic opportunities. Take the well-studied by lowering the barriers of internal divisions. German unification, a merger of two econ- A review of national constitutions from 20 omies with few exchanges of goods and developing countries shows that promoting factors that propelled people from the east unity--reducing divisions--is an important to the west. The opportunity arose in the political objective (see box 8.2). In Nigeria summer of 1989, when people could leave an important article of unity is that people East Germany through Hungary. And with will not be discriminated by sex, religion, the fall of the Berlin wall on November 9, place of birth, ethnic, or linguistic associa- 1989, direct migration from East to West tion. In India Article 16 of the constitution Germany became possible. With the bor- states that "no citizen shall, on grounds der open, 800,000 people left the east for only of religion, race, caste, sex, descent, the west in 1989 and 1990, 5 percent of the place of birth, residence or any of them, be eastern population (see box 8.2).8 Clearly, ineligible for, or discriminated against in the German unification started a move to respect of, any employment or office under density. It suggests that reducing distance the State." For the most part, constitutions to economic density improves people's wel- do not make places paramount--they focus fare, and labor mobility is the strongest instead on the welfare and unity of people natural mechanism for this. (see box 8.3). Unity does not mean uniformity. India's Countries seek unity national motto, for example, is "unity in Many countries have spatial differences in diversity." But in many countries, policy production and poverty, mostly because makers have viewed uniformity as the main of economic distance between lagging and vehicle for unity. The European Union (EU) BOX 8.2 The German integration: convergence and concentration with mobile labor With the fall of the Berlin wall in 1989, Migration from East to West Germany was Convergence of income across German direct migration from East to West Ger- possible after the fall of the Berlin wall counties was noticeable between 1992 many became possible. Opening the bor- Net migration from East to West Germany, and 2005 1991­2005 der led 400,000 people to leave the east to Density the west in 1989 and again in 1990, about 5 Thousands of people 2005 percent of the eastern population (see the 180 1.2 left gure). In later years, incomes started 160 1.0 to converge and the process slowed. In 140 2001 a recession in Germany again led 0.8 120 about 100,000 people to leave the east for 100 0.6 the west. By 2007 more than 1.7 million 1992 80 people had left the east (of about 17 million 0.4 60 at the time of the fall of the Berlin wall). 0.2 Migration produced one predictable 40 0 outcome: incomes became more equal 20 1.5 2.0 2.5 3.0 3.5 4.0 4.5 between the two areas (see the right g- 0 1991 1993 1995 1997 1999 2001 2003 2005 Log, (thousands) ure). While the cross-country distribution Year Source: Holzmann, Priebe, and Vollmer 2008. of income in 1992 was clearly bimodal, Source: Federal Statistical Office Germany 2007. with the counties in the east forming the lower peak, this was smoothed by 2005. institutions of higher learning (see the map for them to study or to nd a job in the Although the counties from the east are below). In 2004, in the 18­29 age-group, economically more dynamic parts of still located at the left of the distribution, there were only 90 women for every 100 Germany. Second, it is much harder for their economic distance to western coun- men in East Germany (including Berlin). women to nd an attractive job locally ties has shortened. Why did East German women move to in the lagging areas than it is for men, Convergence in incomes has also pro- the economically dynamic areas, while because these areas are typically domi- duced more surprising outcomes. Almost men stay in the lagging areas? There are nated by traditionally male jobs in agri- all counties with more young women than two explanations. First, women are on culture, manufacturing, and construction. men are in the economically dynamic average more successful in school and Sources: WDR 2009 team, based on Kroeh- areas of Germany, which also have better higher education, which makes it easier nert and Vollmer 2008. Women moved to the west seeking economic opportunity: in 2003, there were fewer women per 100 men in the eastern part of the country (left), and higher taxpayer incomes in the west (right) Number of women, Average income ages 18­29, per 100 men of taxpayers GERMANY 75­85 Top 20% 86­95 96­100 101­105 106­115 Bottom 20% 116­126 Source: Kroehnert, Medicus, and Kinglolz 2006. Source: Federal Statistics Office of Germany. 235 236 WORLD DEVELOPMENT REPORT 2009 BOX 8.3 Constitutions call for unity, not uniformity National constitutions re ect the evolu- hibited." Article 15 (3)--"For the purpose Brazil's constitution calls for regionally tion of political discourse in most coun- of promoting national integration, it shall balanced economic development, setting tries and de ne interactions between be the duty of the State to: (a) provide out guidelines to promote capital ows the state and its citizens. A review of adequate facilities for and encourage free to lagging areas. The translation of these constitutions from 20 developing coun- mobility of people, goods and services guidelines into practice has not produced triesa nds that all emphasize national throughout the Federation; (b) secure the expected economic gains. And the unity as an important national objective. full residence rights for every citizen in all programs have been costly. The Constitu- Uniformity in economic outcomes across parts of the Federation." tion Funds--a prominent regional eco- areas is mentioned only in one--Brazil's Côte d'Ivoire: Article 30--"The Republic nomic development program--provided constitution. Universal access to primary of Côte d'Ivoire shall be one and indivisible, subsidized credit worth more than $10 education and basic health care are con- secular, democratic and social. The Repub- billion between 1990 and 2002 to help stitutional mandates in most countries. lic shall ensure equality before the law to all rms locate in lagging areas. The ine ec- Excerpts from the constitutional arti- without distinction as to origin, race, sex or tiveness of these interventions is evalu- cles of three African countries highlight religion. It shall respect all beliefs." Its prin- ated elsewhere in this chapter. the importance that societies place on ciple shall be: "Government of the people, national unity and integration: by the people and for the people." Source: WDR 2009 team; based on a review of Nigeria: "The motto of the Federal Uganda: Article 3--"(i) All organs of the constitutions from 20 developing countries. Republic of Nigeria shall be Unity and State and people of Uganda shall work a. The countries are Argentina, Bolivia, Faith, Peace and Progress." Article 15 towards the promotion of national unity, Brazil, Colombia, and Mexico (LAC); Côte (2)--"Accordingly, national integration peace and stability. (ii) Every e ort shall d'Ivoire, Ghana, Nigeria, and Uganda (AFR); Bangladesh, India, Nepal, and Pakistan shall be actively encouraged, whilst be made to integrate all the peoples of (SAR); Kazakhstan, Poland, and the Russian discrimination on the grounds of place Uganda while at the same time recogniz- Federation (ECA); China, Indonesia, and the of birth, sex, religion, status, ethnic or ing the existence of their ethnic, religious, Philippines (EAP); and the Arab Republic of linguistic association or ties shall be pro- ideological, political and cultural diversity." Egypt (MENA). policies to integrate new member states call and training systems.11 The cohesion policy for "cohesion." But the objective of cohe- aims to improve economic performance of sion--or unity--is pursued through poli- specific areas and help them catch up with cies for convergence. Convergence targets the rest of the union. include eliminating territorial disparities in Resource allocations of this scale to sup- economic development (economic cohesion) port integration may reflect the redistribution and in access to labor and income (social preferences of member states (particularly the cohesion). This "cohesion" is matched by the ones whose residents are footing the bill), but EU regional policy, which allocates about 60 do these policies stimulate overall growth? percent of its funding to support areas of low Are they paying for the "wrong" type of development (less than 75 percent of the EU assets? Academic research shows that they are average GDP per capita).9 not well suited to maximizing aggregate eco- Accounting for 35 percent of total spend- nomic growth because they try to promote ing of the EU, the EU cohesion policy is trans- spatial evenness and not agglomeration. Nor lated into practice through structural funds are they especially well suited to promoting (90 percent of spending) and the cohesion catch-up by lagging areas. Traditional cohe- fund (10 percent). The Agenda 2000 package sion policies that provide "hard" infrastruc- comes with a price tag of 236 billion, with ture and assistance to firms are unlikely 195 billion for structural funds; 18 billion to increase the competitiveness of lagging for the cohesion funds for Greece, Ireland, areas.12 Moving away from these programs to Portugal, and Spain; and 22 billion for new supporteducationprogramsandinstitutional member states, in view of their 2004 acces- development could do much more. sion.10 Agenda 2000's objectives include the Trying to use the same instrument to development and structural adjustment of pursue dual challenges of internal and lagging areas, the development of border external convergence is likely to make the areas and areas in industrial decline, and the policy lose focus. The EU's fourth report adaptation and modernization of education on economic and social cohesion provides Unity, Not Uniformity 237 a candid assessment of convergence across per capita incomes. But the concentration its regions and within individual coun- of economic production within member tries.13 Between 1995 and 2004, there was states--new and old--has been increas- a tendency of aggregate convergence in the ing, led by market forces driving faster eco- EU, with new member states with lower nomic growth in their leading areas. Indeed GDP per capita growing faster than the per capita incomes in several areas in some EU-27 average. But at a more disaggre- new member states--Bratislavský kraj in gated Nomenclature of Territorial Units for the Slovak Republic, Közép-Magyarország Statistics (NUTS) 2 regional level--sub- in Hungary, Mazoweickie in Poland, and national areas larger than administrative Zahodna Slovenija in Slovenia--have risen units in most countries--the results show to more than 75 percent of the EU average. little effectiveness of directed interven- Ireland took a different approach for tions in improving economic performance using EU funds. Rather than try to use the of lagging areas. For the better-off EU-15, EU funds to achieve both international the number of people in lagging regions catch-up and to disperse economic produc- has remained almost unchanged at 32 mil- tion domestically, Ireland focused on one lion, around 8 percent of the total. When objective--national economic growth. From the new member states are considered, being one of Europe's poorest countries, it is there has been international convergence in now one of the richest (see box 8.4). Between BOX 8.4 An instrument per objective: Ireland used EU funds for international convergence Between 1977 and 2000, Ireland's GDP nied by a rising spatial concentration of eco- Today, almost all regions in the new per capita grew from 72 percent of the EU nomic activity. Compared with the other member nations in Eastern Europe qual- average to 116 percent. What was behind cohesion countries--Greece, Portugal, and ify for EU nancial support. They should Ireland's success? Spain--Ireland's economic concentration consider the Irish example of using the Since joining the European Union in rose much more (see the gure below). But funds for international convergence and 1973, Ireland received approximately its per capita income grew much faster too. not--until later stages--for spatially 17 billion in EU Structural and Cohesion In 1977 Greece, Ireland, and Spain had per balanced economic growth within their Funds through the end of 2003. In the rst capita incomes of about $9,000; Portugal's borders. two rounds of EU funding, the entire coun- was about $6,000. By 2002 Portugal had an try was classi ed as an Objective One area. income of $11,000, and Greece and Spain Sources: Dall'Erba 2003, WDR 2009 team. Between 1993 and 2003 cohesion funds close to $15,000. Ireland's per capita income a. Ireland's National Development Plan supported 120 infrastructure projects at had risen to $27,500. (NDP) 2000­06. the cost of about 2 billion.a The choice of projects was based on a national develop- ment plan, which focused on investments in economic infrastructure that stimulated As Ireland's income rapidly grew, economic concentration increased within the country long-term national economic growth. Natural log of variance These included investments in leading 17.0 areas and in connecting leading and lag- Ireland ging areas, such as the M50 (Dublin Ring 16.5 Road), M1 (Dublin-Belfast), and improve- 16.0 ments in the N4 (Dublin-Sligo), N7 (Dublin- Spain 15.5 Limerick), and N11 (Dublin-Rosslare). The Irish also invested in education, 15.0 Portugal training, and lifelong learning in all of 14.5 Ireland to provide investors with a good business environment countrywide. With 14.0 its skilled labor force and good logistics, 13.5 Greece Ireland has become a popular destination 13.0 for American rms wishing to reach Euro- pean markets. In 2004 Irish-based U.S. 12.5 rms exported $55 billion worth of goods 12.0 and services, mostly destined for Europe. 1977 1979 1981 1983 1985 1987 1989 1991 1993 1995 1997 1999 2001 Ireland's rapid convergence toward the Year incomes of Europe's leaders was accompa- Sources: Dall'Erba 2003; WDR 2009 team. 238 WORLD DEVELOPMENT REPORT 2009 1991 and 1998 Ireland was just one region for places. Indonesia's transmigration program the purposes of Objective One support from tried to relocate people from densely popu- the EU.14 When the country grew past the lated Java to less densely populated areas of 75 percent threshold, in July 1999, Ireland Kalimantan, Papua, Sulawesi, and Sumatra. created two regional "assemblies"--Border- At its peak between 1979 and 1984, 535,000 Midland-Western, and the Southern and families or almost 2.5 million people were Eastern. But spatial concentration of eco- relocated. The objective was to promote nomic production has increased in Ireland "balanced demographic development" and relative to Greece, Portugal, and Spain, the reduce poverty by providing land and new other three cohesion countries.15 economic opportunities for poor landless settlers. But the program made almost no A policy framework for integrating dent in the population density of Java, nor lagging and leading areas did the high cost of the program reduce People seek opportunity, and countries poverty much among the migrants.16 seek unity. Policies that integrate lagging Relying solely on spatially targeted inter- and leading areas can help with both. This ventions is a common mistake. It is far bet- section outlines a framework to guide ter to rely on institutions that work less policy making. It proposes a calibrated noisily. In France the concentration of eco- combination of institutions, infrastruc- nomic mass and convergence of disposable ture, and incentives to address the domes- incomes between leading and lagging areas tic challenges posed by density, distance, have been concurrent, producing a "scissors and division. Used well, a combination of effect" in the geographies of production and these measures can help countries reap the disposable income (see figure 8.1).17 The economic benefits from increasing con- effect appears to be driven not by spatially centration of economic activity, as well as focused interventions but by spatially blind the social, political, and economic payoffs or "universal" progressive income taxation, associated with converging living standards social security, and unemployment benefits. between lagging and leading areas. Although space is not explicitly considered Policy makers, often viewing economic in such policies, their effects and outcomes concentration as inconsistent with spatial can vary considerably across locations. equity in living standards, have sought to As the base of economic integration, such reduce concentration through spatially "institutions" capture the benefits of spatial targeted interventions. Many governments concentration of production and deliver fight market forces that promote the con- convergence in living standards. centration of people in economically dense Even in the EU as a whole, the rising inequality of market incomes between 1985 Figure 8.1 France has benefited from increasing concentration of economic production and and 1995 was partially offset by progressive declining spatial disparities in disposable income tax and transfer policies. Increases in the Coefficients of variation income of skilled people were moderated GDP per capita Income per capita through higher taxes, and the unskilled were 0.190 0.085 aided with transfers.18 Similarly, the pro- Per capita disposable income (right axis) 0.185 0.080 gressive tax structure in the United States 0.180 reduced disparities in disposable incomes 0.075 across states, while production became 0.175 0.070 more concentrated, although the extent has 0.170 varied greatly over time as government poli- 0.065 0.165 cies changed (see box 8.5). 0.060 0.160 The experience of the EU and the United GDP per capita (left axis) States in addressing spatial equity with 0.155 0.055 aspatial tax systems is instructive. Skep- 0.150 0.050 1982 1984 1986 1988 1990 1992 1994 1996 1998 2000 2002 tics might counter that the coverage of Years the tax system is low in developing coun- Source: Martin 2005. tries and that weak tax administration and Unity, Not Uniformity 239 BOX 8.5 Taxation against spatial inequality? The U.S. federal income tax system A progressive federal income tax in tax income ratios for two groups of 10 1). When Congress introduced payroll the United States has reduced income states are calculated: one recording the withholding and quarterly tax payments inequalities among people. An unin- highest pretax incomes (Group 1), and during World War II, the progressive tended e ect has been to reduce income the other the lowest (Group 2).a The g- tax system was reinforced. But these inequalities across states, showing that a ure below shows the change in post-tax inequality-reducing e ects started fading spatially blind policy can be a sharp instru- income di erences between the richest away between 1950 and 1970. A brief rise ment for reducing spatial inequalities. and poorest states. Directly imputable in these e ects during the late 1970s was Data from the Internal Revenue Service to taxation, it shows how the U.S. taxa- followed by a fall in the 1980s and 1990s. show how much the income tax reduced tion system has helped to reduce income spatial inequality. To see this, rst, pretax inequalities across states. Source: WDR 2009 team. incomes of the top percentile of earners Never had the tax rate been higher a. Although both groups belong to the rich- in each state are divided by the U.S. per- than in 1918, at 77 percent to nance the est 1 percent of their respective states, keep sonal income, as published by the Bureau war. After World War I, tax rates declined. in mind the sharp di erences from one state to another. In 1940, for instance, an income of Economic Analysis. Next, the same Lowered to 24 percent in 1929, the tax of about $47,000 made a tax ler part of the income dispersion ratio is calculated with rate for top incomes rose again during wealthiest 1 percent in Mississippi, but part post-tax incomes. Then, the percentage the Great Depression (­26 percent in of the wealthiest 10 percent in the District of decrease between the pretax and post- 1940 versus ­7 percent in 1930 for Group Columbia. Without an explicit spatial focus, U.S. federal income taxes reduce spatial disparities Percentage decrease in inequality due to taxation 0 ­5 ­10 ­15 The 10 richest states converge ­20 to the mean 3 times faster than ­25 the 10 poorest ones. Increasing tax rates ­30 during Great Depression ­35 Government increases Bridging the gap after WWII taxes for the bottom group ­40 1920 1930 1940 1950 1960 1970 1981 1990 2000 2003 Year 10 U.S. states with highest income levels before taxes (within top percentile) 10 U.S. states with lowest income levels before taxes (within top percentile) Source: WDR 2009 team, based on analysis by Estelle Sommelier. widespread informality in the economy will the automobile industry in 1925.20 The tax reduce the potential benefits of progressive coverage and revenues were low, with only taxes. But it is worth remembering that a 10 percent of the households filing an indi- century ago, when its tax system was intro- vidual tax return in 1916; today the ratio is duced, the United States exhibited many 93 percent. of the attributes of a developing country Some evidence from developing coun- today. Its population was mainly rural, tries points to the fact that income redis- with just 28 percent living in metropolitan tribution through a progressive tax areas in 1910.19 Wages were $0.72 an hour in system is difficult, and that the targeting 240 WORLD DEVELOPMENT REPORT 2009 of expenditures by individual or household tion regulations, local infrastructure income levels and the level of the average development, and targeted investment tax rate are more important for the post- climate reforms, such as special regula- transfer income distribution. Thus a high- tions for export processing zones. yield proportional tax system may have a larger equalizing impact than a low-yield These instruments for integration-- progressive tax system.21 institutions, infrastructure, and incen- Spatially blind tax and transfer policies tives--span the range between universal will form the bedrock of public policies to and geographically targeted policies. Each integrate countries spatially and help them of the three categories can include taxes, benefit from concentration and conver- public spending, and regulations. gence. But these policies may not be enough. Adverse physical geography generally Depending on their conditions, nations need increases economic distance, reducing a broader range of instruments for domestic trade of goods and services and the flow integration. The challenge of economic inte- of labor, capital, and information, making gration can be seen as reducing the distance delivery of public services harder. In Papua between people--especially the poor--and New Guinea, with the transport system economic opportunity. Misplaced popula- fragmented by a rugged mountainous ter- tion densities, and barriers to mobility of rain, the average travel time from a rural workers and entrepreneurs between leading community to the nearest road is two and and lagging areas posed by internal divi- a half hours. and to the nearest govern- sions, complicate the challenge. ment station is more than three hours.22 In general, a policy framework for eco- In the more rugged parts of Peru, the nomic integration includes the following: coverage of public infrastructure is low.23 Other such places include Chile's Zonas · Institutions (spatially blind policies). The Extremas, western China, Upper Egypt, term is used here to categorize policies the outer areas of Nepal, and northeast- that are not explicitly designed with spa- ern Russia. Because of adverse conditions, tial considerations, but that have effects poverty rates can be high in these areas. and outcomes that may vary across But for the same reasons, unless prevented locations. These include such national from leaving by government policies or policies as the income tax system, inter- sociopolitical reasons, or enticed into stay- governmental fiscal relations, and gov- ing by incentives, not many people live in ernance of land and housing markets, these areas. as well as education, health care, basic Integration reduces the economic dis- water and sanitation, and other govern- tance between lagging areas and more ment initiatives. dynamic places. The most successful ini- · Infrastructure (spatially connective poli- tiatives, which balance economic efficiency cies). The term is used here as shorthand and political feasibility, are adapted to to include all investments that connect country circumstances. The circumstances places and provide basic business ser- that matter most are the population den- vices, such as public transportation and sities in lagging areas and the extent to utilities. These include developing inter- which domestic divisions weaken market regional highways and railroads to pro- forces. Where few people live in lagging mote trade in goods--and improving areas, as in northeastern Russia, integra- information and communication tech- tion policies should be different from nologies to increase the flow of informa- those in places such as northeastern Brazil, tion and ideas. where lagging areas are densely populated. · Incentives (spatially focused policies). Where lagging and leading areas share a The term is used here to include spatially common language and customs, as in Bra- targeted measures to stimulate economic zil and China, integration policies have to growth in lagging areas. These include exert less effort than in areas where dif- investment subsidies, tax rebates, loca- ferences in language, ethnicity, or religion Unity, Not Uniformity 241 divide one part of a country from another, them into one of these types. In Thailand as in India or Nigeria. the northeast is densely populated and dis- In Brazil the distance between the lag- tant from the economically dense capital ging northeast and the leading southeast area, but the sparsely populated south is is coupled with high population densities home to its Muslim minority. In India the in the coastal areas of the northeast. But lagging northeast is sparsely populated, many Nordestinos have found opportuni- whereas lagging areas in central India have ties by moving to the dynamic southeast. almost two-thirds of India's poor. Sec- As many as 4 million residents of Greater ond, countries classified as Type 1 (with São Paulo are Nordestinos.24 This indicates sparsely populated lagging areas) can be the high population density in the north- unified or internally divided. But the strat- east and the strong market forces of labor egies for integration in these two types of mobility, made possible by factors such as countries--unified countries with sparsely a common language and a strong national populated lagging areas such as Russia, or identity. divided countries with sparsely populated Recall the maps of India, where some lagging areas such as the Philippines--will lagging areas have a high poverty mass and not be different. high poverty rate (see map 8.3). Integrating these areas is especially challenging when An instrument per dimension subnational geographic groupings reflect Institutions to overcome distance. In ethnic, linguistic, or social differences.25 countries with sparsely populated lagging The movement of people out of these areas areas, the integration challenge is mainly has been limited because of local prefer- one of reducing economic distance. Poli- ences and discrimination against particu- cies that are universal--spatially blind in lar groups (see chapter 5). Market forces their design and national in their cover- of factor mobility have been weakened by age--can shoulder much of the task of suc- internal political and social divisions--wit- cessful economic integration. Developing ness the hostility that Bihari workers have countries in this category include Chile, encountered in the more prosperous parts China, Ghana, Indonesia, Kenya, Mexico, of India. In such nations, the integration Mongolia, Kazakhstan, Russia, Sri Lanka, challenge involves overcoming economic Uganda, and Vietnam. The primary objec- distance, misplaced density, and domestic tive of these policies should be to encour- division. age people migrate to places with economic Using the spatial dimensions of dis- opportunities. For example, as Russia tance, density, and division to characterize moves further from plan to market, it will conditions in a country, a suggestive tax- have to offset a legacy of policies industrial- onomy can be developed to help countries izing its vast territory. Even today, millions tailor integration policies to their specific of people are subsidized to live in "cold" economic geography. At least three types of and isolated places in the northeast (see countries can be distinguished: map 8.4), where they cannot take advan- tage of new economic opportunities in the · Type 1: countries with sparsely populated dynamic areas of the west.26 lagging areas Correcting land market distortions, · Type 2: unified countries with densely removing restrictions on mobility, and pro- populated lagging areas viding essential services such as basic edu- · Type 3: divided countries with densely cation, health care, water, and sanitation populated lagging areas should be universal policy priorities. The costs associated with land sales--including This taxonomy can characterize lagging fees, survey costs, and transfer fees--can areas in most countries, but two qualifica- make land transactions prohibitively expen- tions are necessary. First, lagging areas in sive. In Russia the fees for private survey- some countries may be sufficiently hetero- ing are equivalent to two years' minimum geneous that it is difficult to neatly classify wages. These costs could slow migration by 242 WORLD DEVELOPMENT REPORT 2009 Map 8.4 In the Russian Federation, population densities are highest in the economically vibrant and warmer west, but a communist legacy has left some people in the cold interior Population per km2 Degrees Celsius 0­4.86 RUSSIAN 0­-3 4.86­10.75 FEDERATION -3­-7 10.75­20.14 -7­-11 20.14­75.64 -11­-15 >75.64 -15­-17 Seasonally frozen Permafrost extent Sources: Population density: Fay, Felkner, and Lall 2008; Ice Thickness: National Snow and Ice Data Center 2007. reducing the ability of less wealthy people Other countries may have just a few lag- to transact in land.27 China's household ging areas that are sparsely populated and registration system (the hukou system) has divided. In Indonesia, an otherwise unified been a barrier to rural-urban migration. country, places like Aceh are considered lag- Not having an urban hukou in urban areas ging areas, with divisions that weaken labor means that migrants may not qualify for and capital mobility. Policy makers may be public education or health benefits. This tempted to provide economic incentives for can produce large interregional wage dif- firms in these areas to compensate for the ferences. Recent research indicates that lack of factor mobility, but the accompany- removing such mobility restrictions would ing risk of creating enclaves of development reallocate labor across areas, reduce wage and deepening existing divisions should be differences, and lower income inequality.28 considered. Instead, initiatives that promote But the benefits depend on the response economic integration by increasing fac- of the urban housing market to additional tor mobility may be better suited for both demand from newcomers. economic and political reasons. Examples Some countries can have sparsely popu- include spatially targeted programs to lated lagging areas and domestic divisions. improve education and equal opportunity In Lao People's Democratic Republic, ethnic legislation to ensure that workers from lag- heterogeneity may make labor less mobile. ging areas do not face labor market discrim- Vientiane, the leading area, has a relatively ination in other parts of the country. low poverty rate, while the provinces in the Institutions and infrastructure to over- north and south have higher rates (see map come distance and density. When distance is 8.5). But the poor are spread out quite uni- coupledwithhighpopulationdensitiesinlag- formly across the country. In such cases, gingareas,spatiallyconnectiveinfrastructure much of the policy response still should is also necessary. Countries in this category be spatially neutral, with special efforts include Bangladesh, Brazil, Colombia, the to ensure equal access to public services Arab Republic of Egypt, Thailand, and Tur- to people in these areas. Afghanistan and key. Isolation from markets in more dynamic Tajikistan are other examples of countries parts of the country (or the world) can reduce with divisions and sparsely populated lag- consumer welfare, as residents face higher ging areas. prices because of market fragmentation, and Unity, Not Uniformity 243 workers and producers have less access to Map 8.5 In the Lao People's Democratic Republic, the poverty rate is high outside the capital markets. In principle, infrastructure invest- region, but poor people are scattered in remote communities ments that connect peripheral areas to mar- kets should improve both consumer welfare and productive efficiency. LAO With sizable concentrations of the poor P.D.R. in lagging areas, spatially blind institutions that promote the mobility of labor and cap- ital and ensure the provision of basic ser- vices must be aided by policies to improve the access of entrepreneurs in lagging areas to markets. Although migration will aid spatial efficiency and equity, with large numbers of the poor in lagging areas, this could take a long time (see chapter 5). Bet- ter infrastructural links between lagging Poverty rate: and leading areas, by improving market proportion of poor (%) Poverty density access, may allow some activities to flour- 21­40 Each dot represents ish in lagging areas. But they may increase 41­60 61­80 2,000 poor persons the concentration of economic activity 81­100 in leading areas, because firms that value agglomeration benefits will now be able Source: Richter, van der Weide, and Souksavath 2005. Note: Dots are placed randomly within each province and do not reflect population distribution. to serve lagging area markets from farther away (see chapter 6). Activities that respond to better infrastructure in lagging areas are those that do not exhibit agglomeration spatial efficiency in production and spatially economies--agriculture, agroprocessing, egalitarian living standards. and labor-intensive manufacturing such as Thereisalonghistoryofusingconnective leather and wood products (see chapter 4). infrastructure to integrate peripheral areas A useful way to conceptualize how infra- with national markets. When accompanied structure investments improve connectiv- by institutions that integrate nations, such ity is to think about a measure of market infrastructure investments can pay off. In access that captures the size and density the United States, the Congress passed the of market centers and the quality of trans- Appalachian Regional Development Act in port networks that link different locations 1965, relying on spatially blind institutions to these centers. The measure comes from and connective infrastructure to integrate the gravity model used to analyze trade the 22 million people in this lagging area, between areas and countries, with the inter- which spans 13 states, with the rest of the action between two places proportional to country.29 The basic strategy combined their size (population or economic density) regionally coordinated social programs and inversely proportional to the distance and physical infrastructure. The 1965 Act between them (see box 8.6). allocated 85 percent of the funds for high- Consider the Arab Republic of Egypt. ways--seen as critical to meeting other The location of human settlements has been socioeconomic objectives--and, cumula- dictated by a dominant natural geography tively, highways have accounted for more constraint--access to water. Most people, than 60 percent of the appropriated funds in leading areas around Cairo and Alexan- through the mid-1990s. Other investments dria as well as lagging areas in Upper Egypt included hospitals and treatment centers, around Aswan and Qena, live along the Nile land conservation, mineland restoration, (see map 8.6). As in the densely populated flood control and water resource manage- coastal zone in Brazil, institutions to inte- ment, vocational education facilities, and grate Egypt need to be complemented with sewage treatment works. Between 1965 and connective infrastructure to ensure both 1991, total personal income and earnings 244 WORLD DEVELOPMENT REPORT 2009 BOX 8.6 Low market access in Mexico's lagging south Quantitative information on regional that center. It is computed using an up- the federal district, thanks to concentra- or local market integration is scarce. to-date digital map of transportation tions of people and infrastructure. A quar- Summary statistics--such as the road infrastructure from the Mexican statistical ter of Mexico's GDP is generated within length in a state or province or the agency (Instituto Nacional de Estadística two hours' travel time from the center of straight-line distance to ports or urban y Geografía, INEGI).a For each road seg- the Federal District. The southern states agglomerations--are poor proxies for the ment, the database indicates the number of Chiapas, Guerrero, and Oaxaca, the complexity of a national or regional trans- of lanes and whether those lanes are poorest areas, have low market access. portation network. To improve on them, paved or unpaved--and for railroad lines, a geographic representation of Mexico's the number of tracks. For each category Source: Deichmann, Fay, Koo, and Lall 2004. transport network is used to compute an of road or rail, average travel speeds are a. The digital road and rail network includes index of accessibility for each municipio in estimated to calculate how long it will 171,000 kms of roads, of which 84,000 kms are paved roads; 51,000 are unpaved; and the country as a simple measure of poten- take to traverse each segment in the 36,000 are paths and breaches. The rail tial market integration. transport network.b Urban population network has an estimated total length of This index summarizes the size of the data from the INEGI database indicate 14,000 kms. These values are calculated by potential market that can be reached the location and population size of about a geographic information system (GIS) from from a particular point given the density 700 cities and agglomerations in Mexico. 1:1 million scale digital maps and may not and quality of the transport network in These urban centers accounted for about necessarily match o cial statistics. b. Using travel time on a transport network that region. For any point in the country, 68 million of Mexico's 97 million people provides a more accurate measure of acces- it is the sum of the population of urban in 2000. sibility compared with the computationally centers surrounding that point, inversely The map of market access (below) much simpler straight-line distance, as weighted by the travel time to reach shows high values of the index around employed, for example, by Hanson (1998). Market access in Mexico is highest around the national capital and low in the lagging southern states Accessibility Index Low High City population, 2000 (thousands) 50­100 Mexico City 100­500 500­1,000 1,000­5,000 > 5,000 GUERRERO OAXACA CHIAPAS MEXICO Source: Deichmann, Fay, Koo, and Lall 2004. grew 48 percentage points faster on aver- caused by ethnolinguistic or religious het- age in the Appalachian counties than in erogeneity, the forces of factor mobility can their economic "sisters," population grew 5 be weak even over relatively short distances. points faster, and per capita income grew 17 Examples include India and Nigeria. In points faster.30 such cases, spatially focused incentives Institutions, infrastructure, and incen- may need to complement institutions and tives to overcome distance, density, and infrastructure to encourage economic pro- division. When countries face divisions duction in lagging areas. Commonly used Unity, Not Uniformity 245 Map 8.6 In the Arab Republic of Egypt, population densities are high in lagging areas, and connective infrastructure is needed to help spatial integration Primary roads Secondary roads Population density High ARAB REP. Low OF EGYPT Source: World Bank 2008b. incentives include fiscal incentives and of people, encouraging them to seek eco- subsidies, special economic zones, industry nomic opportunities wherever they arise. location regulations, and investment cli- Second, they should help unify countries in mate improvements. the long term. Internal divisions may be a Only a few countries face the triple chal- short-run constraint for economic integra- lenges of distance, density, and division. tion, but the diminution of division should Nevertheless, incentives to promote eco- be a long-term objective. Third, policies nomic development in lagging areas have that try to offset constraints posed by divi- been widespread. The experience has been sions should not inadvertently strengthen disheartening.31 In good measure this has them. For example, making land and labor been because incentives have been used laws or school systems different in lagging instead of, rather than in addition to, inte- areas may weaken economic and political grative institutions and infrastructure. integration. More often than not, geographically tar- geted incentives would not even be neces- The framework in action sary if the policy objective was to integrate This section uses the framework to discuss leading and lagging areas, rather than pro- how countries have used specific policies to mote industry in economically unfavorable integrate lagging and leading areas. places. Table 8.1 summarizes policy options Institutions that promote portable for domestic integration using a calibrated investments combination of institutions, infrastruc- Investing in human capital. Universal ture, and incentives. Policy makers should primary education and basic health are keep three points in mind in designing mandated across a broad range of develop- integration strategies. First, policies should ing countries, as shown by the review of focus primarily on improving the welfare national constitutions discussed earlier. For 246 WORLD DEVELOPMENT REPORT 2009 Table 8.1 An instrument per dimension--a framework for area, territorial, or regional development policies Country type Densely populated lagging areas in Densely populated lagging areas in divided Sparsely populated lagging areas united countries countries Examples (countries) Chile, China, Ghana, Honduras, Bangladesh, Brazil, Colombia, Arab India, Nigeria Pakistan, Peru, Russian Federation, Rep. of Egypt, Mexico, Thailand, Sri Lanka, Uganda, Vietnam Turkey Dimensions of the Economic distance (1-D) Economic distance Economic distance integration challenge High population densities in lagging High population densities areas (2-D) Internal divisions (3-D) What policies should Labor and capital mobility Labor and capital mobility Labor and capital mobility facilitate Market integration for goods and Market integration for goods and services services Selected economic activities in lagging areas Policy Priorities Spatially blind Fluid land and labor markets, security, Fluid land and labor markets, security, Fluid land and labor markets, security, institutions education and health programs, safe education and health programs, safe education and health programs, safe water water and sanitation water and sanitation and sanitation Spatially connective Interregional transport infrastructure Interregional transport infrastructure infrastructure Information and communication Information and communication services services Spatially targeted Incentives to agriculture and agro-based incentives industry Irrigation systems Workforce training Local roads Source: WDR 2009 team. example, constitutions state that primary between lagging areas and potential des- education should be free and universal, tinations have deterred migration. Reduc- regardless of the place of residence, and sup- ing economic distance, an additional year plementary national laws specify how many of education increases out-migration from years of instruction are necessary to com- remote areas by 40 percent.32 As Brazil plete primary education. In conflict-driven transformed from an agricultural to a man- or postconflict countries, basic education is ufacturing economy, migration flows from viewed as a tool for national reconciliation the lagging northeast to the dynamic south and ensuring territorial integrity. and southeast increased between 1960 and Despite such legislation, education, 2000. In the northeast people who have at health, and poverty levels vary consider- least a primary education migrate more fre- ably among areas in many countries, par- quently than less educated people.33 ticularly in Asia and Africa. In China, the One of the biggest success stories is in the human development index (a combination United States, where a rise in the schooling of education, health, and income levels) of of African Americans is believed to have the leading area in 2003 was 0.97, close to been an important causal factor behind their the Republic of Korea's index, and that of "Great Migration" out of the South. In 1900, the lagging area was 0.59, about the same as 90 percent of African Americans lived in the Lao PDR's index (see figure 8.2). Chapter 2 South, and only 4.3 percent of those born in pointed out that these gaps were even higher the region were living elsewhere. By 1950 the some years ago. proportion in the South had declined to 68 Developing human capital is essential percent, and 19.6 percent of those born in whether policies aim to bring jobs to peo- the region had left it. Census data for 1900, ple or encourage the movement of people 1940, and 1950 show that better-educated to jobs. One of the main gains comes from people were more likely to migrate because helping people in lagging areas migrate to schooling increased their awareness of dis- areas with better opportunities. In Russia tant labor market opportunities and their the large economic and physical distances ability to assimilate into a different social Unity, Not Uniformity 247 and economic environment, thus lowering Figure 8.2 Living standards can vary considerably between leading and lagging areas the costs.34 In another U.S. study of people Human development index tracked between 1968 and 1982, those with 1.0 high education levels showed less inclination 0.97 to change professions but were more likely 0.9 to move geographically. A person with a col- 0.82 0.8 lege education was likely to move three times 0.78 0.76 0.75 0.74 more often than a person with an eighth- 0.7 grade education or less.35 0.67 0.6 0.67 Opening options for migration stimu- 0.60 0.58 0.58 0.55 lates greater human capital investments: 0.5 0.49 0.50 people consider not only the local returns to education but also the returns in other loca- 0.4 tions. If schooling options are available in 0.3 poor areas, potential migrants will invest in Lao PDR Indonesia Mongolia Vietnam Philippines Thailand China 2000 2002 2002 2001 2003 2003 2003 additional human capital, anticipating that jobs in leading areas require higher skills. Source: Gill and Kharas 2007. Employers in those areas are likely to favor educated workers who signal themselves as education quality because of the higher more "able" than other workers from lag- costs of delivering services in the northern ging areas. In the United States, African region. With poor implementation capacity American school enrollment rates were and underspending in lagging areas, the gap significantly higher in southern states that between "regional needs" and allocations previously had experienced high rates of out- from higher levels of government becomes migration. An increase in earlier migration even wider. Although this could be seen rates explains 7.4 percent of the increase in as spatial targeting of public spending, an African American enrollment rates between outcome-oriented policy framework would 1910 and 1930. As more African Americans regard such efforts as spatially blind. migrated from the South, migration became Transfermechanismsforpublicservices. more common and feasible, and school Redistributive transfers from higher levels enrollments rose in response. of government can reduce disparities in fis- Schooling has a strong effect on welfare, cal capacity and public service provision as in Brazil. Nine states in the northeast across subnational jurisdictions. At least have the worst education attainments, with three criteria motivate their allocation: gains smaller than in the rest of the coun- · Need. Areas with lower incomes would try. Average illiteracy in the nine states fell receive more investment, but richer 42 percent, less than the 49 percent in other areas may also demand more resources states, and is still twice the national aver- to meet the needs of population growth age (18 percent versus 9 percent). Differ- and congestion. ences in schooling explain more than half · Efficiency. Areas with higher returns the income difference between the north- to investment would receive more east and the leading southeast. If the local allocations. populace had the same education profile as · Equality. Spending is equalized across people in the southeast, average incomes in locations, so that public investments do the northeast would increase by more than not give an advantage to any single area. half, moving from 62 percent of the Brazil- ian average to 93 percent.36 Need-based transfers can improve pub- Government programs, such as that for lic service delivery in lagging areas, because the universal primary education in Uganda, local tax bases may be inadequate to gen- often reflect national priorities. Uganda's erate enough revenues. Intergovernmental program increased enrollments in the transfers can help provide similar access to north--the country's poorest area (see box public services for residents anywhere in 8.7). But more effort is needed to improve the country. Such transfers are particularly 248 WORLD DEVELOPMENT REPORT 2009 and households) benefit from agglomera- tion economies in leading areas. External BOX 8.7 Universal primary education in Uganda increased economies induce mobile factors to cluster access to schools in the northern areas geographically and turn them into quasi- President Museveni's decision to UPE has had an equalizing e ect fixed factors. So if residents see benefits implement universal primary educa- in access but not in quality and per- from locating near other similar residents, tion (UPE) in 1996 made Uganda the formance. In the northern area, high they become locked into these locations, rst African country to institute such a rates of teacher absenteeism, low less sensitive to tax differences. Moderate policy. UPE abolished tuition and con- nancial incentives to teachers, and intergovernmental transfers financed by tributions to parent-teacher associa- limited education infrastructure and leading areas thus can finance public ser- tions and school building funds. The teaching materials produce low edu- impact on primary school enrollment cation performance. And the per cap- vices in lagging areas. has been large, with those in the poor- ita budget allocations to the region Although transfers can bridge short- est quintile gaining most and, within do not always re ect the higher costs term fiscal constraints in lagging areas, the poorest quintile, the enrollment of delivering services there. A spa- fiscal dependency is a danger. If intergov- of girls more than tripled between tially blind program of education that ernment transfers finance a large share of 1992­93 and 2002­03. In the lagging emphasizes outcomes would not be expenditures, subnational governments are northern area, girls' enrollment rose geographically neutral. unlikely to improve local revenue collec- from 40 percent to 73 percent. Source: Bird and Higgins 2008. tion or be accountable to local residents.38 OECD countries have recognized these disincentives, and many have reduced the important for subnational governments equalization component of revenues and that depend heavily on federal transfers grants (Italy and Spain, for example). to cover spending. They finance about 60 In India, where federal transfers redistrib- percent of subnational spending in devel- ute resources to poor areas, average incomes oping countries and transition economies, inlow-incomestatesare40percentofthosein compared with about a third in member high-income states. With local tax revenues countries of the Organisation for Eco- linked to local incomes, the fiscal capacity of nomic Co-operation and Development low-income states is worse than that of their (OECD). In India central government high-income counterparts.39 Compensating transfers finance more than 30 percent of for this difference is a progressive fiscal redis- state spending. In China central-provincial tribution system. Low-income states receive and provincial-local transfers financed 67 48 percent of total central government trans- percent of provincial, 57 percent of prefec- fers, compared with a 17-percent allocation ture, and 66 percent of county and lower- to high-income states. The progressiveness level spending in 2003.37 of transfers is also evident per capita--Bihar, The allocation rules for transfers thus the poorest state receives Rs 501 per person in have a direct bearing on the potential for tax transfers. Maharashtra--a high-income welfare improvement in different areas. But state and home to India's leading urban intergovernmental transfers that finance a center, Mumbai--receives only Rs 298 per large share of subnational expenditures are person. But the translation of resources into rarely made with spatial equity in mind. services on the ground is not always visible in Indeed, the large transfers go to areas where India's lagging areas. people already receive high-quality services Many developing countries are collecting (see box 8.8). and disseminating credible information on Fiscal equalization transfers to lagging service entitlements to increase the account- areas are financed by a net tax on the resi- ability of service providers and improve dents of leading areas. A common concern outcomes. Increasing access to reliable in the fiscal competition literature is that quantitative information about service deliv- higher effective taxes in some areas will ery outcomes makes it difficult for providers stimulate the out-migration of productive to ignore this information as anecdotal or factors. The new economic geography pro- irrelevant. Involving community members vides some hope that tax-induced migra- in identifying concerns and encouraging tion will be limited if residents (both firms them to do their own monitoring can create Unity, Not Uniformity 249 BOX 8.8 Improving the spatial progressivity of Nigeria's intergovernmental transfers Poverty and service quality in Nigeria are are lower than those in the southwest The proposed changes would shift worst in the north, particularly the north- (3,700) despite the north's having the intergovernmental transfers toward east, and much better in the south, particu- highest overall poverty and worst service states with the greatest need. The maps larly the southwest.a, Nigeria's states rely b delivery.e To determine what would hap- below show this shift in per capita on scal transfers from the center to pro- pen if transfers were spatially progressive, terms. vide most services. Nigeria's allocation of an illustrative policy experiment was statutory grants (NGN 700 billion in 2006)c developed by the World Resources Insti- a. National Bureau of Statistics 2006. is not targeted using a clear principle that tute to identify the implications. Popula- b. See numerous measures of service qual- ity and access (National Bureau of Statistics supports poverty reduction: 54 percent of tion and land area, re ecting demand for 2006). Core Welfare Indicator Questionnaire the funds are divided equally among all services, are used to allocate 50 percent (CWIQ). Data can be obtained from the states regardless of population, land area, of the statutory grant. Equal allocations Nigerian Bureau of Statistics electronically poverty, or other measures of need.d Indi- are reduced to 5 percent (from 54 per- from http://www.nigerianstat.gov.ng/ cators of health care and education make cent). Education and health care are split cwiq/2006/survey0/index.html. up only 7 percent of the transfer. between measures to support current c. The 13 percent derivation of oil proceeds and disbursements from the oil fund (under The indicators chosen to direct that service levels and those to support pro- which 9 of the 36 southern states where small percentage are regressive in that gressive funding for states with the poor oil is produced receive 13 percent of oil they favor states with the best service services. For education, school enroll- revenues) represents about NGN 330 bil- delivery and strongest infrastructure. ment (increased from 4 to 5 percent) is lion, or about one-third of total transfers to Basing education transfers purely on used for the former purpose, and lack states in 2006. While the derivation strongly enrollments favors states that already have of access to schooling (increased from 0 shapes Nigeria's overall transfers, this trans- fer is stipulated by the constitution, not by education infrastructure and teachers, to 5 percent) for the latter. Health care statute and so is not included here (Nigeria penalizing those that do not. Basing health also received a 10 percent weight. The Federal Ministry of Finance, downloaded transfers purely on hospital beds similarly allocation for revenue e ort is kept at 2.5 and compiled from http://www.fmf.gov.ng/ supports better-o states that have the percent. Poverty was added as a category portal/detail.php?link=faac). resources to build more hospitals. by weighting the number of poor and d. Revenue Mobilization Allocation and Fis- Per capita transfers to states in the the poverty rate (headcount ratio) at 2.5 cal Commission 2003. e. National Bureau of Statistics 2004. north (about NGN 3,300 per person) percent apiece. Nigeria's statutory grant transfers per capita with actual transfers in 2006 (left) and with equity considerations (right) Actual Proposed NIGERIA Naira per capita Naira per capita 180­435 255­400 435­752 400­755 752­1375 755­1036 No data No data Sources: World Resources Institute; www.funnelthemoney.org; Layke and Adam 2008. 250 WORLD DEVELOPMENT REPORT 2009 a constituency of users willing to hold pro- below the market rate and further restricted viders accountable. mobility. Tenants do not want to lose their Effective service delivery thus depends favorable position in existing contracts.42 In on having enough financial resources and Chile, meanwhile, the government's success translating them into outcomes on the in targeting housing subsidies to the poor in ground. Investing in public goods in lag- lagging areas has also created a strong incen- ging areas reduces migration possibilities tive for people in these areas not to migrate, in the short term, because it improves the impeding convergence across locations even attractiveness of lagging areas. But targeted in a country that has a small population and social investments to develop portable homogenous society. assets can improve the welfare of people and facilitate their longer-term mobility by Institutions and infrastructure to making them better prepared for work in connect lagging and leading areas leading and intermediate areas. Transport infrastructure. Consider Portable pension benefits. Even with Bangladesh, a two-dimensional country, portable assets, people may find it hard to where the most lagging areas are distant move. In several countries, a lack of pen- from economic density but densely popu- sion portability is likely to slow the pace of lated. Selected investments in correc- migration. Facing a potential loss in pension tive infrastructure can help greatly. The benefits because of differences in schemes bridge over the Jamuna River opened mar- or a lack of portability, workers may be less ket access for producers in the northwest likely to move even when leading areas offer around the Rajshahi division. Built at a higher wages. The European Commission cost of almost $1 billion, this bridge pro- finds precisely this problem.40 vides the first road and rail link between Better land market regulations. Well- the northwest, an intermediate area with functioning land markets make it possible for the more developed east, which includes people to acquire land, exchange it with oth- the national capital region. Better market ers,anduseiteffectively.41 Andpoliciesthatset access and reduced input prices encouraged up defensible and tradable property rights for farmers to diversify into high-value crops, land and housing are likely to facilitate peo- such as modern varieties of rice and per- ple'sgeographicmobility.Theabilitytodefend ishable vegetables.43 The government has rights legally rather than physically allows complemented connective infrastructure people to leave their land to take advantage policies with spatially blind institutions of short-term opportunities. And the ability to improve coverage of social services. The to use land as collateral or sell it allows them Expanded Program on Immunization aims to finance migration costs and benefit from to immunize all children less than one year economic and social opportunities elsewhere. old against the six vaccine-preventable dis- Government involvement in managing eases. The Health, Nutrition, and Popula- land markets and enforcing property rights tion Strategic Investment Plan for 2003­10 prevents households from wasting pri- improves coverage in districts with poor vate resources. But too much government health indicators.44 involvement can hurt efficiency. Large-scale In the Islamic Republic of Iran, another public ownership can withhold land from two-dimensional country in the terminol- the market and artificially increase prices, ogy of this Report, connective infrastruc- precluding many poor people from entering ture improvements are necessary for spatial the market. And high direct costs and com- integration. In addition, spatially blind plicated procedures can reduce the incentive education policies of improving schools as for people to formally exchange land. well as conditional cash transfers for chil- Policies to safeguard the poor from rising dren to attend schools can improve welfare land and housing costs often have hindered in lagging areas.45 Turkey's lagging eastern market functioning. In the Czech Republic, areas have 44 percent of the land but only where there is a large rental market, de facto 5.7 percent of national motorways, and rent control has kept prices significantly asphalt road coverage is 40 percent that Unity, Not Uniformity 251 of the leading regions. The government's slower pace. Transport connectivity improve- Village Infrastructure Support Project ments in China's intermediate areas can be (KÖYDES) and Municipal Infrastructure economically beneficial for lagging areas. By Support Project (BELDES) have improved reducing the transport cost from the west to living standards in rural areas and small the coast, infrastructure investments in the towns by paving roads and providing sani- central (intermediate) transportation hubs tation and drinking water networks.46 And in Henan, Hubei, and Hunan provinces may investments in human capital are likely to well have greater effects on the west's devel- benefit potential migrants as well as those opment than improvements in the western who stay behind. area itself. But if China's overall growth is the Information and communication tech- main objective, infrastructure investments nologies. Mobile phones have driven down in the dynamic economic centers along the provision costs, boosting penetration and coast--Hebei, Jiangsu, and Shandong-- improving information flows. In 2003 China could still provide the highest payoffs.50 had more mobile phone users (269 million) In Brazil improvements to the road than fixed-line users (263 million). For 29 network between the 1950s and 1980s did areas in China between 1986 and 2002, tele- reduce transport and logistics costs. But communication infrastructure was strongly most of the economic gains accrued to the associated with subnational GDP growth.47 center-west, with only small gains to the lag- Because telecommunication investments ging northeast. During this time, its share are subject to diminishing returns, lagging of the national network increased from 15 areas can gain the most from them. percent to 25 percent. Even so, such invest- New technologies have lowered the costs ments did bring economic density closer to of delivering financial services, making the lagging northeast. them more affordable. Many people in lag- In Colombia, with water and land suitable ging areas have limited access to financial for agriculture, the mountainous topography services, relying on cash-based transactions makes freight transport difficult. So some outside the banking system. But with ris- intermediate areas are not well integrated ing international and national remittances, better access to financial services can help people in these areas overcome credit constraints. The proliferation of mobile BOX 8.9 Exporting by mail in Peru--connecting small services, even in remote areas, opens new producers to markets opportunities to provide financial services In many countries small enterprises need to use a customs agent, logistics over a mobile phone network (m-banking). are often excluded from export agent, or freight forwarder or to con- Reports from the Philippines indicate that chains because they operate in vil- solidate the merchandise; even the 3.5 million people have access to mobile lages or small towns or do not have packaging is provided. Firms or indi- phones that can transfer money.48 the needed information to export. viduals need only to go to a post o ce Producers in lagging areas can receive In Peru a trade-facilitation program with a scale and a paper scanner and to use the Internet to complete the better information on prices they can get called "Easy Export" connects small producers to markets. The key to this export declaration for the tax agency. for their products. In Kerala, India, mobile program is the most basic of transport Has it made a di erence? Within six phones reduced the dispersion of market networks--the national postal service. months of inception, more than 300 prices so much that prices differed only by How does it work? An individual or rms shipped goods totaling more the transport cost between markets.49 And rm takes a package to the nearest than $300,000. Most users are new in Peru, a ubiquitous but often undervalued post o ce, which provides free pack- exporters--microentrepreneurs communication system is connecting small aging. The sender lls out an export and small rms, producing jewelry, alpaca and cotton garments, food producers with markets--the postal service declaration form, and the post o ce weighs the package and scans the supplements (natural products), (see box 8.9). export declaration form. The sender cosmetics, wood art and crafts, shoes Greater benefits in intermediate areas. pays the fee for the type of service and leather, and processed food. Intermediate areas closer to centers of eco- desired. Goods with values of $2,000 And many of them are in the poorest nomic mass are likely to gain more from or less can be exported. The main areas of the country. connectivity-enhancing infrastructure, and bene t is that the exporter does not Source: Guasch 2008. lagging areas are likely to gain less and at a 252 WORLD DEVELOPMENT REPORT 2009 with large domestic and international mar- the development of extensive rural road kets. Casanare, the nation's largest rice pro- networks led to significant economic bene- ducing area, has good potential for biofuels fits, with considerable gains to the poor. But from corn and palm oil. But it takes 18 hours the poor's share often is not large enough to reach Bogotá and 50 hours to access a main to significantly reduce income inequality, port. Improving road quality would increase because the benefits from road extension market access and help the area's economy. could be greater for the rich.55 In Ghana La Mojana, an area with 5,000 square kilo- the benefits of improving access to infra- meters of flatlands, close to the Atlantic ports structure by the poor could be increased and most Colombian cities, is often flooded, by complementary spatially blind policies, because it lies in the buffer zone of two major such as education and health improve- rivers. Improving ecosystem management ments, which would boost the use of that along with transport connectivity would infrastructure.56 improve its access to cities and ports.51 Interregional infrastructure improve- Institutions, infrastructure, and ments can bring higher economic concen- incentives to overcome the 3-D tration. The potential benefits of improving challenge of distance, density, and marketaccessforperipheralareasmayinstead division accrue to firms in larger agglomerations.52 Incentives to promote economic investments And improving transport connectivity can in lagging areas have been widely used by further concentrate economic activity. Roads countries to accelerate national economic and rails run both ways--better transport growth and balance growth outcomes across connectivity not only provides market access places. They seem to work better when they to firms in lagging areas, but also allows reinforce market signals and address coor- firms in leading areas to reach markets. A dination failures. They are less successful decline in transport costs helps competitive when governments pick the places to sup- firms in leading areas easily scale up produc- port growth. The following lesson seems tion to reach these new markets at lower cost to emerge: let markets pick the place, while relative to local producers in lagging areas. governments help to push the pace. So improving market access may hurt the Incentives that exploit geographic production of standardized goods in lagging advantages are more likely to succeed. To areas. But lower prices and better access to stimulate economic growth, many gov- new products are likely to improve consumer ernments have offered tax holidays, reli- welfare. able infrastructure, and improvements Experience validates this conjecture. In to the business environment. Often the Italy reducing transport costs between the incentives are geographically focused--in north and south in the 1950s deprived Mez- special economic zones--to quickly cre- zogiorno firms of their previous protection ate enclaves of growth, leaving nationwide and accelerated their deindustrialization.53 infrastructure and governance shortfalls And in France, where transport costs within to the longer term. There is an ongoing the country fell by 38 percent between 1978 debate over whether focused incentives and 1993, the geographic concentration of slow the pace of economic reforms, but the employment increased.54 interest here is in identifying where these In addition to growth effects that vary incentives are more likely to succeed. Is across areas, it is also useful to consider it desirable to provide incentives in areas the distributional effects of infrastructure that already have good geography and improvements. Are the benefits of infra- human capital? Or should they be reme- structure improvements large enough and dial measures to offset market forces and distributed progressively enough to reduce help develop lagging areas? overall income inequalities, with more In China and India, spatially targeted benefits accruing to the poor than the non- incentives are most likely to succeed when poor? Empirical evidence on this question is they reinforce geographic advantages, par- limited, but a study from Nepal shows that ticularly in areas advantaged by good access Unity, Not Uniformity 253 to domestic and international markets Let markets pick the places. The Repub- (see box 8.10). In Uganda the returns from lic of Korea is one of the few success stories infrastructure development in highways and involving spatially targeted incentives. To power supply are highest in areas that already support economic growth in specific areas, have a skilled labor force and a diverse mix the national government worked with the of industrial activities.57 These happen to be private sector to identify areas offering pro- along the corridor linking the country's two duction advantages. Consistency between main agglomerations, Kampala and Jinja. national industrial policy and regional pol- Using infrastructure to spread out manufac- icy objectives was instrumental. Although turing, instead of facilitating its concentra- deconcentrating economic activity from the tion, can slow national economic growth. Seoul metropolitan region was an implicit Incentives that enhance market links policy objective in the government's tax and improve agriculture performance in sharing, decisions to promote export- areas with good natural geography can be oriented "strategic industries" were at the a part of development strategies for densely core of industrial and regional policies.61 populated lagging areas where factor Spatial equity did not guide national mobility is constrained. But before offering industrial policies. In fact, areas picked by incentives, agriculture needs to be assessed the market in different phases of industri- as an economic driver in the local economy. alization were encouraged. In the 1960s and World Development Report 2008 provides a 1970s, national industrial policies created useful diagnostic technique to identify sub- new industrial cities--Ansan, Changwon, national areas as agriculture based, trans- Kumi, Kwangyang, Pohang, and Ulsan. forming, or urbanized--a country's "three The private sector (chaebols) established worlds" of agriculture--based on the share large branch plants with imported technol- of aggregate growth originating in agricul- ogy and borrowed foreign capital. Market- ture and the share of aggregate poverty in driven industrial and regional policy led to the rural sector. Applying this technique different specializations across the country, can help identify whether agriculture will with chaebol headquarters concentrating in remain a prominent feature of lagging areas Seoul, and production functions decentral- in the short to medium term. ized to areas outside the capital. Since the Consider Malaysia, where agriculture mid-1980s, industrial policy to support is important in the regional economies of high-technology activities triggered indus- the lagging areas, which account for more trial reconcentration in the capital region. than 40 percent of the country's people To speed the growth, incentives were (Sabah and Sarawak are home to more complemented by infrastructure invest- than 2 million).58 To encourage agricul- ments that connected the southeast to the tural development in the eastern peninsula, capital region. The Gyeongbu expressway, the national government has been offering which connects Busan, Daegu, Daejeon, and reinvestment allowances for capital expen- Seoul, enabled industries in the southeast to ditures related to farming, providing cold- reach the capital region within five hours. chain facilities and services for perishable Thus industries producing standardized agricultural produce and exempting food intermediate goods in the capital and south- processing from tax.59 east regions benefited from considerable cost But in Ghana, where the lagging north is reductions. In the Republic of Korea, it may mostly in the arid Savannah zone and pop- be fair to conclude that markets picked the ulation densities are low, expanding agri- place and governments pushed the pace.62 culture is less likely to facilitate territorial Many countries have offered incentives to integration. Without allowing for large-scale create economic mass in lagging areas. The migration or structural transformation, even idea is that to attract firms, lagging areas a sharp acceleration in productivity growth need to offset higher transport and logistics in groundnuts and other northern staples is costs, weaker infrastructure, higher factor insufficient to bring the north up to par with prices, and lower levels of public services. the south in the medium term.60 European countries have a long history of 254 WORLD DEVELOPMENT REPORT 2009 BOX 8.10 Special economic zones bring growth if they exploit advantages in natural and economic geographies Many developing countries have loca- ince and Xiamen in Fujian Province (see rior states of Haryana, Karnataka, Punjab, tions where infrastructure conditions the map below). In 1984 14 coastal cities and Rajasthan. Even in the coastal states, and economic regulations are more hos- opened their doors to overseas invest- many SEZs are not along the coast.b pitable than those typical in the rest of ment, and in 1988 the entire island of Closer inspection reveals consider- the country. These locations, often called Hainan was assigned SEZ status. Around able diversity across product special- special economic zones (SEZs), enhance the same time, the coastal belt around ization, which range from standardized industry competitiveness, attract foreign the Yangtze River Delta, the Pearl River manufacturing to information technol- direct investment, and diversify exports. Delta, and the Xiamen-Zhangzhou- ogy and pharmaceuticals. Compared Recent estimates suggest that there are Quanzhou Triangle in south Fujian with standardized manufacturing 2,300 such zones in developing and tran- opened for business with the world. In products, human capital­intensive sition countries.a the early 1990s the government opened products depend more on reliable up 11 border cities and six ports along telecommunication infrastructure Look at China the Yangtze River. The developments and access to airports, not harbors. The earliest developing-country SEZs reflected a strategy of exploiting the Moreover, India's economic zones also were established in China under Deng best locations to access external mar- target the large domestic market. Take Xiaoping's leadership in the early 1980s. kets. Gurgaon, a satellite town a stone's In 1978 the government decided to throw from Delhi's international air- open the country's economy to the And at India port, which 20 years ago was a cluster outside world. SEZs and "open" coastal A cursory glance at India's SEZs suggests of villages (gaon is the Hindi word for cities were integral to this process. In that they were not nearly as well located. village). Now it is one of the main ser- 1980 SEZs were established along the In 2007 SEZs were approved in coastal vice-oriented corridors in the country, southeastern coast in Shenzhen, Zhu- states of Andhra Pradesh, Gujarat, Maha- sitting in the middle of India's largest hai, and Shantou in Guangdong Prov- rashtra, and Tamil Nadu, as well as inte- consumer market. It houses such infor- China's special economic zones opened the country to external markets 4 first special economic zones (1980) Province of Hainan (1988) 14 coastal cities (1984) 6 ports of Yangtze (1992) 3 deltas (1985) 11 border cities (1992) Source: WDR 2009 team, based on Huang and Luo forthcoming. Unity, Not Uniformity 255 BOX 8.10 Special economic zones bring growth if they exploit advantages in natural and economic geographies--continued mation technology service providers in the 1980s. They are generally more geographic advantages rather than try as IBM and Microsoft, along with such profitable and have better social and to o set them. consumer goods providers as Coca- environmental track records than pub- Cola, Gillette, and Nestlé. lic zones, except in East Asia's govern- Source: WDR 2009 team. Unlike China's government-led ment-run zones. a. SEZs take a variety of forms, which SEZs, India's are being developed by Location is the key: poor location is include free trade zones, export processing the private sector, by such companies the main obstacle to success. It seems zones, enterprise zones, freeports, and spe- as Infosys and HCL, promoting infor- to matter more where the zones are cialized zones (Gauthier 2007). mation technology development,c b. Location of SEZs is based on SEZ INDI- located, not who owns and operates AWEB at http://www.sezindiaweb.com/ reflecting broader global trends. Of them. The lesson from China and India is SEZ_map.html. SEZs in developing countries, 62 per- that spatially directed interventions are c. Information on SEZs in India is drawn cent are private, up from 25 percent more likely to succeed when they exploit from the Indiastat database. using industrial policies to attract firms to incentives and building grants to lagging lagging areas. But fiscal incentives, while areas through the Local Employment politically efficient, have not transformed Act, along with relocation of 250,000 the economic fortunes of lagging areas:63 manufacturing jobs from prosperous to lagging areas. · In Italy one of the main objectives of the national economic program begun in 1965 A common theme in traditional wastoeliminatedevelopmentgapsbetween regional policies focusing on taxes, sub- the south and the rest of the country. To sidies, and regulations is that they were support this program, financial incentives mainly central government initiatives to wereprovidedtofirmsinthesouththrough create employment and provide infrastruc- partialexemptionsof welfarecontributions. ture with the objective of dispersing (or Through 1992 public infrastructure and fighting the concentration of) economic financial incentives promoted industrial activity. They targeted specific firms in the development in the region. hope that they would become anchors in · In France the Fifth National Plan the local economy and have large multi- (1966­70) provided assistance to agricul- plier effects.64 But for the most part, these ture and began to direct industrial invest- incentives have not stimulated sustained ment away from Paris toward low-income growth in the lagging areas, imposing areas in the west. Light industries--with large costs on taxpayers. lower transport costs and higher labor Most European countries now focus more intensity--were targeted for relocation. on "soft" interventions, such as investing in · The United Kingdom has supported innovation and supporting research insti- economic development in northern tutes and science and technology parks (see England, Scotland, Wales, and north- table 8.2).65 Central government programs ern Ireland. During World War II, war- have been replaced by greater cooperation time building controls directed industry between the public and private sectors. And out of the south of England and into rather than target specific firms, coordinated northern and peripheral areas. During measures are attracting clusters of interre- the postwar recession in 1958, employ- lated firms. Rigorous evaluations of these ment in coal mining, textiles, and ship- programs are hard to come by. But innova- building declined, renewing interest in tion policies that bring new information and bringing jobs to the north. The 1960s technologies to lagging areas should in prin- saw regionally differentiated investment ciple help in the long term. 256 WORLD DEVELOPMENT REPORT 2009 Table 8.2 The OECD's experience with incentives to stimulate lagging areas has been evolving locating outside these three large cities were "Traditional" "New" regional policies The economic eligible for a 50- to 100-percent reduction in Strategy regional policies perspective import duties and income, sales, and capital Objective Create jobs Provide complementary Incentives should be gains taxes, as well as accelerated deprecia- and provide services, subsidize the provided only to "new" tion and lower interest rates. Their impact large-scale cost of innovation-- activities--products new on economic decentralization was insignifi- infrastructure research institutes, to the local economy science and technology and new technologies to cant because import duties on raw materials parks produce existing products and capital goods were low to begin with, Main players Central Public-private Public sector has limited so the reductions had no effect on location government partnerships information on what firms decisions and lost revenues.69 want In India the Industrial Policy Resolu- Focus area Incentives to Encourage development Public sector support tion of 1956 set up a strict licensing system attract individual of clusters--both must target activities, firms sectors and business not sectors; subsidized to direct investment into lagging areas. development activities must have The Indian government decided that no strong spillovers licenses would be issued to new industrial Source: WDR 2009 team. units in the vicinity of large metropolitan areas. And state governments and financial The U.S. federal government is also institutions were asked to deny support to involved in smaller "economic develop- new industries in metropolitan areas even ment" programs. A recent review identi- when they did not require an industrial fies 180 programs of U.S. federal agencies license. Large public sector projects (steel addressing issues as diverse as planning plants, for example) were located in the lag- and economic development strategies, ging states of Bihar, Madhya Pradesh, and industrial parks, infrastructure repair, and Orissa. Industrial estates (or growth cen- building renovation. The agency with the ters) received infrastructure investments greatest visibility in this group is the Eco- and financial incentives for private indus- nomic Development Administration (EDA) trial investment in designated lagging dis- of the U.S. Department of Commerce.66 tricts. The policies effectively stifled growth The EDA has spent more than $188 billion in areas that had good market access and on economic development, but with little human capital and did not allow exit of coordination among initiatives, or a com- unproductive activities from lagging areas. mon policy objective. The EDA reports that Even after more than 30 years of draconian its investments leverage about $37 in pri- regulation, few districts in backward areas vate sector investment for every dollar that became major industrial centers. it spends.67 No independent evaluation of The economic reforms in 1991 scrapped these programs is available. these licensing policies. The decline of pro- Area incentives, popular in developing duction in inland areas continued, and countries, have produced mixed results places with good market access and good at best (see table 8.3). In Brazil, where the local business environments flourished. goal has been to attract "dynamic" indus- The ten best-performing industrial dis- tries to the lagging north and northeast tricts are now located south of the Vindhya by providing fiscal incentives, expendi- mountain range, which divides north from tures have reached $3 billion to $4 billion south.70 a year. A recent impact evaluation shows Similarly, in the former Soviet Union, that the allocation of these "constitutional central planners decided where firms would funds" did induce the entry of footloose locate and tried to spread economic activity manufacturing establishments into lagging throughout the country's landscape. They regions--but incentives were not attrac- spread production facilities across the former tive enough for vertically integrated indus- Soviet Union's millions of square kilometers. tries.68 Between 1970 and 1980 the Mexican Far from markets and lacking specialization, government used fiscal incentives to pro- their productivity suffered. How did the mote industrial development outside the transition to markets change things? New three largest urban agglomerations. Firms firms located closer to markets and old ones Unity, Not Uniformity 257 in remote areas closed down. The result was Table 8.3 A range of instruments has been used by governments to create economic mass in a 2.5 percent gain in firm productivity annu- lagging areas, with modest results ally between 1989 and 2004.71 Instrument Examples Coordinating local and national incen- Investment Brazil: Constitutional funds (interest rate subsidies)--induced entry tives. Incentives for lagging areas are best subsidies of footloose firms, but not for firms in vertically integrated industries (Carvalho, Lall, and Timmins 2005) coordinated with national sectoral policies: taking stock of various national economic Tax holidays Thailand: Income tax exemptions; sales tax reductions for firms locating in secondary cities in the 1970s--unsuccessful as deductions from promotion initiatives and aligning spatial taxable profits did not induce firms to locate in unprofitable locations interventions with these policies can help. (World Bank 1980) For instance, spatially targeted subsidies Reductions in Mexico: Import duty and tax exemptions for deconcentrating represent only 12 percent of Brazil's export import duties manufacturing out of the three largest agglomerations--unsuccessful promotion and industrialization subsidies, as tax rates were low to begin with (World Bank 1977, Scott 1982) which favor the industrial southeast. Esti- Industrial estates/ Chile: Free trade zones in zonas extremas with exemptions for customs, mates suggest that these industrial subsidies free trade zones value added tax, corporate profit, and real estate taxes--successful in the high-tax, high-tariff period until the mid-1990s, performance cost $42 billion in 1999, or 4.4 percent of declined with national import duty reduction from 35 percent in the GDP.72 In Brazil initiatives to recruit firms 1980s to 6 percent in 2000 (World Bank 2005b) into the northeast clearly were fighting an Regulation India: Preference to backward areas in industry licensing (1956 uphill battle against broader industrial industrial policy), with public sector­led industrial growth in lagging incentives that were better aligned with areas and regulations to stop industrial expansion in leading areas-- few backward areas took off, and when regulations were relaxed, these market forces. lagging areas declined further (Chakravorty and Lall 2007) In India, too, common pricing poli- Source: WDR 2009 team. cies to reduce overall inequalities hurt the economic prospects of lagging areas. The Freight Equalization Policy of 1956 stan- and public works, an assured loan from the dardized the prices for transporting "essen- national development bank of R$500 mil- tial" items such as coal, steel, and cement lion, and exemptions from local taxes for nationwide regardless of distance. Lost in 10 years. When the state government tried the process were the location-based advan- to renegotiate the deal fearing that it was tages of resource-rich areas. The affected too generous, Ford moved to Bahia, which areas included southern Bihar, eastern offered a package similar to the original one. Madhya Pradesh, and western Orissa, each Evaluations show that these "fiscal wars" cost among the poorest, least industrialized Brazilian taxpayers around $172,000 per job parts of the country. The policy weakened created--five times the cost of job creation the incentives for private capital to locate in a General Motors plant in Tennessee.73 production in lagging areas. Thinking through the design of incen- Decentralization often has been accom- tives. Before using incentives to promote panied by the efforts of subnational gov- economic development in lagging areas, ernments to create economic mass to meet national and subnational governments expenditure responsibilities. They offer fis- should first find out why some areas are cal incentives and tax expenditures to attract being bypassed by the market. Is it because firms to their jurisdictions. But if not coor- of the low social returns to economic pro- dinated, these incentives can be wasteful and duction in these places, the low ability to counterproductive. capture these returns, or the high cost of Look at the competition between states finance?74 Have policies actively or inad- in Brazil, where Bahia and Rio Grande do vertently blocked local economic growth? Sul competed to attract a Ford Motor Com- The success of incentives depends on how pany plant in the 1990s. Rio Grande do Sul well the problem is diagnosed, perhaps offered a package of incentives to Ford that starting with area-specific natural, human, included a R$210 million (around US$200 and infrastructure endowments.75 "Know million) loan from the state at extremely thy economy," a phrase used in the World favorable conditions (6 percent interest, 15 Development Report 2000/01, should be the years to repay), additional state expendi- motto of subnational governments. Good tures of R$234 million on infrastructure information can promote constructive 258 WORLD DEVELOPMENT REPORT 2009 debate on development options and build republican borders, limited interrepublic consensus around a development strategy. trade, and duplicated economic produc- If the information and subsequent anal- tion. In 1987, for example, 70 percent of all ysis points to specific opportunities for production in Serbia was consumed in the growth, the next step is to identify whether local market.77 the planned incentives are to subsidize cap- As discussed in chapter 5, analysis sug- ital formation or to promote innovation. gests that factor mobility equalizes wel- If they are to attract firms with potential fare across areas, weakening incentives to local multipliers, it is important to know break away from an economic and political whether the product lines value agglomer- union.78 By contrast, persistent inequalities ation economies, which would reduce the across areas fuel disintegration movements. power of the incentives. For firms in sectors Unity, not uniformity, is the valid principle in which economies of scale and agglom- for both political and economic integration. eration are important for production, it is This chapter has provided a framework less likely that spatially targeted interven- for integrating lagging and leading areas as tions will attract them to lagging areas. countries address economic distance, mis- Industrial surveys in Brazil, China, India, placed density, and internal divisions. Eco- Indonesia, and Mexico show that manufac- nomic forces are likely to produce spatial turing firms in many product lines value divergence in growth outcomes. Economic both internal scale economies from mar- models of geography and growth show that ket access and agglomeration economies increasing returns to scale and agglomera- in deciding their location.76 Firms produc- tion economies can start and sustain a vir- ing standardized products serving local tuous circle of growth and investment in a markets, and those specializing in natural few areas. resources, are less likely to value agglom- For valid reasons, though, policy mak- eration economies than are those depend- ers are concerned with reducing geographic ing on skilled labor, business services, and imbalances soon, sometime between now access to information. and the long term. And sometimes, politi- And from a national growth perspective, cal pressures can be such that widening it is important to find out whether relocating divergence at any point is unacceptable. "targeted" industries produces net additional The typical territorial development policy employment and output nationally. If not, response has emphasized targeted incen- local efforts of attracting industry may be tives and large-scale infrastructure to zero-sum games. If the relocated industries encourage economic production in lagging are less productive, policy makers may face a areas. However, the evidence reviewed in negative sum. If incentives are being used to this chapter shows that many such policies promote innovation, it is important to ensure have led to waste. In the meantime, poli- that local production processes can accom- cies that address institutional bottlenecks modate the innovations. that can help people seize opportunities elsewhere or improve their living standards Avoiding Balkanization: locally may be ignored. the political benefits of Even with such compromises, the big- economic integration gest part of the policy challenge lies in Economic and political objectives can identifying the outcomes that can be real- clash, but more often they coincide. In the istically sought, that is, which depend on Western Balkans, the former republic of the stage of development and the fiscal and Yugoslavia became a federation after World institutional capacities of a country. Where War II but disintegrated when its republics incomes are low, it may be feasible only to declared themselves independent in the reduce spatial disparities in poverty rates early 1990s. Fueling the disintegration was and in access to essential shelter, water, rising autarky and fragmentation in Yugo- health, nutrition, and education services. slavia since the mid-1970s, with barriers China's 11th Five-Year Plan passed by the to movements of people and capital across National People's Congress in 2006 states Unity, Not Uniformity 259 that "the construction of public finance Table 8.4 Assessing the performance of area development policies system should be accelerated . . . to gradu- Reduce inequalities Pro-poor? ally equalize basic public services." In Octo- Performance across regions? (interpersonal Avoid tradeoff with ber 2007 the 17th Congress again pointed criteria (interregional equity) equity) spatial efficiency? out that, to narrow regional disparities, Institutions Yes Yes Yes equalization of basic public service provi- Infrastructure No No Yes sion would be the priority. Upper-middle- Incentives No No No income countries can be more ambitious Source: World Bank 2008b, based on country-specific case studies. in equalizing basic consumption indica- tors across areas, and developed countries such as those in the EU more ambitious density and division, targeted incentives still. Reducing spatial inequality in dispos- are necessary. able incomes may be the relevant target for This framework was tested using high-income countries. country-specific case studies of spatial But at all stages of development, forcing integration, which included Brazil, Ghana, economic production to spread evenly across India, Mexico, Russia, and Uganda. Each areas is both elusive and expensive. Growth set of integration policies is examined using generally is unbalanced, but it always three criteria: (1) Do they reduce economic brings more resources for societies to bal- distance across subnational areas? (2) Are ance development outcomes. Policy makers these policies pro-poor? (3) Are these poli- should identify and execute strategies that cies spatially efficient (that is, do these poli- balance development outcomes across areas cies avoid tradeoffs with spatial efficiency)? by means other than resisting the forces of Table 8.4 summarizes the findings. Efforts unbalanced growth--because that is tanta- to strengthen institutions fare well on all mount to fighting economic growth itself. three criteria. Although infrastructure The framework in this chapter is investments may not reduce economic dis- intended to help policy makers identify the tances or help the poor, they can be spatially policies best suited to addressing domes- efficient. Geographically specific incentives tic integration. The suggested solutions do not fare well. consider country-specific conditions. The Perhaps most important, the chapter main points? First, integration strategies identifies the point at which all discus- should increase the access of the poor in sions of territorial development policies lagging areas to opportunities, through a should start--with spatially blind institu- set of spatially blind institutions. Second, tions. Infrastructure that connects lagging infrastructure that connects lagging to areas to markets can help nations integrate. leading areas is needed when the problem Sometimes, not always, these discussions of distance between lagging and leading should include spatially targeted incen- areas is coupled with misplaced popula- tives. The right mix of integration instru- tion density in the lagging areas. Third, ments will bring the benefits that come when the problem of economic distance from both unbalanced growth and inclu- comes accompanied by both misplaced sive development. CHAPTER 9 Winners without Borders Integrating poor countries with world markets M any leaders in Africa called for a economies within their neighborhoods and political union of the continent with the rest of the world. Regional and at the time of independence. global integration have been complements, Félix Houphouët-Boigny,Côte d'Ivoire's first not substitutes, in the development of these president, was more pragmatic, promoting regions (see box 9.1). a gradual increase in economic cooperation But in most of the developing world, con- with neighboring countries. He proposed centration and convergence have been slow, one of the first regional economic agree- often because of persisting economic, politi- ments in Africa, the "Conseil de l'Entente," cal, and cultural divisions between coun- backed by a solidarity fund provided mainly tries (see chapter 3). These divisions make it by Côte d'Ivoire. The key elements of the hard for countries to take advantage of scale Entente were free trade and free movement economies (see chapter 4), mobile labor and of people.1 capital (see chapter 5), and falling trans- The preferred destination of migrants port costs (see chapter 6). Some developing was, naturally, Côte d'Ivoire. Its share of countries have tried to globalize through foreigners increased from 5 percent in unilateral liberalization; others have tried 1950 to 26 percent of its 16 million people to integrate regionally. There have been suc- in 1998--making the country one of the cesses and failures with both strategies. This top dozen destinations for international chapter deals with ways to combine these migrants in the world. Côte d'Ivoire ben- strategies by increased cooperation among efited as foreign workers contributed to neighbors and strong connections to world export-led growth in industry and agricul- markets, while recognizing and avoiding the ture. Sending countries--especially Benin, tradeoffs that can arise between these two Burkina Faso, Niger, and Togo--benefited approaches. from remittances and increased trade. The The chapter proposes regional integra- political crisis triggered by a coup in 1999 tion as a mechanism to increase local supply affected the entire region. But Houphouët- capacity and global integration to improve Boigny had vested his country's neighbors access to markets and suppliers. Integra- in its future, earning the nickname of "The tion means cooperation between countries Sage of Africa." in trade, domestic regulations and policies, Côte d'Ivoire reflects the main lines of regional infrastructure, and other cross- argument in this chapter. In today's devel- border initiatives, including public goods. oped regions--Europe, North America, Regional integration implies cooperation and Northeast Asia--most economic activ- within a neighborhood of countries. Global ities are highly concentrated, their exports integration implies cooperation at an even are specialized, and living standards are wider international level. converging. These regions have overcome This chapter's framework for policy 260 national borders and have integrated their action uses a taxonomy of neighborhoods Winners without Borders 261 to organize thinking about how best to con- front the development challenges of each of BOX 9.1 Are the policy messages of this Report the developing world's regions. anti­global integration? No. The main strategies are as follows: hood, which a ected the long-term · World Development Report 2009 Countries close to large world markets focuses on regional integration stability of the agreement and the should strive to benefit from proximity because that is where considerable willingness to respond to unexpected to high economic density and become scope for policy action now lies. But events. With many previous e orts an extension of the large markets. Mex- this does not imply that the message at regional integration having failed, ico, the Caribbean, the European Union is against global integration. Quite the pursuit of further regional agree- (EU) accession countries, and the Repub- the contrary. This chapter argues that ments has drawn considerable skepti- lic of Korea are linked, respectively, with regional cooperation boosts the sup- cism in development circles. ply capabilities of a neighborhood This chapter argues that, given the U.S., EU, and Japanese markets. But by providing regional public goods current conditions, this skepticism integration must go beyond a simple and taking advantage of regional is misplaced. Instability stemming free trade agreement to gain significant specialization. In this way, it can from macroeconomic policy and development benefits. The biggest chal- broaden the gains for each country poor governance is far less common lenge is to make domestic markets attrac- from global integration. In this sense, today than even a decade ago, so it tive enough to investors to be seen as an regional and global integration are is less likely that a country will import extension of the large market nearby. complements, not substitutes. With- problems from its neighbor even if out global integration, the bene ts their economies are integrated. And · Countries with big neighbors but far from regional cooperation would be with the decline in transport costs and from world markets should develop small or negative, as was true of many expansion in global trade, the bene ts their regional market. This requires past regional agreements. But with- from successful export-led growth are two instruments: institutional reforms out regional integration, the bene ts higher than ever. To compete, coun- that facilitate intraregional trade and from globalization might simply be tries are now more willing to harmo- nize their policies and institutions with factor mobility--and infrastructure unattainable for some countries, because they cannot compete on a others, so the prospects for regional investments that link lagging to lead- global scale by themselves. cooperation have grown substantially. ing countries and the region to major For many countries, especially in That may be one reason why, in June world markets. Regional integration can Africa where global export market 2006, 56 regional, 49 regional exten- naturally support regional production shares have fallen, the bene ts of sion (cooperation between a regional networks. These networks maximize global integration have been ephem- agreement and an individual country), production-cost advantages that come eral. Global integration is sometimes 5 superregional (cooperation between two or more regional agreements), with increasing returns to scale, and they seen as risky, and progress in the Doha Round on several issues central and fully 118 bilateral agreements allow small countries to specialize in to developing countries, such as agri- were signed or initiated under the niche products in regional supplier net- cultural trade, has been slow. In the World Trade Organization (WTO). By works. Greater cost efficiency on the sup- same vein, past regional cooperation acting under the global rules of the ply side makes it easier for such regions also did not yield signi cant bene ts, game, these agreements strive to rec- to then integrate with global markets. and many regional agreements fell ognize and avoid tradeo s between · regional and global integration. Countries far from world markets in apart. Those experiences also high- lighted the uneven gains across large Central Asia, the small Pacific Islands, and small countries in a neighbor- Source: WDR 2009 team. and Sub-Saharan Africa--the world's "bottom billion"--face the stiffest chal- lenges to economic growth and need a strong commitment for cooperative solu- munity can support these integration tions. Regional integration can occur in efforts through coordinated incentives. "natural" neighborhoods with three sets of instruments. They need close institu- East, Central, and West Africa fall into tional cooperation and comprehensive the third category. Resource-poor coastal regional infrastructure investments, as countries in these neighborhoods have with the others. But they also may need been the poorest growth performers in the cross-country compensation mecha- world relative to other world regions.2 For nisms to sustain the integration effort them, the Report suggests a pact involving because deep integration is likely to lead regional governments and the international to uneven short-term gains and losses community to improve social services and across countries. The international com- human capital in lagging countries and to 262 WORLD DEVELOPMENT REPORT 2009 improve infrastructure in leading countries chains) than they can by relying on sup- where takeoff is most likely. This should be pliers in one country alone (see box 9.2). augmented by preferential access to devel- Global integration provides the demand oped country markets for regional exports. and incentive to develop such efficient In return, both leading and lagging coun- regional supply networks. This combina- tries in these "natural neighborhoods" tion of regional and global integration has would allow freer intraregional movements produced successful developers in today's of labor, capital, goods, and services. rich neighborhoods. Today's developing countries, as late- Plant data add further detail to aggre- comers, face a stark choice: stay divided gate econometric findings.5 Firms export- and lose ground, or become winners with- ing to regional markets are hurt more by out borders. power outages and inefficient border pro- cedures than are firms exporting to global Regional integration to scale markets, although firms exporting time- up supply, global integration sensitive products such as textiles to global to scale up demand markets are hurt by inefficient borders as well. The efficiency of firms dictates where Some countries, such as Chile, Mauritius, they sell their products: the least efficient and the well-known East Asian tigers, have sell only in domestic markets, others serve integrated globally without much coop- both regional and domestic markets, and eration within their world region. They the most efficient are involved in domestic, enjoyed significant first-mover advantages. regional, and international markets.6 But many other developing countries have A successful integration policy will con- found this hard to achieve, and some won- centrate economic activities in places with der if the emergence of highly competitive better access to markets and inputs, whether exporters like India and China makes the subnational, national, or regional. Integra- likelihood of a successful export-led strat- tion could lead to income divergence in a egy even lower today. regional neighborhood for a while, before The counterargument is that the range successive waves of lagging countries catch of goods in which a country can develop a up with the leading countries as growth comparative advantage has expanded along spills over to the neighborhood. When the with the growth in global trade. Intermedi- integration process is market driven, as in ate goods and services, more tradable and East Asia, production factors will relocate traded, provide developing countries with and promote convergence in country per a broader range of diversification oppor- capita incomes within the neighborhood tunities than before.3 Empirical evidence (see chapter 3). But when it is institution suggests this is true even for Sub-Saharan driven, as in most developing neighbor- Africa.4 Across individual countries within hoods today, political economy challenges each of nine Sub-Saharan African neigh- can become major concerns.7 borhoods, imports in the previous year of intermediate goods from neighbors are Regional and global integration positively correlated with total exports in imply tradeoffs the current year. As the level of interme- Regional integration agreements, complex diate imports grows larger and crosses a to negotiate, implement, and maintain, threshold, this effect becomes noticeably are intensive in the use of administrative stronger. resources. Efforts to align regional institu- These findings show that higher exports tions through such agreements can come at occur when countries cooperate regionally the expense of domestic administration and (in terms of scale economies, greater fac- unilateral liberalization that can determine tor mobility, and lower transport costs) a country's integration with the rest of the as well as integrating globally. Regional world. Regional agreements also prevent cooperation means that firms in neighbor- countries from pursuing more rapid global ing countries can produce final goods more integration, when some members within a cheaply (by building international supply region want to move more slowly. Winners without Borders 263 BOX 9.2 Diversifying production through regional cooperation Diversifying an economy is no easy task. be facilitated in several ways. Foreign Nearly all members of the West African Hidalgo, Barabasi, and Haussman (2007) direct investment can provide much of Economic and Monetary Union can ben- show that the current export structure the required information and know-how. e t from cooperation in at least seven of a country determines how easy it will So can learning from one's neighbors. product clusters (fruits and vegetables be to diversify its production base over Cooperation between neighboring coun- and their products, wood and its manu- higher-value products. They use the tries can therefore help, providing the factures, cotton, low-tech manufactures, metaphor of a forest representing the scale attractive for foreign investors and chemicals, and minerals) to reduce their product space (the same for all countries the access to critical intermediate goods overdependence on traditional agricul- in the world). Each tree is a product, and that makes the leap to a new product less tural exports, such as co ee and cocoa. rms are monkeys that can climb higher costly and risky. Cooperation can provide Southern Africa Customs Union mem- on a tree to improve their value added an outlet for intermediate goods produc- bers, except for South Africa, can gain (intensive diversi cation) or jump to ers who sell to innovating rms elsewhere signi cantly more than other unions from another tree with higher value (extensive in the neighborhood. cooperation in natural-resource-based diversi cation). When African exports during 1980­ and manufacturing clusters, because they Developing country rms nd it easiest 2004 are mapped against a global prod- have much easier diversi cation options to grow through intensive diversi ca- uct space of some 800 products (four-digit driven by the logistics, nance, skills, and tion, which builds on capabilities they industries), the Central African Economic infrastructure that re ect their middle- already possess. The alternative, required and Monetary Community appears to income status. at higher incomes or in response to even have only a few options for diversi cation By looking at which areas of economic lower-cost competitors, is to jump to (wood and its manufactures). Members activity o er the most promise for further higher value trees. Even if a country is of the East African Community have development, countries can focus coop- lucky enough to have such higher value more options because their exports are eration on sector-speci c infrastructure, trees close to its production base, the more diversi ed (fruits and vegetables, such as common standards, compliance jump remains costly and risky. It may prepared food, sh, wood and its manu- and metrology systems, and speci c cur- require physical infrastructure, speci c factures, cotton, textiles, low-tech manu- ricula to build a skilled labor force and know-how, knowledge of the tastes and factures, metallic products, chemicals, adapt new technologies. That can serve standards in the targeted markets, and and minerals). Other countries with similar as a complement to the general areas of easy and cheap access to speci c inputs. production structures have gone on to cooperation in regional infrastructure, Haussman and Rodrik (2003) called these diversify into such clusters as cotton, tex- better business regulations, and a strong initial investment needs "cost discovery," tiles, and garments, which currently enjoy judicial systems. a search by the rst rms to explore these preferences under the African Growth and Based on contributions from Vandana Chan- new opportunities. Cost discovery can Opportunity Act in the U.S. market. dra, Jessica Boccardo, and Israel Osorio. The regional versus global debate is not welfare improving than multilateral trade new. It revolves around the welfare impli- agreements if intercontinental transport cations of potential trade diversion and costs are much higher than intracontinental trade creation compared with the first-best trade costs.11 There are also noneconomic welfare- improving effects of unilateral lib- gains to regional integration initiatives, eralization or multilateralism.8 Yet a "new such as greater peace and security as well as regionalism" debate has been launched increased bargaining power in international with the recent proliferation of free trade forums.12 These noneconomic motives are agreements. One side of this debate sees in sometimes more important than the eco- regional integration a competitive liberal- nomic in the decision to sign regional inte- ization process that will ultimately support gration agreements. global integration.9 The other side sees the Regional integration can take many emergence of "spaghetti bowls" impeding forms, from formal treaties regulating global integration.10 many aspects of economic exchange and This debate will not be readily con- cooperation to informal, de facto integra- cluded. But the lens of the new economic tion that follows from the private sector­ geography gives it a different perspective. led deepening of economic ties. This variety Some have argued that when physical geog- allows for a different dynamic. While global raphy is properly included in trade models, agreements are comprehensive and rare, regional trade agreements can be more regional agreements can start small and 264 WORLD DEVELOPMENT REPORT 2009 move at a pace and scope with which each Regional integration implies comple- party is comfortable. Each region needs to mentary policy actions by participating find the path that allows it to benefit from countries. The larger the number of par- both regional and global integration. ticipants, the more complex the coordina- tion, with a higher risk of failure. Specific Developed neighborhoods provide agreements based on country interest can useful insights--think big, start small build variable-geometry regional integra- Successful neighborhoods in Europe, North tion in which countries (or areas within America, and Northeast Asia provide three countries as with the "growth triangle" in lessons for the design and implementation East Asia) deepen their cooperation at their of regional and global integration initia- own speed. Such cooperation on trade and tives: think global, start small, and com- nontrade issues can gradually build a stron- pensate the least fortunate. ger neighborhood. This does not preclude Think global. For all developing neigh- specific continentwide initiatives to carry borhoods, the most important export mar- out projects with high fixed costs, such as kets are outside the region. The Republic of launching and maintaining a satellite. Korea, Mexico, and Romania are fortunate to Compensate the least fortunate. be close to one of these large world markets, Regional integration can produce win- but most nations are not. The main goal of ners and losers across countries--at least any regional integration process should thus in the short term.16 If two countries with be to promote sound export-led growth. different domestic infrastructure integrate, Indeed, the success factor of regional inte- the country with the better infrastruc- gration agreements is "open regionalism," ture will attract more industrial activities, setting low external tariffs and suppressing which may deepen differences in income all the internal ones.13 This is a key differ- and employment.17 Building a sustainable ence from the first wave of regionalism in neighborhood of countries with different the 1970s, which simply extended inward- endowments is thus helped by a compensa- looking import-substitution policies from tion mechanism to ensure equitable shar- countries to regions. ing of the gains from integration. In the Start small. Regional integration ini- EU, rich members subsidize infrastructure tiatives do not need to address all issues development in poorer member nations. In immediately. Nor do they need to involve a the Association of Southeast Asian Nations whole continent at once. The Latin Ameri- (ASEAN), richer member countries have can and Sub-Saharan experiences in the programs specifically designed to assist 1970s show that comprehensive agreements poorer member countries--the Integrated involving a large number of countries often ASEAN Initiative. Some regions also have remain "paper agreements."14 The Euro- bilateral aid programs for their poorer pean Union started with a narrowly focused neighbors. agreement--the European Coal and Steel One approach to compensation is pool- Community (see "Geography in Motion, ing customs revenues collected in customs Overcoming Division in Western Europe"). unions and redistributing them accord- The North American Free Trade Agreement ing to each member's development needs. (NAFTA) started with a free trade agree- The West African Economic and Monetary ment for automobiles, between the United Union (WAEMU) adopted a common exter- States and Canada.15 East Asia's regionaliza- nal tariff in 2000, and introduced a 1-percent tion accelerated in the 1980s, with Japanese levy on all third-party imports to build a multinationals setting up manufacturing compensation fund. By September 2006, export platforms across the region. Often $500 million had been collected and shared. regional integration can start without a for- Côte d'Ivoire and Senegal, the richest mem- mal agreement of any kind but with a state- bers of WAEMU, contributed 60 percent of ment of intent for strategic cooperation that the funds but received only 12 percent. Such gives firms comfort that any disputes will transfers are politically feasible if the wealth- be resolved quickly and fairly. ier countries realize that they will benefit in Winners without Borders 265 the long run if their neighborhood prospers. enoughtotriggerorsustainindustrialization, Revenue-sharing initiatives are strengthened or that lack the capabilities to diversify and by the involvement of a developed country as advance up the value chain. Different coun- an external partner willing to subsidize the tries thus face different problems that require process. The Economic Partnership Agree- different policy responses to integrate them ments (EPAs) currently being negotiated into the global economy. Integration happens between the EU and African, Caribbean, largely through private activity in trade and and Pacific countries are examples (see box factor mobility. But most of the institutions 9.3). or infrastructure needed to connect a region to the global economy are public goods, Building integrated requiring collective action to overcome coor- neighborhoods: a framework dination problems and externalities. Three types of policy instruments can be The "thickness" of country borders is a self- used to pursue regional integration. They imposed obstacle to development, with iso- also help with global integration. lation increasing the economic distance to markets (see chapter 3). On top of division, · Institutional cooperation can address coor- some neighborhoods have small countries dination problems within neighborhoods whose local markets are simply not large and foster greater scale economies. BOX 9.3 Economic partnership agreements between the EU and African, Caribbean, and Pacific countries can be made better Until 2007 the EU granted nonreciprocal The goal is ambitious. The EPAs give thus improving political and economic trade preferences to African, Caribbean, incentives to ACP countries to increase stability and creating a better business and Paci c (ACP) countries. This policy regional trade and cooperation, unlike and investment climate. did not comply with the WTO principle of the previous arrangements that favored One of the most di cult issues is the most-favored-nation treatment, but got a a hub-and-spoke structure, discourag- expected loss in tari revenues, which temporary waiver that expired in Decem- ing interaction with neighbors. And are, on average, about 2 percent of gross ber 2007. The economic partnership agree- while the previous trade preferences domestic product (GDP) for Sub-Saharan ments (EPAs) between the EU and the ACP were determined unilaterally by the EU, countries. But for some, the loss can be 4 countries are a new approach to promot- the EPAs are jointly negotiated. Under- to 6 percent of GDP, a sizable fraction of ing trade and achieving more general standably, some countries are unwilling the public purse. A phased reduction in development goals at the same time. to cooperate on issues in which they tari s is designed to mitigate big declines In 2003 the EU started negotiating EPAs might lose. But the EU can provide in government revenues. Over the long with six self-de ned ACP regions: the incentives--like aid--to help overcome term, the lost tari revenues need to be Caribbean (CARIFORUM), Central Africa such di erences. replaced through reforms of domestic tax (CEMAC), Southeast Africa (ESA), West Experience shows, however, that and tax administration. A more radical Africa (ECOWAS), Southern Africa (SADC), (North-South) trade liberalization alone approach would be for the EU to provide and the Paci c. does not promote economic develop- budget support to the most a ected At the core of the EPAs are regional ment. So the EPAs try to improve the countries over a predetermined transition trade agreements between the EU and coherence between trade and devel- period. each of the six regions. The export struc- opment. Besides trade in goods, the Another issue involves complicated ture from these regions to the EU is het- EPAs include trade in services as well as rules of origin that need to be simpli ed erogeneous, often re ecting dependency investment, public procurement, and and liberalized. Technical assistance is on just a few products. But the EPAs are competition law. Although the agree- also needed to enable developing coun- broader in scope. They will extend 100 ments on trade of goods and services tries to ful ll EU standards and stimulate percent duty-free and quota-free market are about mutual--though asymmet- a supply response to enhanced market access into the EU from each region (with ric--trade liberalization, the trade- access. "Aid-for-trade" programs provide simpli ed EU rules of origin) while permit- related issues follow another route. resources for such e orts. ting ACP countries to open their markets They aim to support regional integra- to a lesser extent (on average 80 percent tion by common regional regulation, within 15 years). harmonization, and implementation, Contributed by Sebastian Vollmer. 266 WORLD DEVELOPMENT REPORT 2009 · Regional infrastructure, strategically attractive the neighborhood of a country, linking the neighborhood to the lead- the less attractive the individual country, ing world markets, can reduce transport particularly when its local market is tiny. costs. Now that tariff preferences have fallen, · Coordinated incentives involving all the behind-the-border barriers are more impor- neighborhood's stakeholders and donors tant determinants of the pattern of trade. from the leading world markets can pro- And by aligning domestic and international mote factor mobility and converging standards and institutions, a neighborhood living standards between leading and can improve its attractiveness for foreign lagging countries in the neighborhood. direct investment (FDI) and increase its opportunities for trade, particularly impor- Institutional cooperation tant given the need to connect to regional Behind-the-border reforms. Institutional and global production networks and mar- cooperation--such as mutual recognition kets. For instance, the crisis facing the fish- agreements on technical and business pro- processing sector in Kenya in the 1990s cedures, adoption of international stan- would have been less severe if raw and semi- dards, and macroeconomic convergence processed fish providers in Kenya, Tanzania, frameworks--expands the size of regional and Uganda had all cooperated to adjust to markets, supporting scale economies. EU hygiene standards.20 Many countries Indeed, domestic and foreign firms assess in Sub-Saharan Africa are now aiming for investment opportunities and related gov- such cooperation.21 ernment policies and the business environ- At-the-border policies. Facilitating the ment--such as property rights, regulation, flow of capital, labor, and intermediate taxes, finance, infrastructure, corruption, inputs is a precondition for cross-border and macroeconomic stability--as part production networks. The WTO provides of a package that determines a country's a framework for such liberalization that attractiveness for investment.18 Another permits the scope of agreements to vary. part is the quality of the legal system, Almost all new regional trade agreements which increases equity investments and include provisions on service liberalization, firm sizes.19 These effects spill over even to but some of these services are embodied in countries with better institutional endow- people and require corresponding agree- ments in leading world markets. The less ment on labor mobility, on which there is little uniformity (see table 9.1).22 Movement of labor raises economic and political con- cerns that appear to be far higher than for Table 9.1 Few regional agreements provide for full mobility of labor traded goods or investments, so few agree- Degree of mobility ments provide the kind of mobility required stipulated Agreement for countries and people to benefit fully. Full labor mobility European Union, Agreement on the European Economic Area, Financial and monetary cooperation European Free Trade Association, Australia­New Zealand Closer Economic Relations, Economic Community of West improves capital mobility and increases a African States region's attractiveness to FDI, especially for Market access for certain Caribbean Community, North American Free Trade Agreement, small countries.23 Indeed, small financial groups Europe agreements, Group of Three, and Canada-Chile, U.S.- markets tend to be less competitive and less Singapore, U.S.-Chile, Japan-Singapore Free Trade Agreements efficient because they cannot exploit the sub- Based on GATS mode 4, ASEAN Free Trade Area, Euro-Med Association Agreements, stantial economies of scale in financial mar- with additional provisions New Zealand­Singapore Closer Economic Partnership, or limitations Southern Common Market agreement, and EU-Mexico, kets. Some market segments may be missing, EU-Chile, MERCOSUR, U.S.-Jordan Free Trade agreements and small markets are less able to diversify No effective provisions for Asia Pacific Economic Cooperation Forum, South Asian investments and operational risks. The labor mobility Association for Regional Cooperation, Central European regulatory structure tends to be more costly Free Trade Agreement, and Common Market for Eastern and and of lower quality in small markets, and Southern Africa ancillary services such as credit information Source: World Bank 2004a, updated by the WDR 2009 team. are more difficult to maintain. Regional and Note: ASEAN = Association of Southeast Asian Nations; GATS = General Agreement on Trade in Services; MERCOSUR = Southern Common Market. global trade in financial services is the best Winners without Borders 267 way to cope with being small--by opening administrations gradually align their pro- national markets to foreign financial inter- cedures with EU standards. The goals are to mediaries, by fully or partially integrat- reduce the processing time for traders and ing with a regional financial system, and transporters, reduce facilitation payments, by gradually opening national markets to reduce corruption related to international international capital flows. The benefits of transport and trade, and improve the effec- regional financial integration increase as a tiveness of controls and antismuggling group of countries moves toward a single efforts. The results provide an encourag- currency, a single central bank, and a single ing precedent for replicating and scaling up licensing and regulatory system for financial regional trade elsewhere. services firms.24 But such integration also reduces the policy flexibility in responding Regional infrastructure to shocks. Regional transport infrastructure reduces Efforts beyond borders. Developing the economic distance between trading countries, particularly the landlocked, partners, both within the neighborhood are hurt by high transport costs due to and between the neighborhood and leading expensive and unreliable freight services. world markets. Electricity, water, telephone They have overregulated transport sectors, lines, and Internet access all raise produc- inefficient logistics services, oligopolistic tivity but are severely inadequate in many freight forwarders, as well as roadblocks developing regions (see table 9.2). Many and demands for bribes along international countries could benefit by coordinating corridors.25 Each day a product is delayed and cooperating in infrastructure provi- before being shipped is estimated to trans- sion. Hydropower development launched late into an increase in the distance to its in 1997 by Mali, Mauritania, and Senegal trading partners by 70 kilometers, reduc- lowered costs and improved access, reliabil- ing its trade volume by 1 percent.26 Land- ity, and quality of electricity supply.29 The locked countries, in particular, would enjoy East Caribbean telecommunications proj- greater exports if their neighbors improved ect, implemented in 1998, increased access the quality of their transport logistics and to telecom services, reduced prices, and customs procedures: it is estimated that a increased employment opportunities. one standard deviation improvement in a landlocked country's logistics together with one standard deviation improvement in its neighbors' logistics would raise the land- Table 9.2 Sub-Saharan Africa, South Asia, and the Middle East and North Africa are most locked country's exports by 74 percent.27 affected by unreliable infrastructure, East Asia the least Beyond-the-border institutional reforms World regions facilitating trade and transport in a neigh- EAP ECA LAC MNA SAR SSA OECD borhood can greatly increase the efficiency Delay in obtaining an 19.4 9.3 32.9 53.7 56.3 43.8 9.7 and reliability of logistics chains. Central electrical connection (days) Asia and Sub-Saharan Africa, whose inter- Number of electrical 9.3 14.0 17.8 46.1 121.5 56.4 1.5 national competitiveness is seriously affected outages (days) by high transport costs, are now exploring Value lost due to electrical 2.5 3.1 3.6 4.2 5.6 5.7 2.3 corridor approaches that have worked well outages (% of sales) elsewhere, as in Southeastern Europe.28 Number of water supply 3.5 7.5 14.5 41.7 12.0 37.2 0.3 In 1998 six countries asked for World failures (days) Bank support in designing a regional pro- Delay in obtaining a mainline 15.8 13.4 45.1 49.9 66.3 58.4 9.0 gram of trade and transport facilitation in telephone connection (days) Southeast Europe. By 2004 eight countries Firms using the Web in 23.7 56.7 40.9 34.2 29.2 20.4 80.2 were involved: Albania, Bosnia and Herze- interaction with clients/ govina, Bulgaria, Croatia, FYR Macedonia, suppliers (%) Moldova, Romania, and Serbia and Mon- Source: World Bank ICA database. tenegro. The initiative reduces transport Note: 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 Region; SSA = Sub-Saharan Africa; OECD = costs, fights corruption, and helps customs Organisation for Economic Co-operation and Development. 268 WORLD DEVELOPMENT REPORT 2009 Regional infrastructure is an important neighboring countries to the backbone part of regional integration, but it often infrastructure and free entry by firms into requires considerable outside financial sup- national markets. But the rapid spread of port because the upfront costs can be high. mobile phone coverage in Africa still leaves Cross-border project preparation is com- out many areas (see map 9.1). plex, and individual countries may not have Mobility-enhancing regional infrastruc- local capacity to conceptualize the technical ture. Cooperation in higher education and design and to build a consensus.30 And the training can not only increase the endow- legal and regulatory framework to facilitate ment of skilled workers but also enhance the provision of cross-border infrastructure labor mobility as students from different is often lacking. All these constraints can countriesestablishcross-countrynetworks.31 prevent promising regional infrastructure French cooperation and the EU Commission projects from getting to the bankable stage. sponsor a network of three statistical schools Three types of regional infrastructure and in Abidjan (Ecole Nationale Supérieure related services enhance scale economies, fac- de Statistique et d'Economie Appliquée, tor mobility, and trade between countries. ENSEA), Dakar (Ecole Nationale d'Economie Productivity-enhancing regional infra- Appliqué, ENEA), and Yaoundé (Institut structure. Power, mobile phones, Internet Sous-regional de Statistique et d'Economie connectivity, and major trunk roads can all Appliqué, ISSEA), training highly qualified generate revenue through fees. The produc- statisticians for French-speaking African tivity increases from these infrastructure private and public enterprises.32 Recogniz- services translate into a high willingness ing the importance of mobility-enhancing to pay. Private firms will provide regional regional infrastructures, a high-level panel infrastructure when it is profitable--as of the African Development Bank has pro- with the South Atlantic 3 (SAT3) marine posed centers of excellence in research, ter- cable connecting West Africa to the global tiary education, and vocational training in fiber optic network, or the Regional African collaboration with the private sector.33 Satellite Communications Organization Trade-enhancing regional infrastruc- (RASCOM) public-private partnership to ture. Good transport infrastructure reduces provide satellite telecommunications cov- transport costs, which in turn increases erage in Africa. Regional cooperation can trade flows.34 Some observers have argued provide a sound regulatory framework that there is little potential for intraregional that, for example, permits free access of trade within developing neighborhoods Map 9.1 Mobile phone coverage has spread rapidly in Africa Global System for Mobile communications network coverage January 1999 March 2002 September 2006 Source: Buys and others 2008. Winners without Borders 269 because the small size of economies will not reason for the collapse was that El Salvador create significant trade flows.35 If so, improv- gained much more from regional coopera- ing the quality of regional roads would have tion because of its better infrastructure.39 In no impact on intraregional trade. But recent 1977 the East African Community of Kenya, studies suggest otherwise. Tanzania, and Uganda also collapsed after Trade models show that regional invest- disagreements over the benefits that would ments to pave all the unpaved interstate be received from common regional services roads would increase the intraregional trade such as airline, harbors, and telecommunica- of West African countries threefold--and tions--as well as over ideological differenc- boost the region's trade with the rest of the es.40 Sound compensation mechanisms and world.36 Upgrading the main highway net- better communication about longer-term work in Sub-Saharan Africa could expand gains for all participants can reduce the risk overland trade by about $250 billion over 15 of failure of such initiatives.41 years, with major benefits for the rural poor, Consider a taxonomy that incorporates while requiring about $20 billion for initial the three essential properties of public upgrades and $1 billion annually for main- goods: nonrivalry, nonexcludability, and tenance.37 In Central Asia road upgrades aggregated contributions (see table 9.3).42 could increase trade by half, exceeding the · Nonrivalry implies that several groups or expected gains from tariff reductions or individuals can consume the good with- trade facilitation programs of comparable out diminishing its value. Clean air and scope. Total intraregional trade in Eastern water are common examples. Europe and Central Asia could be increased · Nonexcludability means that no one can 30 percent by upgrading roads in just Alba- nia, Hungary, and Romania.38 be prevented from consuming the good. There is an incentive to leave the cost of Coordinated incentives provision to a third party. Coordinated incentives can address market · Aggregated contributions relate to resource failures and disputes between countries in a pooling to finance public goods. Com- regional association. The Central American monly, the willingness to contribute Common Market, created in 1960 by El Sal- decays over time. vador, Guatemala, Honduras, and Nicaragua, Each of these properties requires a coordi- faced periodic complaints about redistribut- nated response or some mechanism for equi- ing benefits to Honduras and Nicaragua. tably matching benefits and costs, or else the The agreement collapsed in 1969 following good will be underprovided. The quantity conflict between El Salvador and Honduras. and quality of the public good both depend Some studies suggest that the underlying on member contributions. In some instances, Table 9.3 Regional "club goods" can easily be provided because costless exclusion is possible Regional public goods, types, and examples Impure public goods Pure public goods (nonrival, Impact of aggregated contributions nonexcludable) Goods for which exclusion is easy Shared public services Each contribution has the same impact on A clean lake Transnational park Preserving the rain forest the quality and quantity Countries more interested in the good can Curbing the spread of HIV/AIDS Power grid Eliminating transnational contribute more terrorist threats Contribution of weakest member Implementing international financial Airport hub-spoke network Preventing and mitigating determines the quantity and quality standards natural disasters Contributions of weaker members Forestalling the spread of pests Transport infrastructure Providing Internet determines the quantity and quality connectivity Contribution of leading countries Eradication of a disease Satellite launch facility Regional peacekeeping determines the quantity and quality Contribution of strongest member Discovering an effective treatment Biohazard facility Agricultural research and determines the quantity and quality bioprospecting Sources: Sandler 2002, adapted by the WDR 2009 team. 270 WORLD DEVELOPMENT REPORT 2009 each member is equally important. In oth- Development Community (SADC), for ers, the public good depends on the weakest instance, promoted the Southern Africa or strongest member, or some combination. Power Pool to take advantage of the dis- This taxonomy suggests that the nature of tribution of power sources in the region. regional cooperation varies depending on The Central American Electricity Connec- the goal. tion System was initiated in 2005 under the When the regional public good is sensi- umbrella of the Central American Com- tive to the performance of the weaker mem- mon Market (CACM). So an umbrella bers, as in a hub-and-spoke airport network, agreement can spawn smaller agreements, the challenge for the other members is to or small agreements can be consolidated raise the performance of the weaker links into umbrella agreements. The path is a to an acceptable standard. This can be done tactical choice. through cross-country subsidies, as in the In the same vein, there is a choice EU structural funds. In poor neighbor- between starting with aggregate political hoods, foreign aid may be the only feasible agreements, as in the EU enlargement, or way to ensure the provision of such public starting with economic ties, as in East Asia, goods. If the good depends on the best- with ASEAN+3. Both approaches have performing member of the neighborhood, seen success and failure. The United Arab such as targeted agricultural research, the Republic joining Egypt and Syria in 1958 weaker members may be asked to contrib- foundered in part because of its limited ute to stronger members, or foreign assis- economic advantages. The First East Afri- tance can facilitate its provision.43 can Community started in 1967 as an eco- Trust is especially important in regional nomic grouping, but collapsed 10 years later cooperation. For the waters of the Nile, the because of political divisions between the Arab Republic of Egypt and Sudan, two major countries. It has since been revived, countries that were culturally and politi- but the forces for economic and political cally closer, built the Aswan High Dam union remain divided. near their common border instead of coop- erating with Ethiopia, where a dam might The geography of regional integration have been more efficient for the electricity Looking at the world's neighborhoods and water needs of all three countries.44 through the lens of market access highlights International organizations can help build the role of the three major world markets: trust, as in the Aral Sea Basin rehabilita- Europe, North America, and Northeast tion. Another example is the "development Asia, rich neighborhoods where most of diplomacy" used to resolve the Indus River the world GDP is clustered (see chapter 3). Basin dispute between India and Pakistan, Proximity to these markets, the thickness with the World Bank facilitating coopera- of borders, and the fragmentation of world tion. This diplomacy was recognized by the regions reveal the potential market access then­World Bank President Eugene Black of all countries (see map 9.2).48 as "the most important thing the Bank has Adding up the country scores for poten- ever done, by far."45 tial market access produces three broad Specific regional agreements can get types of developing regions: things started, but they can also lead to multiple and at times overlapping agree- · Type 1 countries are in regions close to large ments, weakening coordination. Many world markets, where the market access developing regions need to rationalize their score is dominated by proximity to the regional economic communities and clar- densest areas in the world. They include ify relations with river basin or power pool those on the periphery of the two larg- organizations.46 Broader regional agree- est markets: North America and Western ments can foster trust, provide an institu- Europe. The neighborhoods are Cen- tional framework for compensation that tral America and the Caribbean, Eastern facilitates bargaining, and allow for more Europe, and the Middle East and North effective sanctions.47 The Southern African Africa. Winners without Borders 271 Map 9.2 Density, distance, and division combine to determine access to markets Real market access, relative to the United States in 2003 Real market access relative to the United States, 2003 < 0.040 0.040­0.090 0.091­0.240 0.241­0.910 > 0.910 No data Source: Mayer 2008 for this Report. Note: To compute potential market access: Each country is assigned a score for the size of its own market (real GDP) and the size of international markets with which it can trade. This is computed by weighting the GDP of other countries by the inverse of a measure that combines physical distance, transport costs, and barriers to trade to show how difficult it is to access these markets. The measure, which is expressed relative to the market access of the United States, essentially combines all three spatial dimensions of density, distance, and division into a composite of potential market access. This map is a complement to the map showing foreign market access in box 6.6. For these countries, the major prob- North America, they can serve as a con- lem is division between themselves and duit to accessing markets everywhere. In major markets. The main instruments some regions, like those in South Asia, for integration will be institutional: for- political considerations also preclude mal regional trade agreements, more economic integration of all the countries limited sector-specific agreements (on in the neighborhood. labor mobility or natural resource-shar- These countries have moderate ing), and harmonization of standards market access potential (see map 9.3). and regulations--all implemented with Their distance from major markets or without formal regional bodies. holds down their overall score, but the · Type 2 countries are in regions with big presence of large developing country neighbors far from world markets. They neighbors can offset this score to some include the neighborhoods of the devel- degree. Brazil, China, Nigeria, the Rus- oping world's giants--Brazil, China, sian Federation, and South Africa are India, and South Africa. Although these examples of large emerging economies are potentially large markets, growth has that add considerably to the market not yet been sustained long enough and access scores of their immediate neigh- many domestic distortions remain.49 bors. For countries in these neighbor- Integration with them runs a risk--to hoods, division is compounded by different degrees in different parts of the distance. Appropriate instruments world--of exposing a neighbor to vola- include institutional and infrastruc- tility and of importing inefficiency from ture development, including regionally the large neighbors' domestic structures. shared utilities, transport corridors But because their market potential is and hubs, and a range of other regional attractive to enterprises in Europe and public goods. 272 WORLD DEVELOPMENT REPORT 2009 Map 9.3 Potential access to major world markets distinguishes the developing world's regions Regions with High-income countries Countries close to world markets Large countries far from world markets Small countries far from world markets Source: WDR 2009 team. · Type 3 countries are in regions far from markets pose barriers to the beneficial flows world markets, without a big neighbor. of people, goods, capital, and ideas. They make up the "bottom billion" For each of these country types, the eco- described in Collier (2007) and con- nomic integration strategies and priorities sist of Central Africa, Central Asia and will differ (see table 9.4). As the potential Caucasus, East Africa, the small Pacific for market access becomes lower, the com- Islands, and West Africa. Many of these plexity of the integration problem becomes countries are falling behind because greater, and a broader range of instruments they are trapped in conflict, suffer from is required to manage integration effec- a natural resource curse, are landlocked tively. For each dimension of the integra- with bad neighbors, or are small with tion challenge, this chapter proposes an bad governance. instrument for integration--"an I for a D." A range of countries, mostly small, Some countries do not fit neatly into have low market access potential. Hav- any of these three types, such as Chile and ing to contend with being far from major Russia. Chile is a relatively small country markets, these countries face division, far from major markets. But it has grown distance, and low economic density. In by exporting to world markets without addition to institutional and infrastruc- significant regional integration. Russia is ture instruments, they need coordinated another special case because of its peculiar incentives for regional integration. The economic geography that spans eleven time incentives include transfers from cus- zones, connected to Europe at its most pop- toms unions and other revenue sources, ulated and most developed western part, direct aid, and preferential market and connected with Northeast Asia through access, such as relaxed rules of origin. the inhospitable and sparsely populated Siberia.50 One part of Russia, and some of All three types of countries have much the former Soviet republics with political lower market access potential than rich and economic ties, could be considered a countries, implying considerable potential neighborhood with a big country far from for more effective economic integration. But world markets. But given that its economic their persistent divisions from major world center is in the western part, Russia is more Winners without Borders 273 Table 9.4 An instrument per dimension--a simple framework for regional integration Region or neighborhood Close to world markets With big countries far from world markets Small countries far from world markets World neighborhoods Central America and Caribbean, South America, Southern Africa, East Asia, Central Africa, East Africa, West Africa, North Africa, Middle East South Asia Central Asia and Caucasus, small Pacific Islands Dimensions of the regional International division (1-D) Regional division, economic distance (2-D) International division, economic distance, integration challenge low density (3-D) What policy instruments Integration with large nearby Regional integration Regional integration should facilitate markets Regional and global connectivity Regional and global connectivity Regional compensation mechanisms Priority instruments Institutions Agreements on trade and factor Agreements on trade and factor mobility Agreements on trade and factor mobility mobility within region and with within region and with large markets within region large markets nearby nearby Shared facilities (research, central banks, Regional provision of public goods regulatory bodies) Infrastructure Transport corridors connecting to large Hub-and-spoke infrastructure regional economy Regional power grids, telecoms, water Regional power grids, telecoms, water management management Incentives Subsidized human development investments in lagging countries and areas Productive investments in leading countries and areas Preferential market access Source: WDR 2009 team. appropriately considered close to world Integration options for countries close markets. to world markets Russia also highlights the point that the Market access is essential for growth, concept of market potential is not country- and proximity is an asset for just-in-time wide but more spatially specific. It is con- production. Many examples in car man- venient to measure it as a single number ufacturing and in segments of the gar- for all localities within a country, but many ment industry demand short-term repeat developing economies have areas where orders. Perishable goods (fresh fruits and markets in other countries are potentially vegetables) are easier to export to nearby more accessible than their own domestic markets. Tradable services--such as mar- markets because of poor local infrastruc- keting, research, and complex information ture. Northern areas of Pakistan are closer technology tasks--benefit from frequent to Afghanistan and western China than to face-to-face interaction, easier if the client the major markets in Karachi and Lahore. is nearby. Countries close to world markets Medan in Indonesia is closer to Penang in thus have an intrinsic advantage in con- Malaysia than it is to its own capital city. necting to markets, suppliers, and ideas. The principles of economic integration in Conversely, for the wealthy world regions-- the real world and the use of the instru- Europe, North America, and Northeast ments can be applied as readily at the sub- Asia--neighboring developing regions national level as at the country level (see expand their growth potential as domes- table 9.4). tic markets mature, while also delivering lower-cost platforms for their firms. There The framework in action are mutual gains to regional cooperation What concrete steps can countries take and ongoing processes to further deepen toward regional integration to build bet- integration. ter neighborhoods and increase global The Euro-Mediterranean Forum is a competitiveness? long-standing coordination mechanism 274 WORLD DEVELOPMENT REPORT 2009 between Europe, the Middle East, and and Association Agreements between the North Africa. The Caribbean Basin has EU and the Balkans specify the legal and benefited from privileged access to the U.S. regulatory reforms to be undertaken before market through various preferential trade joining the EU. The Balkans also have schemes, including NAFTA, the Carib- signed an intraregional free trade agree- bean Basin Initiative, and the Dominican ment, the Central European Free Trade Republic­Central America Free Trade Agreement (CEFTA), to replace the patch- Agreement (DR-CAFTA). China, Japan, work of 32 bilateral agreements formerly and the Republic of Korea are intensifying governing their intraregional trade. The their relations with Southeast Asian coun- new agreement simplifies and harmonizes tries through the ASEAN+3 initiative. The rules of origin and extends the trade and long-term benefits are clear to all sides, but transport facilitation initiative launched the short-term risks and adjustment costs in 2000. The region has also established a have to be managed. common power market and signed an open Institutional reform. The key for coun- sky agreement with the EU that could boost tries close to world markets is to undertake tourism. institutional reforms and improve domes- The Balkan region is close enough to tic governance to fully integrate with the the EU to permit tight integration of its large markets nearby. Free trade alone companies into pan-European produc- does not bring the full benefits of integra- tion networks. Governments can facilitate tion. Although Turkey has had a free trade regional production chains linking their agreement with the EU for many years, supply capacity to that of the EU by signing it did not receive significant FDI until it mutual recognition agreements, confor- embarked on major institutional reforms mity assessments, and other trade-related associated with membership talks. The coordination initiatives. Besides trade pro- policies and governance standards in motion, government policies can attract countries close to large world markets direct investment by multinationals to have to converge with those in the nearby help countries move from agriculture and high-income region. Indeed, multinational basic manufacturing to higher technology firms are more likely to locate in a coun- production. In the 1990s El Salvador and try if it has both institutional and physical Costa Rica diversified their exports from connections to a larger market. The large traditional products (coffee for El Salvador market nearby also has a strong incentive and bananas for Costa Rica) by develop- to foster sound policy and governance ing export processing zones, tax incentives, frameworks in nearby small markets to and FDI promotion in high-tech activities. ensure the stability of its neighborhood. They more than doubled their exports These two factors make the coordination in a decade. In Costa Rica and Mexico, of national policies in neighborhoods close human capital and FDI have jointly stimu- to large world markets both desirable and lated knowledge-intensive manufacturing feasible. The prospect of joining the EU activities.52 has accelerated the pace of reform in Cen- Small countries usually lack the eco- tral Europe. And the prospect of better nomic and political weight to bargain access to the U.S. market triggered policy with wealthier regions. But the Caribbean reforms in Mexico long before NAFTA Regional Negotiating Machinery, created took effect.51 in 1997, has the goal of formulating and Institutional reforms include moving implementing a joint Caribbean negotiat- to a sound macroeconomic environment ing strategy in international trade forums.53 that contains inflation and an efficient fis- The countries now have technical special- cal system that does not rely on distorted ists to deal with each area of negotiations trade policies for budget revenues. They in the WTO. The machinery also facilitates also include establishing a sound institu- the transition of Caribbean Community tional framework that limits corruption (CARICOM) countries toward a single and improves governance. The Stabilization market, with a common external tariff as Winners without Borders 275 the basis for a common trade policy. And it distant regions should try to bridge the has been involved in the negotiations of the gap with world markets by reducing border EPA between the EU and CARICOM. barriers, but they suffer from late-mover To enter the world market for tradable disadvantages in major markets. They can accounting and back-office functions, complement their global integration with countries need an efficient telecommuni- efforts to build a stronger regional market cation system and a highly educated work- centered on a large neighbor. force. The small countries of the Caribbean The competitive advantage of neighbor- region have pooled resources to establish hoods with big countries is size: large local the Eastern Caribbean Telecommunica- markets, abundant human capital, and tions Authority (ECTEL) and the Carib- substantial remittances. Economic activi- bean Knowledge and Learning Network ties generating scale economies--such as (CKLN). petroleum and coal products, refineries, Contrast that with the lack of coordina- pharmaceuticals, electric and electronic tion in the Middle East and North Africa. machinery, iron and steel, instruments, The regional economy is based mainly and nonelectrical machinery--benefit from on oil revenue and cannot create enough being concentrated in leading countries jobs for the 4.2 million people added to that have strong agglomeration economies the labor force every year.54 Governments and better market access.57 Because most in the region have started the transition investment in these sectors will go to those to manufacturing and services, but the countries, usually the largest in the region, region's investment climate is still weak. this creates tensions. The challenges are to The Pan-Arab Free Trade Area (PAFTA) balance political and economic concerns and the Arab Maghreb Union (AMU) have between leading and lagging countries, to had little impact on export performance. ensure spillovers of direct and indirect ben- The declining imports from the rest of the efits to lagging countries, and to compete world accompanying the increase in intra- with neighborhoods close to world markets PAFTA and intra-AMU exports suggest and such emerging economic powers as that the agreements have been more trade China and Russia. diverting than trade creating.55 The region Meeting these challenges of division and could take greater advantage of its prox- distance requires institutions to ensure pol- imity to European markets by increasing icies and governance that promote trade, exports of high-value agricultural prod- factor mobility, and regional growth-- ucts, especially in the winter. But agri- and infrastructure to connect lagging and cultural expansion will put pressure on leading countries, link regional economic scarce water resources, so regional agree- centers, and favor regional production net- ments for water management and use are works integrated with the global economy. essential.56 Institutional reform to improve regional integration. The provision of public goods Integration options for countries with big within a region depends on each member neighbors but distant from world markets to a differing degree according to the good A large home market gives countries an (see table 9.3). Although regional coopera- advantage in attracting industrial activi- tion is sometimes seen as a process to be led ties. If this market is also well connected by the strongest member economy, this is to world markets, this advantage is rein- valid only for certain types of regional pub- forced. But the second group of countries lic goods, perhaps peacekeeping, research, is far from world markets. South America and specialized shared infrastructure, such is farther than Central America and the as biohazard facilities or satellite launch Caribbean from the U.S. market and even sites. For other types of goods, mainly net- farther from the EU and Northeast Asian work related, institutional reforms depend markets. South Asia is far from Northeast on the contributions of the weaker mem- Asia. Southern Africa is far from all three bers of the region. In these cases, some large world markets. Countries in these assistance to build the capabilities of weaker 276 WORLD DEVELOPMENT REPORT 2009 member states can promote overall regional incentives, and the growing political crisis integration. poses risks to the effectiveness of deeper Countries grow faster when other coun- regional integration. tries in their neighborhood are also grow- Investments in cross-country infra- ing, as several studies confirm.58 For small structure to connect regional markets. In countries far from world markets but close neighborhoods with big countries distant to a large developing country, their best from world markets, the costs and benefits prospects often lie in growth in the domi- of cross-country infrastructure can differ nant economy.59 Regional growth cen- between large and small countries in the ters are one reason for regional economic neighborhood. Where the distribution of groupings and for regional peer surveil- benefits differs from the proposed sharing lance. What happens in one's neighbor- of the costs, there may be underinvestment hood, good or bad, is too important to one's in such infrastructure. One example is a own development prospects to ignore. landlocked country such as Bolivia or Para- Economic advantage may not be the sole guay that needs access to the coast to export determinant of regional integration pros- its products. International transit agree- pects. Conflict in South Asia after the end of ments guarantee this right to landlocked the British colonial rule in 1947 prevented countries, but since they are not always the neighborhood from taking advantage enforced, support from the international of its market size, more than a fifth of the community, or from regional institutions, world's people. It took four decades before may be necessary. Another example is the trade volumes between India and Pakistan potential for better infrastructure to link passed those of the early 1950s.60 A recent India's northeastern lagging regions and study estimates that trade between India Bangladesh, Bhutan, and Nepal. The South and Pakistan would increase by 405 per- Asia Subregional Economic Cooperation cent if the territorial and political disputes (SASEC) initiative of the Asian Develop- were resolved.61 In 2004 the two countries ment Bank suggests that such cross-border engaged in the "Composite Dialogue" on cooperation can be beneficial for all these peace and security issues, including terror- countries. ism and drug trafficking, confidence-build- Several major cross-border infrastruc- ing, economic and commercial cooperation, ture projects are being developed. The and friendly exchanges in various fields. On Maputo Development Corridor between a broader regional basis, the South Asian South Africa and Mozambique was initi- Association for Regional Cooperation is a ated in 1995 to rehabilitate the primary forum to discuss development challenges, infrastructure network along the corridor such as cooperation in energy production (road, rail, port, and border posts), attract and water basin management. The burden investment in the corridor's catchment area, is on India, the largest country by far in the and provide employment opportunities for neighborhood, to take the lead in promot- disadvantaged populations (see map 9.4). ing the common agenda.62 Its structure, led by South Africa, promotes Zimbabwe's political instability since fast-track design and implementation of 1998 has dimmed growth prospects in the bankable private investment projects and Southern African neighborhood. Attempts public-private partnerships. But it risks fail- to mediate by the African Union and the ing to address the social service needs of local SADC have brought limited results. South communities.63 Some ongoing evaluations of Africa, the largest country in the Southern the corridor show that border-crossing costs Africa Customs Union, has a large inter- and delays are common impediments, pos- est in a stable neighborhood. But the large sibly diverting freight to domestic corridors. rents from natural resources along regional This suggests that more formal institutional transport corridors are realized even dur- cooperation between the countries could ing conflict, though most of the benefits generate additional benefits. are not shared widely. So economic rea- South America has been much more sons may be unlikely to provide enough ambitious in its plans with the Initiative Winners without Borders 277 Map 9.4 Building regional infrastructure in Southern Africa The Maputo Development Corridor Pilgrims Rest Graskop Joburg to Maputo Marble Hall Hazyview Railway line MOZAMBIQUE AREA OF Lydenburg Sabie MAP Groblersdal Numbe Joburg-Maputo Highway Gate Port of Matola/Maputo PPP- BOT completed Witrivier Ressano Coal-based Power Station Garcia Upgrades, PPP 2 transmission lines to Nelspruit Komatipoort Brits Matola completed Malelane Moamba Waterval Boven Middleburg Tshwane BelfastMachadodorp Barberton Maputo (Pretoria) Witbank Piggs Peak Matola Carolina Tshaneni Ogies Namaacha Johannesburg Hendrina SOUTH AFRICA Mbabane Al smelter 500ktpa Chrissiesmeer BHPB completed Manzini Evander Ermelo Mpaka Liquid Fuels & Petro- SecundaBethal chemicals: Sasol Amsterdam Vereeniging Pande-Secunda Gas line Big Bend PPP Sasol completed SWAZILAND Ponta do Standerton Ouro Piet Retief Port Main railroads Main roads Existing gas line Golela Source: MINTEK 2007. for Integration of Regional Infrastructure, GDP, most among the world's poorest, launched in 2000 to promote the integra- and far too many prone to conflict. These tion and modernization of the 12 coun- neighborhoods face divisions and barriers tries' physical infrastructure in the energy, to trade and factor mobility, are distant telecommunications, and transport sec- from major markets, and lack the density tors, with the goal of improving global of economic production to benefit from competitiveness. The initiative focuses on agglomeration economies. Collier (2007) 10 hubs of economic integration across the identifies their populations as the "bot- continent and on harmonizing regulatory tom billion." frameworks. It has identified 40 megaproj- The challenge for countries in isolated ects and hundreds of smaller infrastruc- neighborhoods is to find ways to integrate ture improvement projects for potential regionally and globally. Their geographic financing, with an aggregate cost in the situation implies that the degree of integra- tens of billions of dollars. Implementation tion rarely will be as high as in other coun- has been slow, however. tries, so the prospects for manufactured trade are more limited. Conversely, their Integration options for countries isolation provides them with natural pro- distant from world markets and with tection of their home markets. small neighbors Many of these economies have minerals Central Asia has the highest proportion and other natural resources, such as water, of landlocked countries (see box 9.4) with that can best be exploited on a regional many common problems that could be basis. While there is evidence of growth more effectively tackled through better spillover from resource-rich countries to regional cooperation. The small Pacific their neighbors in Sub-Saharan Africa,64 Islands are the most geographically frag- regional integration is the key to getting mented, making them "sealocked," with resource-led growth going and to spread- limited accessibility to world markets (see ing benefits more broadly. These coun- box 9.5). And Africa between the trop- tries face the triple challenges of division, ics has the largest number of landlocked distance, and density. Addressing them countries, many small in population and will require institutional reform, scaling 278 WORLD DEVELOPMENT REPORT 2009 BOX 9.4 Integration in Central Asia Central Asia has ve landlocked countries: Republic, Russia, Tajikistan, and Uzbeki- and feedback on the design and imple- Kazakhstan, the Kyrgyz Republic, Tajikistan, stan. Then there are the Commonwealth mentation of trade and related policies. Turkmenistan, and Uzbekistan. The coun- of Independent States (CIS), the Collec- The region loses an estimated 3 per- tries vary in population, type of govern- tive Security Treaty Organization (CSTO), cent of GDP annually because of poor ment, and willingness to cooperate with the Economic Cooperation Organization water management. Agreements are also each other and the rest of the world. But the (ECO), and the Special Programme for the needed for oil and gas resources to reach region has established national identities Economies of Central Asia (SPECA). international markets. Many environ- and institutions, avoided violent con icts, The sheer number of regional agree- mental problems remain as a legacy from established the foundations for market- ments illustrate the problems that can the Soviet era, such as radioactivity from based economies, and sustained an eco- arise from a disjointed regional approach. abandoned uranium mines and danger- nomic recovery since the end of the 1990s. Regional initiatives in Central Asia can ous remnants of biological and nuclear Consider many regional institutions foster integration but add duplication and tests. Regional organizations could be and initiatives. The Central Asia Coop- complexity to reform. The ongoing WTO rationalized around these key themes of eration Organization (CACO) comprises accession for many of these countries trade and transport facilitation, water, Kazakhstan, the Kyrgyz Republic, Russia, could help, because the WTO has clear energy, and environment management. Tajikistan, Turkmenistan, and Uzbekistan, rules on regional trade agreements. Also They could develop long-term plans for which merged with EURASEC (Eurasian needed are trade and transport facilitation these issues, bringing civil society and Economic Community) in 2005. The Cen- initiatives and behind-the-border reforms academic institutions into the fray. The tral Asia Regional Economic Cooperation to improve the countries' attractiveness international community could facilitate Initiative (CAREC) comprises Azerbaijan, to FDI and bolster their global integration. the strengthening of institutions with China, Kazakhstan, the Kyrgyz Republic, (Countries with the highest cost of busi- clear mandates and targets. Mongolia, Tajikistan, and Uzbekistan. The ness entry have lower imports, exports, Shanghai Cooperation Organisation (SCO) and FDI in ows.) Regional forums for busi- Sources: Linn and Tiomkin 2006; Broadman comprises China, Kazakhstan, the Kyrgyz ness communities could o er suggestions 2005; United Nations 2005a. BOX 9.5 Integrating the small and distant Pacific Islands with world markets Small island developing states face a Paci c Regional Trade and Economic this preferential access to their markets. great risk of marginalization in the global Cooperation Agreement (SPARTECA), a Without signi cant trade preferences, economy because of their small size, nonreciprocal trade agreement for which the Paci c Islands need other ways to remoteness from large markets, and Australia and New Zealand o er duty- integrate with their large neighbors. More vulnerability to economic and natural free, unrestricted, or concessional access radical approaches, including consid- shocks. And with their fragile ecosystems, for almost all products originating from eration of greater labor mobility, could they are highly vulnerable to domestic the countries of the Paci c Islands Forum. be required. Children in island families pollution and rising seas. Their share in To qualify for preferential access, goods receiving remittances from overseas fam- global merchandise trade fell from 0.4 exported to Australia and New Zealand ily members show strong improvements percent of world exports of goods in 1980 must meet the rules of origin set out in in education and health outcomes, sug- to 0.2 percent in 2003, while their share SPARTECA. gesting labor mobility could be a power- of global services trade remained at 0.7 The textiles, clothing, and footwear ful driver for longer-term development in percent. industry has been a major bene ciary. these countries. One e ort to deal with the special But Australia and New Zealand are plan- problems of small islands is the South ning to adopt free trade by 2010, ending Sources: UNCTAD 2002; SPARTECA 1996. up infrastructure investments, and tar- But these agreements are often poorly geted incentives to encourage regional implemented, their effectiveness tends to integration. be low, and they overlap in responsibilities. Identifying natural neighborhoods for The administrative costs of participating institutional reform. Neighborhoods with in such agreements are high in relation to small countries distant from world mar- the small benefits, given the small size of kets need to focus on specific institutional the participating economies. The African needs that drive their cooperation. There Union has spotlighted the inefficiencies of is no shortage of international agreements. 13 or 14 overlapping regional economic Winners without Borders 279 communities and has called for their cross-border agreements. Cameroon shares rationalization.65 twin cities with West African neighbors, but Regional integration can be rooted in none in its Central African neighborhood. the traditional economic and sociocultural Similarly, local integration initiatives, such interactions within natural neighborhoods, as growth triangles starting in the early as building blocks for broader integration. 1980s in East Asia, can take advantage of Trust can be built on a shared language. the economic complementarities in border- East African countries share Swahili, which ing regions. has facilitated trade in the neighborhood A succession of large coastal cities along for centuries. Free trade was established the Gulf of Guinea spans from Abidjan in between Kenya and Uganda during colo- Côte d'Ivoire to Douala in Cameroon, and nial times.66 West African countries share includes Accra, Cotonou, Lagos, and Lomé. the Dioula, Haoussa, and Peuhl cultures, When discussing "growth champions," it which, nurtured by Islam, developed an may be worth keeping in mind the potential impressive trade network.67 of such multicountry agglomerations, rather Interactions between neighboring areas than thinking of some nations as regional or cities across countries can also provide growth leaders. When seen through the lens the base for broader integration--a form of of economic geography, the regional integra- transfrontier regionalism that could follow tion priorities change to prioritizing regional European models.68 Sub-Saharan Africa infrastructure investments in leading areas has many pairs of large cities that are near that span several countries. each other but separated by a national bor- Regional trade in agricultural products der (see map 9.5). This carries hidden eco- can be another entry point for broader nomic costs that can be overcome through regional integration. This requires a revival Map 9.5 Twin cities for local integration City pairs in bordering regions within 150 kilometers and with more than 100,000 inhabitants Maradi Ziguinchor Ndjamena Sikasso Bobodioulasso Katsina Kousseri Bissau Mubi Maroua Conakry Kissidougou Parakou Shaki Jimeta Garoua Freetown Nzerekore Portonovo Abeokuta Sefadu Man Lomé Lagos Calabar Cotonou Kumba Goma Kigali Brazzaville Bukavu Uvira Bujumbura Pointe-Noire Kinshasa Kigoma Cabinda Kalemie MatadiBoma Lubumbashi Chingola Mufulira Mutare Chimoio Maseru Bloemfontein Source: WDR 2009 team. 280 WORLD DEVELOPMENT REPORT 2009 of regional trade agreements, adequate is the center of the cotton basin, with the cross-country infrastructure, institutional most production. Rails connect Ouagadou- reforms, and nonmarket institutions such gou and the port of Abidjan, and roads link as farmer cooperatives.69 The Horn of Africa all three cities. And many ginneries and could build on its livestock trade, though textile industries are located in the region. security poses a problem.70 West Africa By upgrading and pooling infrastructure could build on cotton, if leading agricul- within a regional industrial development tural areas across several countries can be program, input costs could fall and cotton- integrated into a single, efficient production based industries such as textile and gar- and processing zone: the Sahelian cotton ments could become competitive in the basin in the border region of Burkina Faso, global market. Such an initiative would Côte d'Ivoire, and Mali (see map 9.6). This require a strong commitment from the region, dominated by Dioula ethnic groups, participating countries and support from is anchored by three cities--Bobodioulasso regional associations and the international in Burkina Faso, Korhogo in Côte d'Ivoire, community. and Sikasso in Mali.71 In 2000 the popula- Institutional development to increase tion of this area was 4 million (11 percent scale, support labor and capital mobility, of the total in the three countries), with an and improve market access. Some regions estimated gross regional product of 1,000 have taken concrete steps toward integra- billion CFA francs (10 percent of aggregate tion. ECOWAS has signed protocols for the national GDPs). free movement of people, abolishing visa The areas in this region have comple- and entry permit requirements. In fact, mentary economic endowments. Bobodi- labor mobility has always been a hallmark oulasso has an international airport with of Sub-Saharan Africa, where tradition or storage facilities. Korhogo has a regional colonial laws have favored circular labor airport, a specialized university, and train- mobility. Nomads moved across countries ing centers in agricultural science. Sikasso in response to seasonal climatic change, Map 9.6 West Africa has potential for cotton-led industrial development MAURITANIA MALI NIGER SENEGAL CHAD Dakar Niamey Banjul BURKINA THE Bamako FASO SUDAN GAMBIA Ouagadougou Bissau N'Djamena GUINEA- GUINEA BISSAU BENIN Conakry SIERRA CÔTE Abuja Freetown TOGO LEONE D'IVOIRE GHANA NIGERIA CENTRAL AFRICAN REPUBLIC LIBERIA Cotonou Monrovia Yamoussoukro Tema Porto CAMEROON Lomé Abidjan Accra Novo Douala Yaoundé DEM. REP. EQ. GUINEA OF CONGO CONGO GABON Limit of cotton-producing basin: Means of cotton exports: Cotton exports through the ports: 260 Cotton-producing basin Rail National cotton 200 Heart of the cotton basin Road Sahelian cotton 50 Tons (thousands) Ginning factory Sources: Atlas on Regional Integration in West Africa. ECOWAS; Sahel and West Africa Club/OECD 2006. Winners without Borders 281 while sedentary farmers also moved sea- institutional cooperation as well as spe- sonally in search of supplementary income cific infrastructure among countries in the during the dry season.72 But the skills of neighborhood. the workforce need to improve. For higher Providing a regional public good is less education, technical training, and research, complex politically if it is based on a mutu- cooperation within the neighborhood can ally beneficial and profitable project, as support institutions beyond the means of in much of energy, communications, and individual countries. And a better local irrigation infrastructure. But for small business climate and new opportunities in isolated countries, regional infrastructure regional growth centers may induce Afri- projects require considerable outside sup- can migrants with technical and business port. Traffic volumes are too small in most know-how to return from abroad. parts of Africa for toll roads to be feasible. Invest in regional infrastructure. The And regional infrastructure may be more New Partnership for Africa's Development beneficial for one country, even though (NEPAD) spatial development initiative most of the investment costs are incurred identified the Bas-Congo development cor- in another. In addition to reinvigorating ridor involving Angola, the Democratic public-private partnerships in infrastruc- Republic of Congo, and the Republic of ture, there is a need for scaling up Inter- Congo as a region where deep integration national Development Association (IDA) would have large benefits, based on enor- contributions to regional integration, mous hydroelectric power potential.73 Two systematizing Aid-for-Trade initiatives74 other development corridors also have and rationalizing the interaction between promise in West and East Africa: regional development banks and global · financial institutions. Less than 3 percent The Gulf of Guinea development cor- of all international development support ridor--linking Benin, Côte d'Ivoire, now goes for regional programs.75 Ghana, Liberia, Nigeria, and Togo-- Use coordinated incentives to facili- could integrate West African economies tate regional integration. African coun- through transport and energy. It could tries need to make a strong commitment also connect five large coastal cities with to regional integration, sharing the costs a critical mass of economic activities and and benefits from opening borders in natu- administrative service provision: Abi- ral neighborhoods. Multilateral agencies djan, Accra, Cotonou, Lagos, and Lomé. and donors need to commit to long-term · The Mombasa development corridor-- support of these initiatives by providing linking the Democratic Republic of financial and technical assistance and bet- Congo, Kenya, Sudan, and Uganda-- ter access to markets. Concrete steps can could use established infrastructure be sequenced, gradually ensuring the irre- links such as the Northern Corridor to versibility of policy reforms in leading and unlock natural resources in the Demo- lagging countries. Preferential trade agree- cratic Republic of Congo and southern ments and aid flows could be tied to coop- Sudan. eration among recipient countries, with the Even with more regional infrastruc- proposed EPAs with the EU as one model ture, better human capital, and greater (see box 9.3). In cases in which incentives factor mobility, these neighborhoods still for regional cooperation are insufficient for face being latecomers in the global mar- some partners--such as facilitating access ket, where other developing countries with for a landlocked economy to a port in its low-cost advantage dominate the market neighboring country--conditional aid for basic manufactures. African coun- flows with clear performance targets may tries need to diversify their export base to be required. reduce dependence on natural resources. A key incentive for policy reform in Many of these neighborhoods need to Africa is temporary preferential access to design explicit export diversification strat- OECD markets.76 Africa cannot wait for a egies to capture a larger share of the world big wage difference with Asia before start- market (see box 9.2). Success requires ing to attract greater productive investment 282 WORLD DEVELOPMENT REPORT 2009 BOX 9.6 A contract with Africa? The give and take of the world's biggest development challenge Better understanding of the geogra- regional role. For each of Sub-Saharan ports, transport links, and information phy of development can lead to more Africa's regions, the contract would include and communication technology--in e ective development aid. This Report speci c obligations and actions that the coastal countries, as well as corridor advocates di erent strategies for Africa's encourage regional development. The gov- infrastructure to link coastal and inte- landlocked countries and its resource- ernments of East, West, and Central Africa rior markets. poor coastal economies. The former have would commit to the following: · Preferential access for Sub-Saharan natural disadvantages associated with · Establishing "Regional Economic Areas" Africa's exports, with liberalized rules of geography and a large distance to market that would tie the economic interests origin that encourage regional supply that reduces their potential growth by of leading and lagging countries in chains. as much as half a percentage point per Africa's regional neighborhoods tightly year. But what is unusual in Africa is that Things are already headed in this together and provide a framework for resource-poor coastal countries have direction. In 2007 the Government of the provision of regional public goods. underperformed. These are the types of the United Kingdom, through its Depart- countries that act as engines of growth in · Pursuing freer movements of labor, ment for International Development, other world regions. Africa's growth poles capital, goods, and services within allocated $1.4 billion over the coming are still weak. these areas. decade to e orts by the governments of This Report argues, to exaggerate · Maintaining and protecting access Burundi, Kenya, Rwanda, Tanzania, and somewhat, that development strategies routes between landlocked countries Uganda and to revitalize the East African for leading areas should invest in places, and outlets for trade. Economic Community. The European and strategies for lagging areas should Commission is also adopting a regional The strategy would combine insti- invest in people. Seen through the lens of approach with its economic partnership tutional cooperation, investment in economic geography, the thrust of devel- agreements. But all donors could be regional infrastructure, and coordinated opment assistance to Africa that focused bolder in their approaches. interventions that may require giving up on education, health, and other social The experience of Europe after World some hard-won and jealously guarded infrastructure in the late 1990s seems cor- War II illustrates how national determina- attributes of national sovereignty. rect for the lagging, landlocked countries. tion to prioritize reconstruction coupled In exchange for these actions, bilateral But this assistance appears to focus on with international assistance can pay o . and multilateral development partners the wrong priorities for coastal countries, Regional integration in Europe did not go would commit to the following: which need physical infrastructure and smoothly initially. But encouraged by the better integration with global markets. · A large increase in international nan- tough terms of cooperation in the Mar- A better contract between donors cial assistance for improved social shall Plan, a process of integration that and countries would be to di erentiate services and other life-sustaining infra- would have been impossible a generation approaches across countries depending on structure aimed at raising living stan- earlier, created the largest common mar- their potential market access. This Report dards and creating portable human ket for capital, labor, and ideas today. proposes a tailored approach, which would capital in lagging countries. lay out the rights and responsibilities of · Increased nancial support for growth- Source: WDR 2009 team. countries according to their potential sustaining infrastructure--including and larger export shares, especially with Over the past centuries, East, Central, multilateral trade negotiations at the WTO and West Africa have suffered a series of driving down tariffs at a fast pace. Initia- "formative disasters" (see "Geography in tives such as the U.S. Africa Growth and Motion, Density, Distance, and Division in Opportunity Act and the EU Everything Sub-Saharan Africa"). Today, they pose an But Arms could be extended to all Sub- especially difficult development challenge Saharan countries, with more liberal rules spanning the three development dimen- of origin and a longer time span. This may sions--density, distance, and division. allow at least some of these countries to To reshape their economic geography, the break into world markets and could jump- policy response has to be commensurately start export diversification in African calibrated. A three-dimensional challenge neighborhoods. A "contract with Africa" demands employing all three instruments could be a framework for supporting such of integration--institutions, infrastructure, coordinated incentives (see box 9.6). and incentives. Geography in motion Density, Distance, and Division in Sub-Saharan Africa In November 1884,, Chancellor Otto von Bismarck of Germany convened a meeting of 14 European colonial powers in Berlin. After In ovember 18 1884 Chancellor Otto c foffourr ccenturies of coomppet ion and hostility, the time had come to negot ur ent riesntu competition and hostilit the had gotiate and settle territorial claims. Britain, France, Germany, and settle ial claims. and Porrtugal were and Po tugal were the main players; no Africans were invited. thhe mam in fr ited. Four months later, the borders of African countries had been charted fr ted in a ppattern still recognizable today (see map G4.1). Bismarck's disciplined solution remained until the end of World War I, when in atternr ill eecog hen the League of Nations confiscated Germany's four colonies and gave other colonizers the mandate of governing them. At indepen- the League confi cated y' four colonies and other colonizers the mandate them. indepen- dence in the 1950s and 1960s, Sub-Saharan Africa had almost 50 countries, many of them called "artificial states," with borders de fr ifi cutting across ph sical geographic featur and partitioning ethnic groups into more than one country.1 physical aphic features and thnic than F or Sub-Saharan Africa, the Ber- ago, the Sahara has been an inhos- countries. Chad, Côte d'Ivoire, Ethio- lin conference was just the last pitable desert, separating northern pia, Nigeria, Somalia, and Sudan, for in a long line of what geogra- Africa from what we now call "Sub- example, are fragmented into Islamic phers have termed "formative disas- Saharan" Africa. The global warming and non-Islamic parts. Islam came ters," unfavorably altering the human, of that period had other major conse- from the Middle East into North and physical, and political geography of quences. It cut contact between most West Africa by land and into East the continent, creating continentwide of Africa and the emerging civiliza- Africa by sea. Later, European colo- problems of low density, long distances, tions in the Arab Republic of Egypt nialists brought Christianity. The and divided countries. and the Middle East. It turned Equa- superimposition of these great world torial Africa from a temperate savan- religions on top of traditional beliefs · Low density. Sub-Saharan Africa has nah into a hot and humid place where reinforced the continent's divide and long been a continent where people malaria and yellow fever thrived. may have added to conflict.2 are scarce. Its population in the eigh- Proximity to wild animals and the teenth century was about 90 million. absence of frost--a natural disinfec- Sub-Saharan Africa today suffers Eurasia, with an area about twice that tant--increased human vulnerabil- from the triple disadvantages of low of Africa, had more than five times as ity to diseases. And when Africans density, long distance, and deep divi- many people. But these initial condi- settled in healthy, fertile places, they sion that put the continent at a devel- tions were tragically worsened by the were again displaced by colonial set- opmental disadvantage. These spatial slave trade. Between 1700 and 1810, tlers. Since the 1950s armed conflicts dimensions reduce proximity between an estimated 15 million Africans-- in pre- and postindependence move- economic agents within Sub-Saharan one of every six--were taken to the ments have aggravated the problem Africa, and between Africa and the rest Americas. Some areas were depopu- of refugees. The result: movements of of the world. "Cumulative causation" lated, and many more mired in dev- people that have left many in remote between these forces catches many astating conflict as the price put on areas far from the centers of economic countries in Sub-Saharan Africa in a humans turned Africans against each activity. The long distance to density "proximity trap."3 another. Europe finally put an end to still affects the growth potential of a the slave trade, and replaced it with large part of the African population. Low density is linked to weak colonialism in the eighteenth century. · Deep divisions. The partitioning of agglomeration forces A rapacious trade in men was then Africa in 1884 left the continent with The average population density on the replaced by a plundering of the conti- the most countries per square kilo- continent (77 people per square kilome- nent for minerals. But even with new meter of any region in the world. ter) is among the lowest in the world.4 settlers, the continent still suffered Each African country has an average A sparsely inhabited continent can over- from low density in most places. of four neighbors; in Latin America come this by using its land and people · Long distances. Since the Holocene the average is 2.3. There are also reli- well and by concentrating resources in Age that began about 18,000 years gious divisions, between and within urban agglomerations. But Sub-Saharan 284 WORLD DEVELOPMENT REPORT 2009 Map G4.1 Africa's borders were charted before World War I Tangier International zone SPANISH MOROCCO Madeira I. TUNISIA FRENCH (Portugal) MOROCCO Canary Is. (Spain) ALGERIA LIBYA RIO DE ORO EGYPT Cape Verde Is. (Portugal) F R E N C H W E S T A F R I C A SUDAN ERITREA FRENCH GAMBIA SOMALILAND PORTUGUESE TOGOLAND FRENCH GUINEA BRITISH NIGERIA EQUATORIAL SOMALILAND ETHIOPIA SIERRA LEONE IVORY GOLD COAST COAST LIBERIA GUINEA ITALIAN CAMEROON SOMALILAND Fernando Poo (Spain) RIO MUNI BELGIAN UGANDAKENYA EAST AFRICA São Tomé Seychelles (Portugal) BELGIAN (United Kingdom) CONGO GERMAN Zanzibar I. CABINDA EAST AFRICA (United Kingdom) (United Kingdom) Aldabra Is. (United Kingdom) Comoros Is. ANGOLA NYASALAND (France) NORTHERN Possession of: RHODESIA Belgium Mauritius SOUTHERN MOZAMBIQUE (United France GERMAN RHODESIA MADAGASCAR Kingdom) SOUTHWEST Germany Réunion WALVIS BAY AFRICA BECHUANALAND (France) Italian (United Kingdom) Portuguese SWAZILAND Spain SOUTH United Kingdom AFRICA BASUTOLAND Independent nations Source: WDR 2009 team, based on Pakenham 1992. Long distances raise Africa is the world's least urbanized Divisions between countries in Africa continent, with only one-third of the can distort the pattern of urbanization. transport costs and population living in urban areas in 2000, One simulation suggests that if Africa's reduce factor mobility according to the UN's World Urbaniza- 50 countries were 50 states in one coun- Distance reinforces the effects of low tion Prospects. Due to the lack of regular try, like the United States, the largest cit- population density on productivity in and recent censuses, even this may over- ies would be even bigger than they are Africa. While much is made of Africa's state the urbanization in Africa.5 The today, capable of sustaining diversified distance from world markets, the pri- agglomeration index in chapter 1 gives economies and incubating entrepre- mary problem is domestic--long dis- Africa a score of 30 percent, compared neurship,skills,and innovation.Without tances within countries.6 Table G4.1 with about 50 percent for the rest of the such prospects Africa's skilled labor has indicates thatAfrica has one of the lowest world. migrated to other continents. road densities in the world, second only Geography in Motion 285 Table G4.1 The most distant and divided regions--trading and transport are expensive Trading time Average transport Road density Estimated across borders costs Population in Ratio of number (km2 of road per number of civil for exports ($ per container to landlocked countries to surface area) conflicts, Region (days)a Baltimore)b countries (%)b surface areab (1999)c (1940­2000)d East Asia & Pacific 24 3,900 0.42 1.44 0.72 8 Europe & Central Asia 29 ­ 23.00 1.17 ­ 13 Latin America & Caribbean 22 4,600 2.77 1.52 0.12 15 Middle East & North Africa 27 2,100 0 1.60 0.33 17 South Asia 34 3,900 3.78 1.67 0.85 24 Sub-Saharan Africa 40 7,600 40.20 2.00 0.13 34 Sources: a. World Bank 2006b, p. 44; b. Ndulu and others 2007, p. 101; c. Ndulu and others 2007, p. 29; d. Fearon and Laitin 2003, pp. 7­10. Note: ­ = not available to LatinAmerica.But unlike LatinAmer- third of Africans who live in landlocked significant gains can be expected in ica--where the population lives largely countries, the costs of division are even productivity and poverty reduction. along the coast,making it unnecessary to greater. They must move goods long · Territorial development. The guid- build roads into the interior--Africa has distances over land--expensive,because ance from economic geography is a third of its population in landlocked each 1 percent increase in distance unambiguous: firms and workers countries and even more far from access increases transport costs by approxi- seek agglomeration, and migration is to global markets. Economic distance in mately 0.25 percent.8 And landlocked a natural way to increase density and Africa--in the sense of market access countries must rely on the goodwill reduce distance to markets. Chapter 8 (see chapter 2)--is further lengthened (and efficient investment) of neighbors proposes some principles and priori- by armed conflicts and linguistic diver- for access to ports and markets. ties for countries where lagging areas sity (see map G4.1). Economic distance are sparsely populated and divided has isolated a large proportion of Afri- Meeting the challenge--better along ethnic, linguistic, or religious cans from access to domestic and global urbanization,more domestic lines. Agriculture is one priority, but markets. Physical factors, such as the specialization,and more policies to help leading areas exploit relative absence of navigable rivers and regional integration scale economies may be especially natural harbors, have been serious bar- important in Africa as a latecomer to Africa can reduce the limitations of its riers to trade. Low levels of domestic economic development. poor economic geography. Better urban and international trade, in turn, limit · Regional integration. Given its his- agglomerations can deliver scale efficien- the potential for growth. tory, political regionalism may have cies. Transport links can help domestic to take the lead in African regional Deep divisions raise markets grow. And regional and global integration. The experience of West- transport costs integration can promote trade. Regional ern Europe summarized earlier in this integration, labor mobility, investments Sub-Saharan Africa is a highly frag- Report spotlights the importance of in trade, communication and transport mented continent with many borders, starting small and keeping expecta- infrastructure, and peace and stability many neighbors, and high transport tions realistic. Regional integration should remain high on the agenda. They costs. Africa is as physically close to takes time and will not happen in all create good neighborhoods, and better global markets as is East Asia--about parts of Africa at once. Infrastructure neighborhoods will facilitate investment, 7,500 kilometers--and closer than projects are a good place to start. But Latin America (9,000 kilometers).7 But trade, and factor mobility in a cycle of through regional integration, Africa prosperity. it still costs almost twice as much to can undo some of what Bismarck ship a container to the east coast of the · Urbanization. Contrary to some and his guests did in 1884. Chapter 9 United States from Africa as from other thinking, urbanization, done right, showed that many African countries regions (see table G4.1). Compound- can help development more in Africa have taken the first steps, outlining ing expensive access to global markets than elsewhere. Despite five decades what the rest of the world can do to is costly access to regional markets. It of low-quality urbanization, living help. takes an African exporter about 40 days standards in Africa's cities are much to cross the border into a neighboring higher than in the countryside. If country, compared with 22 days for a urbanization can be managed better, Based on a contribution by Wim Latin American counterpart. For the along the lines proposed in chapter 7, Naudé. Bibliographical Note This Report draws on a wide range of World Bank documents Garg, Kristalina Georgieva, Ejaz Ghani, Marcelo Giugale, and on numerous outside sources. Background papers and Sudarshan Gooptu Jose Luis Guasch, Sumila Gulyani, notes were prepared by Maria Abreu, Stephen Adam, Yusuf Jonathan Halpern, Simon Hay, Rui Hao, Jesko S. Hentschel, Ahmad, M Alva, Alberto Behar, Sarah Boeckmann, Marius Santiago Herrera, Andre Herzog, Kate Higging, Larry Brulhart, Massimiliano Cali, Yang Chen, Michael Clemens, Hinkle, Mun S. Ho, Bernard Hoekman, Vivian Hon, Yukon Kirsten Cornelson, Souleymane Coulibaly, Nicholas Crafts, Huang, David Hummels, Djibrilla Issa, Narasimham Vijay Edward Feser, John Felkner, Kali Glen-Haley, Geoffrey J.D. Jagannathan, Zelena Jagdeo, Rachid Jamali, Emmanuel Hewings, Austin Kilroy, Melissa Klink, Christian Layke, Jimenez, Steen Jorgensen, Abha Joshi-Ghani, Ellis J. Juan, Philip Manners, Robert Margo, Thomas Markussen, Thierry Stephen George Karam, Shigeo Katsu, Phil Keefer, Christine Mayer, John McCombie, Topher McDougal, Claudio E. Kessides, Peyvand Khaleghian, Homi Kharas, Nadim Montenegro, Megumi Muto, Andrew Nelson, B. Nelson, Khouri, Jacob Kolster, Aart Kraay, Alexander Kremer, K. L. Wim Naudé, Park Sam Ock, Jinhwan Oh, Keijiro Otsuka, Krishna, Kathy Krumm, Alice Kuegler, Praveen Kumar, Ken Poole, Lant Pritchett, Diego Puga, David Satterthwaite, Caterina Laderchi, Amina Lahreche, Peter Lanjouw, Hedi Andrey Treyvish, Hirotsugu Uchida, Dirk Willem Te Velde, Larbi, Frannie Leautier, Phillippe Leite, Jeffrey Lewis, Cam Vidler, and Nong Zhu. Maureen Lewis, Johannes Linn, Laszlo Lovei, Xubei Background papers for the Report are available either Luo, William Maloney, Jean-Michel Marchat, Katharine on the World Wide Web www.worldbank.org/wdr2009 or Martingale, Ernesto May, David McKenzie, Elisabeth through the World Development Report office. The views Mealey, Shomik Raj Mehndiratta, Barjor Mehta, Mukesh expressed in these papers are not necessarily those of the Mehta, Abel Mejia, Taye Mengistae, Dino Merotto, Stephen World Bank or of this Report. Mink, Pradeep Mitra, Keiko Miwa, Celestin Monga, Nasser Many people inside and outside the World Bank gave Munjee, Zafer Mustafaoglu, Abdu Muwonge, Ijaz Nabi, comments to the team. Valuable comments, guidance, sup- Mustapha Kamel Nabli, Gobind Nankani, Ashish Narain, port, and contributions were provided by James Adams, Benno N'dulu, Vikram Nehru, Eric Neumayer, Cecile Uma Adusumilli, Theodore Ahlers, Alexis Albion, Junaid Niang, John Overholt, Caglar Ozden, Nicolas Perrin, Abhay Ahmed, Halil Ibrahim Akca, Mohamad Al-Arief, Riad Al Pethe, Axel Peuker, Vidyadhar Phatak, Brian Pinto, Mark Khoury, Asad Alam, Ritu Anand, Candra Fajri Ananda, Povey, Habib Nasser Rab, Madhu Raghunath, Robin Rajack, Inger Andersen, Martin Andersson, Patricia Clarke Annez, Anupam Rastogi, Dilip Ratha, S. K. Rao, Martin Ravallion, Jorge Saba Arbache, Omar Arias, Jean Eric Aubert, Fouad Jean-Louis Reiffers, Jose Guilherme Reis, Frederic Robert- Awada, Jit Bajpai, George Banjo, Luca Barbone, Elena Nicoud, Klaus Rohland, Jamal Saghir, Maurice Schiff, Emily Bardasi, Alberto Behar, Connie Bernard, Anthony Bigio, M. Schmidt, Annemarie Schneider. Harris Selod, Marcelo S. A. Dan Biller, Kate Bird, Alessandro Magnoli Bocchi, Selowsky, Ethel Sennhauser, Claudia Sepulveda, Shekhar Zeljko Bogetic, Laurent Bossard, Carlos Braga, Milan Shah, Zmarak Shalizi, Sudhir Shetty, Vera Songwe, Manju Brahmbhatt, Marie-H. P. Bricknell, Robert Buckley, Thomas Sood, Andrew Steer, Jack Stein, Andy Tatem, Ahmet Tiktik, Buckley, Derek Byerlee, Fernando Cabezas, Francisco Christopher Timmins, Tim Thomas, Vinod Thomas, Peter Carneiro, Mauricio Carrizosa, Shyam Chainani, Nadereh Thomson, Mark Tomlinson, Yvonne Tsikata, Laura Tuck, Chamlou, Jaime Saavedra Chanduvi, Robert Chase, Ken Carolyn Turk, Sanjay Ubale, Sameh Wahba, Linda Van Chomitz, Abderrahmane Chorfi, Luc Christiaensen, Gelder, Juergen Voegele, Hyoung Gun Wang, David Wheeler, Marie-Helene Collion, Hector Valdes Conroy, Jean-Marie Dirk Willem te Velde, Jan von der Goltz, Jorge Wilheim, Cour, Xavier Crépin, Aude de Amorim, Augusto de la John Wilson, Zelai Xu, Ahmet Yaman, Evgeny Yasin, Tito Torre, Christopher Delgado, Christian Delvoie, Gabriel Yepes, Yutaka Yoshino, Shahid Yusuf, Ulrich Zachau, Albert Demombynes, Sebastien Dessus, Jaime de Melo, Mamadou Zeufack, Yan Zhang, and Ekaterina Zhuravskaya. Dem, Banu Demir, Michele de Nevers, Jean-Jacques Dethier, We have also benefited from suggestions and feedback Shantayanan Devarajan, Makhtar Diop, Ndiame Diop, of participants at consultation and discussion workshops Frederique Docquier, Paul Dorosh, Cuney Duzyol, William held during the drafting stage of the report. In particu- Easterly, Ibrahim Elbadawi, Lewis Eldridge, Antonio lar, we would like to thank Olav Seim and Marianne Berg Estache, Warren Evans, Shahrokh Fardoust, Marianne Fay, of Ministry of Foreign Affairs (Norway); Per Ronnas, Wolfgang Fengler, Alexander Ferguson, Francisco Ferreira, Klas Groth, and Christina Hartler of SIDA (Sweden); Melissa Fossberg, Vivien Foster, Bernard G. Funck, Saurabh Pekka Hukka and Elina Levaniemi of Ministry of Foreign 287 288 WORLD DEVELOPMENT REPORT 2009 Affaires (Finland); participants in the Berlin workshop: P. R. K. Murthy, Uma Adusumilli, Vidyadhar K. Phatak, Tanja Boerzel, Mantang Cai, Angel de la Fuente Moreno, G. S. Gill, Satyendra Sinha, C. S. Deshpande, Mukesh Nicola De Michelis, Manfred Fischer, Grzegorz Gorzelak, Mehta, Mridula Krishna, Arun Mokashi, Siddharth Roy, Eduardo Haddad, Steven Haggblade, Michael Hofmann, Veena Mishra, Rajan Divekar, Dilip Karmarkar, Dinkar Robert Kappel, Aehyung Kim, Adama Konseiga, Frauke Samant, Shishirkumar Rai, John Alexander, Ramakant Jha, Kraas, Astrid Kuehl, Rolf J. Langhammer, Ingo Liefner, Rahul Srivastava, Shyam H. Chainani, Gerson D'Cunha, Roman Mogilevsky, Hassen Mohamed, Peter Nijkamp, Jockin Arputham, Abhay Pethe, Niranjan Rajadhyaksha, Nils-Henning Meyer, Ulrich Nitschke, Boris Pleskovic, and Sarosh Bana, Darryl D'Monte, Ashley D'Mello, Dilip Gerhard Ressel; participants in the Marseille workshop: D'Souza, Neha Batura, Sadashiv Rao, M. K. Sinha, A. K. T. Mona Yafi, Jean-Eric Aubert, Gilles Pipien, Rachid Afirat, Chari, Sonia Sethi, Ashish Chandak, Anupam Rastogi, Lavi Erfan Ali, Michel Arnaud, Jean-Dominique Assie, Fouad D'Costa, Ritu Anand, Nasser Munjee, Manju Sood, C. B. Awada, Mohamed Basyouny, AbdelKébir Berkia, Morched Juvekar, and R. Ganesh; participants at the Hyderabad Chabbi, Abderrahmane Chorfi, Simon Compaore, Jean- workshop: Kamal Kumar, M. Govinda Rao, K. L. Krishna, Pierre Elong Mbassi, Kamal Hamdan, Gamal Hamid, Rachid B. G. Verghese, Mahendra Dev, Subhashish Gangopadhyay, El-Jamali, Rabea Kharfi, Hugues Kouadio, Frannie Léautier, Narhari Rao, Sanjaya Baru, C. Rammanohar Reddy, Abdellah Lehzam, Pierre Mayet, Jean-Louis Reiffers, Wafaa Amitabh Kundu, Ravi Bhoothalingam, Amb. S. Narayanan, Sobhy, Kian Tajbakhsh, Pierre Veltz, Xavier Crépin, Luc Ranjana Kumar, Alakh N. Sharma, Rameshwar Singh, Dassonville, Thierry Desclos, Olivier Mourareau, Joan A. K. Singh, Sri. B. C. Mohapatra, R. Radhakrishna, K. Parpal, Elizabeth Peri, Michele Reynaud, Paul-Henri Krishnamurthy, Mahesh Mishra, Santosh Mahrotra, D. C. Schipper, Serge Snrech, Louis-Jacques Vaillant, Nathalie Sah, Surjit Singh, Prabhat P. Ghosh, G. C. Paul, Jang Pangi, Tchoumba Bitnga, Loraine Falconetti, Fabien Marry, and Yaduvendra Mathur, Shubas Chandra Garg, Umesh Sinha, Olivier Lavinal; participants in the Dar-es-Salaam work- Gautam Pingle, S. K. Rao, Mohammed A. Abid, Surendar shop: Tom Okurut, Samuel Wangwe, Elly Manjale, Mr. Reddy, Nirmalaya Bagchi, and Anand Akundy; partici- Raphael Mwai, Audance Ndayizeye, Professor Semboja, pants at the Moscow workshop: Evgeny Yasin, Vladimir Dennis Rweyemamu, Amos Majule, Mr Kweyamba, Mau, Vladimir Nazarov, Ilya Mazayev, Mikhail Dmitriev, Professor Kishimba, Furaha Lugoe, Haji Semboja, Dr Ksenia Yudaeva, Ekaterina Zhuravskaya, Evsei Gurvich, Shechambo, Cosmos Sokoni, Prof Yanda, Marc Kabandama, Evgennii Gavrlienkov, Galina Kurlyandskaya, Nadezhda H. B. Lunogelo, Vivian Kazi, Irene Alenga, Stina Peteson, Kosareva, Lidiya Ovhcharova, Natalia Zubarevich, Vladimir and Monica Hangi; participants in the Abidjan workshop: Drebentsov, Eugenii Gontmacher, Aleksey Prazdnichnykh, Mohamadou Abdoul, Kasségné Adjonou, Stéphane Aka Aleksander Morozov, Andrey Treyvish, Vitaly Shipov, Anghui, Alban Alphonse Emmanuel Ahourse, Jean-Marie Dmitry Aratsky, Tatyana Popova, Irina Makeeva, Oksana Akpoue, Koffi Attahi, Djerkbary Bambah, Toyidi Bello, Sergienko, Oleg Zasov, Marina Vasilieva, and Vladislava Joseph Coulibaly, Konan Kouakou Noel Dahoua, Alhassane Nemova; the organizers of the Ankara workshop, State Coster Diaw, Anne Marcelle Douka, Jacques Esso, Siaka Planning Organization, and participants: Necla Akça, Didem Fofana, Agnès Gnamon-Adiko, Joachim Hunlede, John Akman, Enver Aksoy, Ayhan Atli, Yavuz Cabbar, Abdullah Igue, Prosper Kedagni, Constant Koidou, Hugues Kouadio, Çelik, Ayda Eraydin, Oguz Isik, Serap Kayasu, Vural Kural, Alexane Kouame, Kouassi Jean-Baptiste Kouma, Mibanan Cenk Oguzsoy, Elvan Ongun, Mevlüt Özen, Ibrahim Hakki Hippolithe Kone, Jean Claude Koutou, Gabin Kponhassia, Polat, Gül Polat, Faruk Sahin, Zekeriya Sarbak, Zafer Yavan, Owusu-Bonsu Kwame, Moussa Ladan, Youayou Marital and Hami Yildirim Laguidahi, Konon Jules Lella, Edmond Manouan, Isaac Jeff Lecksell, Siobhan Murray, and Brian Blankespoor F. Mensa-Bonsou, Yao Kouadio Antoine N'Gbala, Jérôme prepared the maps for this Report. Other valuable assis- Aloko N'Guessan, N'Da N'Guessan Kouadio, Sidi Ould tance was provided by Gytis Kanchas, Polly Means, Nacer Cheikh Abdallahi, Guillaume Poirel, Adama Sall, Kanon Mohamed Megherbi, Grace P. Sorensen, Catalina Tejada, Seri, Adama Sissouma, Soumaïla Sogodogo, Fidel Yapi and Roula I. Yazigi. Merrell J. Tuck-Primdahl, Kavita Watsa, Amoncou, and François Yatta; organizers of the Mumbai and Prianka Nandy assisted the team with consultations and and Hyderabad workshops: Infrastructure Development dissemination. Finance Corporation and Bombay Chamber of Commerce Despite efforts to compile a comprehensive list, some who and Industry; and Administrative Staff College of India; contributed may have been inadvertently omitted. The team participants at the Mumbai workshop: Jairaj Phatak, apologizes for any oversights and reiterates its gratitude to Sanjay G. Ubale, Manu Kumar Srivastava, Sanjay Sethi, all who contributed to this Report. Bibliographical note 289 Background papers for the World Development Background notes for the World Development Report 2009 Report 2009 Behar,Alberto.2008."Neighborhood Growth Effects:An Annual Abreu, Maria. 2008."Effectively Dealing with Slums." Panel Data Approach." Alva, M., and A. Behar. 2008. "Factors That Contribute to (or Brulhart, Marius. 2008. "An Account of Global Intra-Industry Detract from) Successful Outcomes in African Regional Agree- Trade, 1962-2006." ments." Cali, Massimiliano. 2007. "Urbanisation, Inequality and Eco- Andersson, Martin. 2007."Spatial Disparities in Taiwan." nomic Growth: Evidence from Indian States." Cali, Massimiliano. 2008. "Urban Agglomeration Policy in Clemens, Michael A., Claudio E. Montenegro, and Lant Pritch- China." ett. 2008. "The Great Discrimination: Borders as a Labor Market Chen, Yang. 2008."China: A Case Study of 1D-2D-3D areas." Barrier." Cornelson, Kirsten. 2008."Egypt and South Africa: A Case Study Coulibaly, Souleymane. 2008. "On the Complementarity of of 1D-2D-3D areas." Regional and Global Trade." Crafts, Nicholas. 2007."European Growth in the Age of Regional Hewings, Geoffrey, E. Feser, and K. Poole. 2007."Spatial/Territo- Economic Integration: Convergence Big Time?" rial Development Policies in the United States." Crafts, Nicholas. 2007."Spatial Disparities in 19th Century Brit- Kilroy, Austin. 2007. "Intra-Urban Spatial Inequalities: Cities as ish Industrialization." `Urban Regions'." Hay, Simon I., Dave L. Smith, and Robert W. Snow. 2008. "Is a Kroehnert, Steffen, and Sebastian Vollmer. 2008. "Where Have Future for Human Malaria Inevitable?" All Young Women Gone?" Kilroy, Austin. 2008. "The Role of Cities in Postwar Economic Mayer, Thierry. 2008."Market Potential and Development." Recovery." Montenegro, Claudio E., and Maximilian L. Hirn. 2008."A New Klink, Melissa. 2008."Nigeria and South Africa: A Case Study of Disaggregated Set of Labor Market Indicators using Standardized 1D-2D-3D areas." Household Surveys from Around the World." Layke, Christian, and Stephen Adam. 2008. "Spatial Allocation Nelson, Andrew. 2007. "Accessibility Model and Population of Public Expenditures in Nigeria." Estimates." Manners, P., and A. Behar. 2007. "Trade in Sub-Saharan Africa Nelson, B., and A. Behar. 2008."Natural Resources, Growth and and Opportunities for Low Income Countries." Spatially-Based Development: A View of the Literature." Markussen, Thomas. 2008. "Policies for improved land use in Roberts,Mark.2008."Congestion and spatially connective infra- developing countries." structure: the case of London in the 19th and early 20th century." Naudé, Wim. 2007. "Density, Distance and Division Spotlight Roberts, Mark, and Uwe Deichmann. 2008. "Regional Spillover on Sub-Saharan Africa." Estimation." Oh, Jinhwan. 2008."Korea: A Case Study of 1D-2D-3D areas." Uchida, Hirotsugu, and Andrew Nelson. 2008. "Agglomeration Roberts, Mark. 2008."Social and Spatial Equity." Index: Towards a New Measure of Urban Concentration." Satterthwaite, David. 2007. "Expanding the Supply and Reduc- ing the Cost of Land for Housing in Urban Areas in Low- and Middle-Income Nations." Te Velde, Dirk Willem. 2007."Regional Integration, Growth and Concentration." Treyvish, Andrey. 2008. "The Downfall of the Soviet Union: A Spatial Explanation." Vidler, Cam, 2008. "Turkey and Russia: A Case Study of 1D-2D-3D areas." Endnotes Overview Employment figures are for 2003, population figures for 2002, and 1. Clemens, Montenegro, and Pritchett (2008), background GDP figures for 2005. All GDP figures are quoted using purchasing paper for this Report. power standard (PPS) exchange rates. 2. Collier (2007). 6. Population figures are for year 2007 (www.citypopulation.de/ 3. In Zambia a person in a settlement with more than 5,000 peo- world.htm). ple is considered urban; in India the threshold is 20,000.To compare 7. The tendency for the populations of the primate or largest urbanization across countries, this Report pioneers a new measure few cities in a country to considerably outweigh the populations of of population density, the "agglomeration index" (see chapter 1). cities down the hierarchy of places is reflected in the tendency of 4. See World Bank 2007k. the rank-size rule to break down at the top end of the size distribu- 5. Fujita 2007. tion of cities. 6. Mukherjee 2007. 8. These definitions of the rank-size rule and Zipf's law are 7. Koh and Chang 2005. approximate. For more precise definitions see Gabaix and Ioan- 8. de Blij 2005. nides (2004). 9. Khanna 2008. 9. Eeckhout 2004. 10. World Bank 2008e. 10. Martin 2005. Even in the United Kingdom, which has a 11. Throughout this Report, what is often called"regional devel- highly spatially and institutionally concentrated financial sector, opment" or "territorial development" is referred to as "area devel- large cities such as Birmingham, Manchester, Leeds, and Edinburgh opment." These policies must contend with even greater spatial have fairly large concentrations of venture capital firms, although unevenness as their scope widens to encompass a country. regional stock exchanges had been abolished by the beginning of the 1970s. Navigating This Report 11. Kim and Margo 2004. 1. http://nobelprize.org/nobel_prizes/economics/laureates/1971. 12. Office of the Deputy Prime Minister 2003. 2. http://nobelprize.org/nobel_prizes/economics/laureates/ 13. Thanh, Anh, and Tacoli 2005. 1971/kuznets-lecture.html. 14. Tiffen 2003. 3. Brown and others 2008. 15. Glaeser and Kahn 2001, p. 21. 16. Baker and others 2005, p. 15. GIM 1 17. Peng, Zhu, and Song 2008, p. 7. 1. See Engerman and Sokoloff (2000a) and Acemoglu, Johnson, 18. Glaeser and Kahn 2001. and Robinson (2002). 19. Estimated from the gross product of the world's major cit- 2. For a more detailed discussion of U.S. economic geography, ies published in www.citymayors.com; Price Waterhouse Cooper see Kim and Margo (2004). 2007. 3. Fogel (1979) suggests that no single technology was respon- 20. This is equal to the excess of GVA measured on a workplace sible for this reduction in cost. For example, he estimates that U.S. basis over GVA measured on a residence basis for London. Calcula- output in 1890 would have been only 4 percent lower if goods were tion is based on data from Office of National Statistics (2006). transported by water rather than rail. 21. Figures on flows of earnings associated with commuting 4. McCallum (1995). come from the U.S. Bureau of Economic Analysis. 22. Quote taken from the glossary of the World Urbanization Chapter 1 Prospects: The 2007 Revision Population Database (http://esa 1. Farvacque-Vitkovic, Casalis, and Eghoff (2007), p. 37. .un.org/unup/.). 2. United Nations/Wilbur Smith Associates (1980), p. 2. 23. Even with time-series analysis, some care needs to be exhib- 3. Siegel (1997), p. 61. ited with the World Urbanization Prospects database. For some 4. NUTS provides a single uniform classification of territorial countries the official definition of an urban area has changed over units for producing regional statistics for the European Union. The time. In China, for example, the urban share in 1999 could have first two administrative levels in most member states correspond been 24 percent, 31 percent, or 73 percent depending on the official to NUTS 2 and NUTS 3. NUTS 1, a larger unit representing the definition of urban population used (Satterthwaite 2007). major socioeconomic regions, often does not correspond to exist- 24. Glaeser, Kolko, and Saiz 2001. ing administrative units within member states. 25. Statistics taken from Price Waterhouse Cooper 2007. 5. The figures in this paragraph are derived from data on total 26. The estimates of urban consumption shares are based on land area, proportion of land area dedicated to agricultural uses, available household surveys of different years, and the per capita total employment, proportion of employment in agriculture, GDP, GDP estimates are in 2000 U.S. dollars for the respective years (the and proportion of gross value added (GVA) generated in agricul- data set is described in detail in Montenegro and Hirn 2008). ture (Eurostat 2005; Belgium's Office of National Statistics 2006). 27. Maddison 2008b. 291 292 WORLD DEVELOPMENT REPORT 2009 28. Kim and Margo 2004. 52. This figure, together with all other historical urban share fig- 29. This is the so-called home market effect of the new eco- ures quoted in this section, is based on the definition of an urban nomic geography, also discussed in chapter 4. The concentra- area as a city with more than 5,000 inhabitants and taken from Bai- tion of economic activity in urban areas also entails additional roch and Goertz (1986).All figures on urban shares and populations productivity-enhancing effects, as discussed in chapter 4. for contemporary developing countries in this section have been 30. This tendency for urbanization to produce a divergence in taken, or calculated using data, from the United Nations (2006c).As basic welfare indicators in favor of urban areas distinguishes the previously noted, using this database to compare changes in urban modern-day developing countries from the industrializers of the share across countries is not as problematic as using it to compare nineteenth and early twentieth centuries, discussed below. levels of urban shares across countries. 31. Based on 2003 data from Maddison (2008b). 53. This sample consists of Belgium, Denmark, Finland, France, 32. European Commission 2001. Germany, Greece, Italy, the Netherlands, Norway, Portugal, Spain, 33. Miles 2007. Sweden, Switzerland, and the United Kingdom. Its composition 34. As measured in constant 1990 international dollars using was determined by the availability of historical data from Bairoch data from Maddison (2008b). and Goertz (1986). 35. An area is defined as predominantly urban (rural) if more 54. Figure 1.13 illustrates a number of countries that, accord- than 50 percent of its population lives in urban (rural) areas. An ing to the United Nations (2006), experienced positive change in area is also classed as predominantly urban if its urban population urban share between 1985 and 2005. A handful of countries with share is between 15 and 50 percent and it has an urban center of populations greater than 1 million experienced "negative urban- more than 500,000 inhabitants (1 million for Japan) representing ization" over this period (for example, Armenia, Estonia, Georgia, in excess of 25 percent of the area's population. Kyrgystan, Latvia, Moldova, Tajikistan and Uzbekistan). Although 36. Some care is required in this case. The increase in the U.S. the definitions of an urban area have not changed for these coun- urban wage premium during this period might have had less to do tries, the pattern of declining urban shares may, in these cases, be with better lives in cities than with the effects of large declines in related to the dissolution of the Soviet Union. A further two of the the relative prices of agricultural commodities (Barro and Sala-I- countries (the Czech Republic and the Slovak Republic) were parts Martin 2004, p. 470). of the former Czechoslovakia, and the dissolution of this country, 37. These estimates of urban consumption and urban shares of along with the transition away from a planned economy, may have population are based on country household surveys of different likewise affected their urban shares. For the remaining six countries years, depending on availability. (Iraq, Mauritius, Papua New Guinea, Sri Lanka, the United Arab 38. As illustrated later for China, however, this is not the case for Emirates, and Zambia), a number have unconventional or chang- all developing countries. Although there is a negative relationship ing definitions of an urban area. For example, Papua New Guinea's across Chinese provinces between the urban share and the ratio of definition of an urban area is a center with 500 inhabitants or more, urban to rural income, this relationship has shifted upward since an unusually small threshold, while the United Arab Emirates' the late 1990s. urban population is measured as the population of nine cities. 39."Access to water" is defined here as any access to improved 55. Satterthwaite 2007. water sources. These sources vary according to the local context, 56. United Nations 2006c. but include wells, communal taps, piped water, and trucks selling 57. This conclusion is based on historical estimate of U.S. urban water. Similarly, "sanitation" refers to a range of qualified items population growth derived by combining historical data on urban such as latrines and outhouses. shares from Bairoch and Goertz (1986) with historical data on 40. The countries included in the sample are Bangladesh, Benin, country population sizes from Maddison (2008b). Bolivia, Burkina Faso, Cameroon, Chad, Colombia, the Domini- 58. Gwatkin and others 2007. can Republic, the Arab Republic of Egypt, Ghana, Guatemala, 59. Gwatkin and others 2007. Haiti, India, Indonesia, Kazakhstan, Kenya, Malawi, Mali, Morocco, 60. Woods 2000. Mozambique, Namibia, Nepal, Nicaragua, Nigeria, Peru, the Phil- 61. Williams and Galley 1995. ippines, Tanzania, Turkey, Uganda, Vietnam, Zambia, and Zimba- 62. Orwell 1937. bwe. 63. Crafts 2008. 41. Gwatkin and others 2007. 64. Williamson 1990. 42. Satterthwaite and others 2007. 65. Stockel 2002. 43. The Economist 2007a. 66. Kim and Margo 2004. 44. Satterthwaite and others 2007. 45. Satterthwaite and others 2007. Chapter 2 46. Data on Delhi's population comes from United Nations 1. Armstrong and Taylor 2000. See, in particular, chapter 8. (2006). 2. Beaumier 1998. 47. Smith 1989. 3. Krugman (1991a), pp. 11­13. 48. Crafts 2008. 4. See, for example, Barro and Sala-I-Martin (2004). 49. Williamson 1982. 5. OECD (2007), p. 14. 50. The Economist 2007a. 6. Blanchard and Katz 1992. 51. Satterthwaite and others 2007. 7. OECD (2006), p. 13. Endnotes 293 8. France's National Statistics Office's Web site (www.insee.fr/fr/ 35. The concentration measure in this section focuses only on default.asp). the level of concentration (density) in the leading area of each 9. Länder National Accounts Working Group (http://www country. It does not provide information for the spatial distribution .statistik-portal.de). of density across the remaining areas. 10. This pattern is familiar to development specialists as the 36. GDP per capita figures are taken from Maddison (2008b). "Kuznets curve," named for the Nobel economist Simon Kuznets. 37. Krugman (1991a), pp. 11­13. Although the Kuznets curve depicts a relationship between a coun- 38. Ingram and Whitehead 2008. try's economic development and its interpersonal income inequal- 39. The source of information is a set of more than 120 house- ity, the hypothesis Kuznets forwarded to explain this relationship hold surveys in 75 countries, with some countries having two years corresponds to that underlying the relationship between spatial of observations. For each country and year, we estimate household inequality and development documented in this chapter. consumption for individual households. In some countries, only 11. Poncet 2006. household incomes are available. Then, we aggregate the household 12. Buys, Deichmann, and Wheeler 2006. consumption into total household consumption for each census 13. Poncet 2005. region of the country. The concentration measure for a country is 14. Poncet 2006. proxied by the highest regional share of total household consump- 15. Tobler 1970. tion. 16. Angeriz, McCombie, and Roberts 2008. 40. The source of information is Yale University's William Nord- 17. Lipsey and Nakamura 2006. haus's Geographically Based Economic Database (http://gecon.yale. 18. Henderson, Kuncoro, and Nasution 1996. edu/). This database of population, gross product, and land area of 19. Deichmann and others forthcoming. subnational areas covers more than 90 countries. 20. Henderson, Shalizi, and Venables 2001. 41. In cross-country comparisons, use of a standardized grid cell 21. U.S. Bureau Census, various years. helps avoid the cross-level fallacy, in which the number of parti- 22. Aghion and Williamson (1998), pp. 144­51. tions can affect the variance of the measure. Like the procedure for 23. Taylor and Williamson 2006. estimating the spatial Gini coefficient, we rank the terrestrial grid 24. Angeriz, McCombie, and Roberts 2008. cells in descending order according to their GDP density and plot 25. Fingleton 2003. the cumulative land area on the horizontal axis against cumulative 26. Deichmann and others 2005. gross product on the vertical axis. Then a polynomial function is 27. Hering and Poncet 2006. fitted to the data to predict the share of GDP within the densest 5 28. Da Mata and others 2007. percent of the country's land area. 29. Paillacar 2007. 42. Spatial Gini coefficients consider the distribution of density 30. Head and Mayer 2006. across a country's entire land area. The coefficient equals zero if 31. Collier 2007. economic density is evenly distributed across space, while it equals 32. Estimates generated from Nordhaus's Geographically Based one if all production takes place within a single terrestrial grid cell. Economic Database (http://gecon.yale.edu/), compiled from infor- Spatial Gini coefficients of GDP density are computed in three mation gathered in 1990. steps. First, grid cells are ranked in descending order according to 33. Time-series data on spatial concentration were assembled density of economic mass (GDP per square kilometer). Second, for 10 countries--some, however, with only two observations. cumulative shares of land area are derived, plus a corresponding Countries in the sample are Canada (1890­2006), Spain (1850­ series of cumulative shares of GDP, by grid cell. Third, these two 2000), France (1801­1999), the Netherlands (1850­2006), Japan measures are plotted against each other to produce a Lorenz curve, (1900­2000), United States (1960­2000), Brazil (1960­2004), Chile and the spatial Gini coefficient is calculated as the ratio of the area (1975­2004), Thailand (1975­2004), Indonesia (1989­2005), and between the uniform distribution line and the Lorenz curve to the the Philippines (1980­2005. Except France, Spain, and the Nether- area below the uniform distribution line. lands, where population is used, economic mass is measured based 43. Krugman 1993. on GDP. Data come from a variety of sources. But, as explained 44. Based on constant 1990 international Geary-Khamis dollars below, this combination of information sources does not consti- in Maddison (2008b). tute a significant problem because the concentration measure--a 45. The coefficient of variation increased from 0.107 to 0.148 ratio--is robust to various indicators. (Crafts 2005). 34. In what follows, the concentration measure is defined 46. Comparison based on GDP per capita figures from Maddi- according to the following method. First, we estimate an area's son (2008b). hypothetical share of national GDP under the assumption that this 47. Roberts 2004. GDP exhibits a uniform spatial distribution. Second, we identify 48. Barro and Sala-I-Martin 1992. the area with the highest actual share of national GDP. Third, we 49. Green 1969. divide the actual GDP share of this leading area by its hypothetical 50. We define the welfare measure as per capita gross cell prod- (uniformly spatially distributed) share to get our measure of spatial uct, and spatial disparity as the ratio of the highest-to-lowest per concentration. For example, if the leading area has an actual share capita gross cell product. Other measures of spatial inequality-- of 10 percent and a hypothetical share of 2 percent, then the mea- such as the difference between the 90th percentile and the 10th per- sure of concentration is equal to 5. centile and that between the mean for the top 10 grid cells and the 294 WORLD DEVELOPMENT REPORT 2009 mean for the bottom 10 grid cells--and coefficients of variation 82. GDP per capita figures for Shanghai and Guizho have been show similar stylized patterns. calculated for 2005 by multiplying the estimates for relative pro- 51. For each country and year, we estimate a Mincerian-type vincial performance by Maddison's estimate of national GDP per regression with per capita household consumption as the depen- capita. The figures are in 1990 international dollars. Comparisons dent variable, a set of location dummies as the independent vari- with Britain are likewise based on Maddison's data from Maddison ables, and standard observable household characteristics (such as (2008b). household size and household head's education, gender, age, and marital status) as control variables. The location variables are broad Chapter 3 census regions. The difference between the maximal and minimal 1. Figures are based on Maddison (2006). Between 1820 and coefficients on the location dummy variables is our disparity mea- 1998, GDP increased from $694.4 billion to $33,725.9 billion in sure in consumption between leading and lagging areas, purely 1990 international dollars. Population during that time grew from because of location. 1.04 billion to 5.3 billion, a fivefold increase at about 1 percent per 52. We estimate the relationship between spatial disparity year. and GDP per capita by pooling the disparity estimates from each 2. These estimates are based on the G-Econ project database household survey in a regression analysis that also controls for the assembled at Yale University (see Nordhaus 2006). The database number of census regions within each country and each country's provides estimates for 1990 of GDP, population, and land area of land area. The inverse relationship between spatial disparities in regular grid cells that are about 100 by 100 kilometers at the equa- consumption and the level of development was found to be robust tor. These figures likely underestimate actual concentration. to various model specifications (to the use of different sets of con- 3. Ohmae 1990. trol variables). 4. The numbers include land borders between nation-states 53. These estimates are in 2000 U.S. dollars for the years of the only. So, borders between colonies in the nineteenth or early twen- household surveys during the 2000s. tieth century are not considered. 54. Yemtsov 2005. 5. Blake 2005. 55. Kanbur and Zhang 2005; Milanovic 2005. 6. The measure summarizes four indicators representing 56. Demurger and others 2002. restrictions to the flow of goods, capital, people, and ideas: (1) 57. World Bank 2007. average tariffs (World Bank data), (2) capital openness (Chinn 58. Georgia was the only country in the region not to enjoy a and Ito 2006), (3) proportion of countries that need a visa to visit reduction in absolute poverty over the period. Douthat, Poe, and that country (Neumayer 2006), and (4) a press freedom index Cutler (2006), p. 52, and World Bank (2005e). that includes information such as Internet filtering (Reporters 59. Ferreira 2000. without Borders for Press Freedom 2007). These measures are 60. GDP per capita figures taken from Maddison 2008b. discussed in more detail later in the chapter. All indicators were 61. Naudé and Krugell 2006. normalized and rescaled from more open to more closed bor- 62. Cárdenas and Pontón 1995. ders and summed. An index based on rank sizes provides similar 63. Hill, Resosudarmo, and Vidyattama 2007. results. The data typically refer to 2005 or 2006, but in a few 64. Wisaweisuan forthcoming. instances they refer to 2004. For a few countries with only three 65. Son forthcoming. of the four indicators were available, the missing values were 66. Based on United Nations (1997) and United Nations replaced by the overall mean. Among the larger countries, this (2005). affected Afghanistan, Cuba, the Democratic Republic of Congo, 67. Garcia-Verdu 2005; Fuentes and Montes 2004. Serbia and Montenegro, Tajikistan, and the Republic of Yemen. 68. Public Administration Research and Consultation Centre The index captures many but not all barriers and divisions. For (PARC) 2004. instance, incompatible laws and regulations can have an impact 69. Government of India Planning Commission 2002b. on country well-being. 70. Sri Lanka National Council for Economic Development and 7. There are several globalization indexes that capture similar UNDP (2005). aspects of international barriers. One example is the KOF Index of 71. Government of Sri Lanka 2008. Globalization (globalization.kof.ethz.ch). However, these indexes 72. Kenya Ministry of Planning and National Development and typically mix country regulations such as tariff levels with outcome UNDP (2005). variables such as the share of exports in GDP. For the illustrative 73. Smith (1976), p. 122. analysis in this chapter, we are strictly interested in how countries 74. Venables 2006. themselves manage economic interaction. 75. Chang 2005. 8. See McCallum (1995) and Helliwell (2002). 76. Hanson 1998b. 9. Anderson and van Wincoop 2003; 2004. 77. Demurger and others 2002. 10. Bhagwati and Srinivasan 2002; Alcalá and Ciccone 2004. 78. He forthcoming. 11. Rodriguez and Rodrik 2000; Slaughter 2001. 79. Venables 2006. 12. World Bank 2007j. 80. Crafts (2005), table 4, p. 59. 13. Kee, Nicita, and Olarreaga (2006) develop measures of 81. Calculations are for 2005, based on data from the China trade restrictions that include some of these aspects. Available for Statistical Yearbook (2006), (http://www.stats.gov.cn/tjsj/ndsj/2006/ a smaller set of countries than the average tariff index, this subset indexeh.htm). shows a high correlation with the average tariff index. Endnotes 295 14. See Chinn and Ito (2006). The Chinn-Ito index of capital 46. Leamer (2007) highlights this point. openness is derived from the IMF's Annual Report on Exchange 47. Bourguignon and Morrisson 1998; World Bank 2007a. Arrangements and Exchange Restrictions (AREAER). It is the first 48. Morrisson and Murtin 2005. principle component of four binary variables reported: existence 49. Maddison 2006. of multiple exchange rates, restrictions on current account, capital 50. Bourguignon and Morrisson 2002. account transactions, and requirement of the surrender of export 51. Collier 2007. proceeds. See Chinn and Ito (2006). The original index values were 52. Cutler, Deaton, and Lleras-Muney (2006) discuss the drivers rescaled between 0 (lowest restrictions among all countries) and of falling mortality rates that led to these increases in longevity. 100 (highest restrictions). The regional averages are unweighted. 53. Morrisson and Murtin 2005. 15. McKenzie 2001; Kose, Prasad, Rogoff, and Wei 2006; Chinn 54. The inequality numbers may underestimate the true dis- and Ito 2006; Henry 2007. parities in human capital outcomes. "Years of schooling" relates 16. Quinn and Toyoda 2006. only to primary and secondary education and therefore has an 17. Bekaerta, Harvey, and Lundblad 2005. inherent maximum value. Many developed countries have essen- 18. Kose, Prasad, Rogoff, and Wei 2006. tially achieved a natural maximum some time ago, but they have 19. Pritchett 2006. made further progress in improving the quality of education and 20. Quantifying these benefits is difficult. But a recent study in expanding tertiary education. Developing countries have been estimates that a 3 percent increase in the labor force of industrial catching up on primary and secondary education, but less so in countries could provide a $356 billion gain (World Bank 2005). other dimensions of human capital. 21. Pritchett 2006. 55. Cohen and Soto 2007. 22. Neumayer 2006. 56. These numbers may overstate improvements somewhat, 23. McKenzie 2007. because no data are available for several countries that are likely 24. Linden, Kraemer, and Dedrick 2007. Another example of the among the worst performers. importance of knowledge-intensive activities is given by Oppen- 57. Baumol 1986; De Long 1988; Barro and Sala-I-Martin 1992; heimer (2008): Of each cup of Latin American­grown coffee that Quah 1996. U.S. consumers buy, only 3 percent of the price goes to the region's 58. Maddison 2006. Pre-1990 data for Eastern Europe are likely farmers. The rest funds genetic engineering, processing, branding, somewhat less reliable. marketing, and other knowledge-based activities, which are mostly 59. Lederman, Maloney, and Servén 2005. based outside the region. 60. Pallage and Robe 2003. 25. Criscuolo, Haskel, and Slaughter 2004. 61. Nordhaus 2006. 26. Brunettia and Weder 2003; Keefer and Khemani 2005. 62. Snow, Craig, Deichmann, and Marsh 1999. 27. Data available online at www.rsf.org. 63. Caselli 2005. 28. Gallup, Sachs, and Mellinger 1999. 64. Easterly and Levine 2001. 29. See Collier 2007. 65. Baldwin, Martin, and Ottaviano 2001. 30. Alesina and Spolaore (2003), p. 81. 66. Puga and Venables 1999; Shatz and Venables 2005. 31. Winters and Martins 2004. 67. Levinson 2006. 32. Bertram 2004; World Bank 2006a. 68. Economic Intelligence Unit 2007. 33. Alesina, Devleeschauwer, Easterly, Kurlat, and Wacziarg 69. Wolf 2004. Others, however, such as Bordo, Eichengreen, 2003. and Irwin (1999), think trade and capital markets are more inte- 34. Collier 2007. grated today. 35. Spolaore and Wacziarg 2005. 70. Leamer 2007. 36. Simmons 2006. 71. According to UN Comtrade data for 2004 and using fig- 37. Blake 2005. ures reported by China--for example, Japan has a trade surplus in 38. Collier 2003; 2007. Miguel, Satyanath, and Sergenti (2004) manufacturing of US$21 billion with China, the Republic of Korea show that the impact can also be in the opposite direction. Based US$34 billion, Thailand US$6 billion, the Philippines US$5 billion, on data for African countries from 1981­99 and using rainfall data and Indonesia US$1 billion. to correct for potential endogeneity, they estimate that a negative 72. New York Times, February 22, 1999. growth shock of 5 percent increases the likelihood of conflict by one-half in the following year. GIM 2 39. See endnote 2 for data source. These data by grid cells may 1. Baldwin and Wyplosz 2006; Crafts and Toniolo 1996. still underestimate concentration, because the estimates are based 2. Baldwin 2008. on data for 1990 and high-concentration areas have likely grown 3. Baldwin 1995. faster than low-concentration areas. 40. Maddison 2006. Chapter 4 41. Bairoch 1982. 1. Smith (1976), p. 21. 42. Tilly 1990. 2. See Glaeser, Kolko, and Saiz 2001; Sinai and Waldfogel 2004. 43. Maddison 2006. 3. Discussion of agglomeration economies dates as far back as 44. See box 0.2 for definitions of WDR regions. Smith's consideration of specialization and the division of labor 45. Disdier and Head 2008. (Smith (1976)); Marshall's information spillovers, searching and 296 WORLD DEVELOPMENT REPORT 2009 matching processes, and input-sharing (Marshall (1890)); and the slow diffusion over space of location- and industry-specific more recently, interindustry supply linkages (Chinitz (1961)); information. learning-by-doing (Arrow (1962)); and cross-fertilization of ideas 27. Carlton 1983; Wheeler and Mody 1992; Carlino 1979; and and innovation (Jacobs (1970)). Hay (1979) find that concentration of industrial activity is favor- 4. McCann (2001), p. 56. able for influencing U.S. firms investment and location decisions. 5. See, for example, Henderson (1986); Henderson, Kuncoro, 28. World Bank 2006c. and Turner (1995); Nakamura (1985); Hammond and von Hagen 29. Burchfield and others 2006. (1994); Sveikauskas, Gowdy, and Funk (1988); Greytak and Black- 30. Instituto de Pesquisa Economica Aplicada (ipeadata.gov.br). ley (1985); Glaeser and others (1992); and Bernstein and Nadiri 31. Data from World Bank (2007j). (1988). 32. Devereux, Griffith, and Simpson (2004) for the United King- 6. Prateen 1988. dom; Ellison and Glaeser (1997) for the United States; and Maurel 7. Junius 1997. and Sédillot (1999) for France. 8. Of 27 estimated scale coefficients, 23 show increasing returns 33. Rosenthal and Strange 2004. to scale, most above 4 percent, robust to various specifications of 34. Kim 1995. the production function (Griliches and Ringstad 1971, and Ring- 35. Duranton and Overman 2002. stad 1978). 36. Henderson, Shalizi, and Venables 2001. 9. Baldwin and Gorecki 1986. 37. Henderson, Shalizi, and Venables 2001. 10. Owen (1983) for Europe, and Scherer (1980) and Hall (1988; 38. Henderson, Lee, and Lee 2001. 1990) for the United States. 39. Krugman (1991b) notes that"the most spectacular examples 11. Industry-level scale returns in the range of 1.30 are found of localization in today's world are, in fact, based on services rather for many U.S. manufacturing industries (see Paul and Siegel 1999), than manufacturing . . . and technology . . . will promote more and Chilean plant-level data for eight industries exhibit a compara- localization of services" (p. 66). ble scale elasticity of between 1.20 and 1.44 (Levinsohn and Petrin 40. In 2004, the rest of the nonfarm employment went to trans- 1999). port and trade (22 percent) and forestry, utilities, and construction 12. Antweiler and Trefler (2002) for 71 countries over the period (7 percent) (Kolko 2007). 1972­92. 41. Audretsch and Feldman 1996. 13. The average scale elasticity estimate of 1.051 is within the 42. Feldman (1994) documents that, in the United States, ser- bounds of estimates in the United States (Basu and Fernald 1997) vices, such as innovative or knowledge-intensive activity, are far and in Canada (Fuss and Gupta 1981). more densely packed than manufacturing. Audretsch and Feldman 14. See Roeger (1995) for the United States, and Martins, Scar- (1996), using a count of new product introductions by the Small petta, and Pilat (1996) for the Organisation for Economic Co-oper- Business Administration in 1982, find that knowledge-oriented ation and Development. industries have more spatially concentrated innovative activity. 15. Junius 1997. Dekle and Eaton (1999) provide evidence that external econo- 16. Roeger 1995. mies from clustering are even stronger for business services than 17. Hughes and Mester (1998) detect that, as scale increases, for manufacturing activities. Feldman and Audretsch (1999) find banks economize on resources to manage risks, preserve capital, that sectors in which R&D, university research, and skilled labor are and signal institutional strength. Calem (1994) and Hughes and important inputs tend to be more spatially concentrated. others (1999) suggest that geographic expansion of bank branches 43. Kolko 2007. produces scale efficiency gains from lower deposit volatility, higher 44. Fafchamps and Desmet 2000. expected returns, and more diversified risk. There is also evidence 45. Desmet and Fafchamps 2006. of increasing returns to scale in data processing activities (for exam- 46. At the upper tail of the distribution, however, the opposite is ple, Hancock, Humphrey, and Wilcox (1999) for payment system). true, with large cities losing jobs in favor of intermediate size urban 18. Bossone and Lee (2004), using a sample of 875 commercial areas (Carlino and Chatterjee 2001; Desmet and Fafchamps 2003). banks drawn from the FitchLBCA's Bank Scope database. 47. Kolko 1998. 19. Named after three economists: Alfred Marshall, Kenneth 48. Martin, R. 2005. Arrow, and Paul Romer (Glaeser and others 1992). 49. See, for example, Holmes and Stevens (2002). 20. Hanson 1996. 50. Head, Reis, and Swenson (1995) and Smith Jr. and Florida 21. See www.referenceforbusiness.com/industries/Textile-Mill/ (1994) find strong preferences of Japanese automotive assem- Women-s-Full-Length-Knee.html. blers to locate near each other to take advantage of backward and 22. Donsky 1990. forward linkages. Hammond and von Hagen (1994) examining 23. U.S. Bureau of Census website. employment in four U.S. cities and three two-digit industries, find 24. Henderson (1986), based on cross-sectional data for the that labor-sharing tends to be stronger in expanding markets while United States and Brazil. asset-sharing is more important in mature markets. Holmes (1999) 25. Jacobs 1970. offers the most direct and compelling evidence of input-sharing as a 26. Henderson, Kuncoro, and Turner 1995. Henderson (1997a) source of agglomeration economies: the pantyhose industry is con- finds the largest effects on productivity from own-industry employ- centrated in North Carolina with 62 percent of national employ- ment are lagged two to five years, because it takes time to learn from ment, and purchased input intensity of 53 percent, compared with neighbors. Jaffe, Trajtenberg, and Henderson (1993) corroborate 40 percent nationally. This pattern is found for other concentrated Endnotes 297 industries as well. Holmes (1999) examines the textile industry, These economists failed, however, to provide the technical solutions finding evidence in support of input-sharing. In particular, he finds associated with the modeling of increasing returns to scale. a positive relationship between the size of the textile industry at a 69. Krugman 1991b. location and the presence of local specialized input suppliers to the 70. Warsh (2006) provides an entertaining and accurate account industry. of intellectual progress in dealing with increasing returns to scale. 51. Costa and Kahn 2000. The advances are based on the special features of ideas, highlighted 52. See Bikhchandani, Hirshleifer, and Welch (1998), and Sobel elegantly in Romer (1994). An idea, once developed, may be used (2000) for surveys of the literature on social learning. costlessly by others as a starting point for new ideas, making it 53. Marshall (1890) emphasizes how cities favor the diffusion "nonrival," unlike labor, capital, land, and other inputs for produc- of innovations and ideas while Jacobs (1970) stresses how the envi- tion. Though ideas are nonrival, however, they are generally nei- ronment offered by cities improves the prospects for generating ther free nor nonexcludable. Coming up with useful ideas usually new ideas. Lucas Jr. (1988) also suggests that learning encompasses, requires effort, and through secrecy or the enforcement of intellec- not just of cutting-edge technology, but incremental mundane tual property rights, it is possible to exclude people from using ideas knowledge (knowing how, knowing who, etc.) through intended to improve products or production processes, even if temporarily. and unintended communications, and cities are the best place for This excludability results in knowledge that confers a monopoly knowledge transmission. power on the creators of the knowledge. By adding knowledge 54. Marshall 1890. explicitly to formulations of economic growth, economists are able 55. Jaffe, Trajtenberg, and Henderson 1993. Jaffe (1986) finds to recognize the centrality of ideas and the importance of increas- evidence for local research and development (R&D) spillovers ing returns, but this also requires recognizing the proliferation of among 432 U.S. firms where the number of patents per dollar of imperfect competition. R&D spending is higher for firms located in areas with above-aver- 71. Dixit and Stiglitz 1977. age R&D spending. 72. Henderson 1974. 56. Rauch (1993) shows that wages are higher where average 73. Black and Henderson 1999; Calem and Carlino 1991; education levels are high, because workers will be more productive Combes 2000; Desmet and Fafchamps 2006; Duranton and Puga and employers will be willing to pay high wages in competing for 2005; Duranton and Puga 2004; Feldman and Audretsch 1999; them. Moretti (2004a) finds a positive effect of college graduates Fujita and Ishii 1999; and Glaeser and others 1992. on a city's wages. Charlot and Duranton (2004) use survey data to 74. Glaeser and Maré 2001; Henderson 1986; Henderson, Kun- show that workplace communication is more extensive in urban coro, and Turner 1995; Henderson 1997b; Henderson 1997a; Hen- areas and that this communication increases wages. derson 2003b; Moomaw 1981; Moomaw 1983; Kolko 1999; Naka- 57. This is known as the home market effects in the literature. mura 1985; Rosenthal and Strange 2001; Rosenthal and Strange 58. Holmes and Stevens (2002) look at the relationship between 2003; Sveikauskas, Gowdy, and Funk 1988. firm size and city size across county-industries in the United States, 75. Black and Henderson 1999; Dumais, Ellison, and Glaeser and interpret their findings as evidence of increased productivity 2002; Eaton and Eckstein 1997; Henderson 1997a; Henderson associated with concentration that allows the plants to become 2003a; and Kim 1995. larger. 76. This has been recognized in geography as long ago as since 59. Holmes 1999; Henderson 2005a. Christaller (1933) and Lösch (1940). 60. Davis and Weinstein (1999), examining production across 77. Duranton and Puga 2000. Japanese prefectures, find that an industry`s production is con- 78. Henderson (1997b) for the United States, Fujita and Ishii centrated where the demand for that industry`s good is relatively (1999) for Japan, and Kolko (1999) and Henderson, Lee, and Lee high. Wheat (1986) and Glaeser and others (1992) find evidence (2001) for the Republic of Korea. that manufacturing employment grows faster in subnational areas 79. Henderson 1997b. with more rapid population growth. Justman (1994) finds strong 80. Glaeser, Scheinkman, and Schleifer 1995; Fafchamps and positive comovements of local supply and local demand for manu- Desmet 2000. facturing industries in U.S. cities. 81. Henderson 2005b and Henderson, Kuncoro, and Nasution 61. See, for example, Shefer (1973); Mera (1973); Segal (1976); 1996. Kawashima (1975); Sveikauskas (1975); Moomaw (1981) and 82. Glaeser and Kahn 2001. Moomaw (1983); and Bartelsman, Caballero, and Lyons (1994). 83. Diversity is measured by the inverse of the Herfindal Index 62. Henderson, Shalizi, and Venables 2001). of local employment shares of different sectors. 63. As Glaeser and others (1992) observe "intellectual break- 84. Feldman and Audretsch 1999. throughs must cross hallways and streets more easily than oceans 85. Fujita and Ishii 1999 and Duranton and Puga 2001. and continents" (p. 1127). 86. Duranton and Puga 2005. 64. Feldman and Audretsch 1999. 87. Chandler Jr. 1977 and Kim 1999. 65. Angeriz, McCombie, and Roberts 2008. 88. Toffler 1980; Naisbitt 1995; Negroponte 1995; and Knoke 66. On this point see Feldman and Audretsch (1999), pp. 1996. 409­29. 89. Feldman and Audretsch 1999 and Venables 2006. 67. On this point see Venables (2006), pp. 61­85. 90. Kolko 2000. 68. There were, of course, some exceptions, most notably in the 91. Glaeser and others 1992; Henderson, Kuncoro, and Turner work of Young (1928) and Kaldor (see, for example, Kaldor 1972). 1995. 298 WORLD DEVELOPMENT REPORT 2009 92. Black and Henderson (2003) show that cities that dramati- cent of migrants to Japan come from other countries in East Asia cally grew (the Phoenix population multiplied 10-fold during and the Pacific. 1950­90) or shrank (the Detroit population halved during the same 6. The largest share of emigrants in Sub-Saharan Africa (more period) are the exceptions, as relative sizes of U.S. cities have been than 63 percent) and South Asia (34 percent) settle in countries stable over the last century. Eaton and Eckstein (1997) and Dob- within the same region, typically in an adjacent country. See Ratha kins and Ioannides (2001) also find a pattern of overall stability in and Shaw (2007). This phenomenon is in part a colonial legacy, France, Japan, and the United States. They observe that the relative because borders were drawn arbitrarily across traditional lands populations of the top 40 urban areas in France (1876­1990) and dividing long-standing trading partners and even people of the Japan (1925­85) remained essentially unchanged. same ethnic groups. 93. Henderson (1997b). Kim (1995) shows a high (0.64) corre- 7. Obstfeld, Shambaugh, and Taylor 2004. lation coefficient of regional localization for two-digit industries in 8. Taylor 1996; Obstfeld and Taylor 2002. the United States between 1860 and 1987 at the state level. Dumais, 9. See Caselli and Feyrer 2007. Ellison, and Glaeser (2002) also find that, for most industries, 10. See de Blij (2005) for a review of the geographic and climatic agglomeration patterns were strikingly stable over 1972­92. Hen- drivers of early migrations. derson (2003a) finds stable specialization patterns over 30 years in 11. See Massey (2003) for a review of the modern history of a study of nine three-digit industries. international migration. 94. See, Dunne, Roberts, and Samuelson (1989b); Dunne, Rob- 12. Comparable statistics of emigration from European coun- erts, and Samuelson (1989a); Davis and Haltiwanger (1992); and tries and Japan are available in Massey (1988). Herzog Jr. and Schlottmann (1991). 13. Massey (1988) calculates a correlation coefficient of 0.59 95. Brezis and Krugman 1997. between the onset of industrialization (proxied by the year when 96. Henderson, Kuncoro, and Turner 1995. rail track exceeded 1,000 kilometers), and the initiation of large- 97. Van der Linde 2003. scale international movement of labor (defined as the period when 98. Kolko 1998. emigration first exceeded 10,000 people). 99. United Nations (2006c), Fact Sheet 9. 14. See the International Organization of Immigration (IOM) 100. Gribbin (2000), pp. 30­31. Web page for updated statistics, at http://www.iom.org. An interna- 101. Quote from a top UN official, Lars Reutersward, Nairobi- tional migrant is defined as a person living in a country other than based director of the UN Human Settlements Program. where he or she was born (Ozden and Schiff 2007). 102. Quoting Lars Reutersward in Ward 2006. 15. Massey 2003. 103. United Nations 2004b. 16. Ratha and Shaw 2007. 104. Gill and Kharas 2007. 17. The notable exceptions are Australia, because of isolation, 105. Au and Henderson 2006a. and Israel, because of the religious and political nature of the deci- 106. Rosenthal and Strange 2004 cited from Hoover and Vernon sion to immigrate. 1959. 18. Experts point out that although appealing to governments in 107. Santos and Shaffer 2004. high-income countries, selective immigration policies are probably 108. McCrae 2006. not transferable to other popular destinations. A crucial element to 109. World Bank 2002a. the viability of these policies in Canada and Australia is the relative geographic isolation of these countries, which makes movement Chapter 5 across their borders easier to monitor and control. 1. Bureau of Transport Statistics 2003. 19. Timmer and Williamson (1998), cited in Massey (2003). 2. While returns to capital vary by place (between urban and 20. The best data currently available are still likely to underes- rural areas, leading and lagging regions, and rich and poor coun- timate the internal mobility of labor. Census and survey data used tries), lower marginal returns in a given location can be explained to track internal migration are notoriously inadequate. Census by inadequate complementary factors (Caselli and Feyrer 2007). and survey instruments typically fail to capture seasonal move- 3. In 2000 China sent only 458,000 migrants abroad, while 120 ment to part-time and temporary jobs; movement of migrants million people migrated internally. In the second half of 1990s, who find jobs in the informal economy or who reside in informal fewer than 300,000 people emigrated from Vietnam, while 4.3 mil- settlements; and movement between rural areas. For a detailed dis- lion people migrated internally. See Anh, Tacoli, and Thanh (2003) cussion of the limitations of data on internal migration see Nam, in Deshingkar and Grimm (2004). Serow, and Sly (1990), Lucas (1997), Bell (2003), and Deshingkar 4. Migration from rural areas to cities accounts for only about and Grimm (2004). a third for the growth of urban areas worldwide. Nam, Serow, and 21. Lucas (1997) cites quantitative studies by Connell and oth- Sly (1990) and Deshingkar and Grimm (2004) show that the largest ers (1976) and Skeldon (1986) to support this assertion. flows of internal migrants in developing countries are from rural 22. Rodriguez 2004. areas to other rural areas. In India, where permanent migration 23. Deshingkar and Grimm (2004) give a detailed description seems to have stabilized, temporary "circular" migration--particu- of biases in data collection instruments that lead to a systematic larly by workers in poor households--is increasing. underestimation of rural-to-rural migration, and particularly of 5. More than half the migrants to the United States come from temporary "circular" migration, which leads to the "invisibility" Central and South America. About the same portion of migrants to of a large portion of internally labor movements to policy mak- the EU-15 come from Europe and Central Asia. And up to 70 per- ers that is captured only in qualitative village-level studies. Lucas Endnotes 299 (1997) points out that the failure of the early literature on inter- 60. McInnis 1966. nal migration to capture the dominant rural-to-rural movement 61. European Commission 2007. can be explained by the greater visibility of urban growth, and 62. Decressin and Fatás (1994) examine regional labor markets by the portrayal of rural sectors in the early theoretical models dynamics in Europe and the United States. They analyze the adjust- as homogenous, within which migration did not confer any real ment mechanisms that a typical region's specific shock triggers in benefits. regional labor markets. In Europe they find that for the first three 24. Borjas 1990; Borjas, Bronars, and Trejo 1992; Dunlevy and years, most of the shock is absorbed by changes in the participa- Bellante 1983. tion rate, while in the United States, it is immediately reflected in 25. Deshingkar and Grimm 2004; United Nations 2006b. migration. 26. Lucas 1997; 2003; McKenzie and Rapoport 2007. 63. Tabuchi 1988. 27. Schiff and Özden 2006. 64. Barro and Sala-I-Martin 1992; Brown 1997. 28. Docquier 2006. 65. Fujita and Tabuchi 1997. However, the authors observe that 29. Lucas 1997. inequality then rose in a third stage, after 1975, as the impact of 30. Solow 1956; Swan 1956. agglomeration economies accelerated income growth in large met- 31. Romer 1986; Lucas Jr. 1988. ropolitan areas, such as Tokyo. 32. A recent comprehensive review of economic theories and 66. De Brauw and Giles (2008) on China, and McKenzie and empirical evidence on internal labor migration is available in Lall, Rapoport (2007) on Mexico. Selod, and Shalizi (2006), on which this subsection is based. 67. Overseas Development Institute 2006. 33. Todaro 1969. 68. Soto and Torche 2004. 34. Harris and Todaro 1970. 69. Timmins 2005. 35. Lucas 1997. 70. Koola and Özden 2008. 36. Cole and Sanders 1985; Packard 1997. 71. Beegle, De Weerdt, and Dercon 2008. 37. Romer (1986) and Lucas Jr. (1988), originally delivered as an 72. Adams Jr. (2006) as cited in Özden and Schiff (2007). address in 1985. 73. Acosta, Fajnzylber, and Lopez (2007) in Özden and Schiff 38. Lucas Jr. 1988. (2007). 39. Lucas Jr. (1988) with an acknowledgment of Jacobs (1970). 74. Yang 2008. 40. Moretti 2004b; Rosenthal and Strange 2004; Ciccone and 75. World Bank 2007k. Peri 2006. 76. Garrett 2005. 41. As discussed in other chapters of this Report, urban econom- 77. De Brauw and Giles 2008. ics provide a range of arguments in favor of clustering that are inde- 78. Zhu and Luo 2008. pendent of spillovers to human capital. For example, cities provide 79. Iliffe (1995) on the historical impact of drought on popu- the large markets for producers to exploit productivity gains from lation distribution in Africa; Bryceson (1999) on the Sahel and scale, as shown by Faini (1996), as well as "thick" labor markets that Sudan; Hardoy and Satterthwaite (1989) on Mauritania. lower hiring costs for employers. Clustering workers adds to growth 80. Wandschneider and Mishra (2003), cited in Deshingkar and by increasing backward and forward linkages, as shown by Adelman Grimm (2004), on the drought-induced migration of 60,000 peo- and Robinson (1978) in a study of the Republic of Korea. ple out of Bolangir, in the Indian state of Orissa, in 2001. 42. Clemens, Montenegro, and Pritchett 2008. 81. Venables and Kanbur 2005. 43. Ratha and Xu 2008. 82. Deshingkar and Grimm 2004; Rodriguez 2004. 44. Pritchett 2006. 83. Docquier, Beine, and Rapoport 2006. 45. Andrews, Clark, and Whittaker 2007. 84. Docquier 2006. 46. Cited in Andrews, Clark, and Whittaker (2007). 85. Easterly and Nyarko 2008. 47. In the United States, rates of internal migration among the 86. Sahn and Stifel 2003. native born declined from 1850 to 1940 but accelerated significantly 87. Anderson and Pomfret 2005. thereafter coinciding with the rapid rate of economic growth after World War II. See Rosenbloom and Sundstrom (2003). Chapter 6 48. Schmertmann 1992. 1. Some of these trends were due to the rise in trade with Europe 49. Fujita and Tabuchi 1997. as a share of GDP. But non-European trade declined even as a share 50. Xenos 2004. of GDP. 51. Anh 2003. 2. Krugman 1991b; see also Krugman 2007. 52. Timmins 2005. 3. See, for example, Antweiler and Trefler (2002), who find that 53. Au and Henderson 2006a; 2006b. international trade is fueled as much by scale economies as by dif- 54. Shukla and Stark 1986. ferences in factor endowments. 55. For evidence on nineteenth-century labor market integra- 4. Limão and Venables 2001. tion in the United States, see Lebergott (1964) and Margo (2000). 5. Gallup, Sachs, and Mellinger 1999. 56. Coe and Emery 2004. 6. World Bank 2007d. 57. Hunt 1986. 7. Anderson and van Wincoop 2004. 58. Gallaway and Vedder 1971. 8. It has also meant a change in long-distance trade, mainly in a 59. Rosenbloom and Sundstrom 2003. rise in manufacturing exports (based on differences in labor abun- 300 WORLD DEVELOPMENT REPORT 2009 dance) and a fall in the share of agricultural goods (based on cli- 55. Hulten, Bennathan, and Srinivasan (2003), summarized matic differences) from developing countries. in Hummels, Lugovsky, and Skiba (2007). Scale economies in the 9. Rostow 1960. manufacturing sector derive from the benefits of many indepen- 10. Williamson 1974. dent producers with relatively small market shares moving close 11. Mitchell 1964. together and interacting (monopolistic competition). Indivisibility 12. Crafts, Mills, and Mulatu 2007. and network effects, by contrast, favor monopolies or oligopolies 13. Metzer 1974. in which one or a few providers generate scale economies that are 14. Hurd 1975. internal to a private or public transport provider that dominates a 15. Roy 2002. port, airport, or railway. 16. Mohammed and Williamson 2004. 56. Estache, Gonzalez, and Trujillo 2002. 17. Harley 1980. 57. Hulten 2007. 18. O'Rourke and Williamson 1994. 58. Canning and Bennathan (2007), for example, find a large 19. Yasuba 1978. variation across countries but a general tendency of middle-income 20. Krueger 2006. countries to have a deficit in infrastructure investment, in particular 21. Baldwin and Martin 1999. road infrastructure. The macroeconomic rates of return calculated 22. Krugman 2007. in these studies derive from the agglomeration and specialization 23. Combes and Lafourcade 2005. Costs are approximated by benefits provided by the lowering of transport costs. These benefits ad valorem shares of trade values between geographic regions and increase with an increasing level of development. See Estache and distance and take into account that the real price for transportation Fay (2007) for a review of these studies. depends on the physical transportation network, technology, traffic 59. Brushett 2005. conditions, and structure of the market for transport services. 60. Wilson, Mann, and Otsuki 2004. 24. Ivaldi and McCullough 2007. 61. Wilson, Luo, and Broadman 2006. 25. Gordon 1990. 62. World Bank 2008. 26. Hummels 2007. 63. Evans and Harrigan 2005; Harrigan and Venables 2006. 27. United Nations Conference on Trade and Development 64. Martinez-Zarzoso and Marquez-Ramos 2007. (various years). 65. Arvis, Raballand, and Marteau 2007. 28. Tolofari 1986. 66. Arnold 2006. 29. Button 1999. 67. World Bank 2008. 30. United Nations 1970. 68. Fischer, Harrington, and Parry forthcoming. 31. Gilman 1984. 69. Nordhaus and Boyer 2000. 32. Tramp prices are set in competitive markets and quoted in 70. Tol 2005. U.S. dollars. Two deflators are used to compute the development of 71. World Bank 2008. real prices: the U.S. GDP deflator and the price index for bulk com- modities that typically are shipped by tramp. GIM 3 33. Levinson 2006. 1. Allen and others 2007. 34. Hummels 2007. 2. Smith (1979), p. 189. 35. Harrigan 2005. 3. Kuroda 2007. 36. Hummels 2001. 37. Evans and Harrigan 2005. For a similar analysis on trade Chapter 7 between East and West Europe compare Nordas, Pinali, and Grosso 1. Ravallion 2007. (2006). 2. Soto 2000. 38. Harrigan and Venables 2006. 3. De Long and Shleifer 1993. 39. Krugman 1995; Venables 2001. 4. A passage explicitly mentioning the land law is found in 40. Freund and Weinhold 2004. North (1971), p. 123. 41. Fink, Mattoo, and Neagu 2005. 5. Wordie 1983. 42. Rauch 1999. 6. Craft 1989; Turner 1986; Wordie 1983; Wrigley 1985. 43. Antras, Garicano, and Rossi-Hansberg 2006. 7. Craft 1989. 44. Rajan and Wei 2004. 8. Sokoloff and Engerman 2000b. 45. Amiti and Wei 2005. 9. Sokoloff and Engerman 2000b. 46. Bairoch 1988. 10. Rodrik, Subramanian, and Trebbi 2004. 47. Christaller 1966. 11. Acemoglu, Johnson, and Robinson 2001. 48. Feenstra and Hanson 1997. 12. This discussion of Mumbai is drawn from Buckley, Bertaud, 49. Evans and Harrigan 2005. and Phatak 2005; Buckley and Kalarickal 2006. 50. Duranton and Puga 2004. 13. United Nations 2001; www.mymoneyblog.comarchives/2006. 51. Leamer and Storper 2001. 14. World Bank 2007j; purchasing power parity adjusted per 52. Storper and Venables 2007. capita gross national income (GNI). 53. Winston 1993. 15. Brueckener 2007; Henderson 2007. 54. Nikomborirak 2007. 16. Buckley and Kalarickal (2006), box 3.4. Endnotes 301 17. In 2002, a Ministry of Cities was established to overhaul 57. World Urbanisation Prospects: The 2007 Revision Population the regulatory framework allowing cities to more effectively man- Database (http://esa.un.org/unup/). age land use and subdivision and be more responsive to changes in 58. Cohen and Soto 2001. market needs. 59. The discussion that follows is largely based on Oh (2008), a 18. World Bank 2007b. note prepared for this Report. 19. Chiquier, Hassler, and Lea 2004. 60. The discussion that follows is based in part on Chen (2008), 20. World Bank 2006d. a note prepared for this Report. 21. Lall and others 2004. 61. Figure derived from China Statistical Yearbook 2007. 22. World Bank 2002a. 62. World Bank 2003b; 2007k; Bertaud 2004. 23. The discussion is based on the historical collection of transport 63. World Urbanisation Prospects: The 2007 Revision Population pamphlets the London School of Economics library has converted Database (http://esa.un.org/unup/). into digital format and made available for public download (www.lse 64. Human Development and Reform Commission 2005; .ac.uk/library/pamphlets/Transport/transportpamphletpages/urban_ Changsha-Zhuzhou-Xiangtan Cluster Regional Plan. transport.htm). These pamphlets cover London transport history 65. Kilroy 2007. (eight pamphlets), transport policy in London (nine pamphlets), the 66. Lall 2005. tramways (three pamphlets), and urban transport policy more gener- 67. Lall 2005. ally (three pamphlets). These pamphlets were issued by a variety of 68. Yepes and Lall 2008. organizations, including private-sector providers of transport, local 69. Sohn 2003, based on Oh 2008. political parties, and what might now be called think-tanks. See Rob- 70. Evans 2002, Sohn 2003, respectively. erts (2008), a background paper for this Report. 71. World Urbanisation Prospects: The 2007 Revision Population 24. Hargan 2007. Database (http://esa.un.org/unup/). 25. Land Enquiry Commission 1914. 72. Maddison 2006. GDP per capita is quoted in constant 1990 26. Offer (1981), p. 291. international dollars. 27. GDP per capita figures have been taken from Maddison's 73. Comparisons based on urban agglomeration population (2007) historical database (http://www.ggdc.net/maddison/). They data from World Urbanisation Prospects. are quoted in 1990 international Geary-Khamis dollars. 74. Cohen and Soto 2007. 28. Bairoch and Goertz (1986), table 3, p. 288. 75. World Bank 2007j. 29. Staff City Population Database, Human Settlements 76. Neuwirth 2007, based on Vidler 2008. Group, International Institute for Environment and Development 77. BBC World Service, Thursday, April 27, 2000 (http://news. (IIED). bbc.co.uk/1/hi/sci/tech/727966.stm). 30. Mumford 1963. 78. The discussion of Turkey above is based on Vidler (2008), a 31. Dunlap 1992. note prepared for this Report. 32. Lane 1860. 79. Bertaud 2003. 33. O'Gorman 1939a; 1939b. 80. Government of India Planning Commission 2002a. 34. Santos and Shaffer 2004. 81. Selod 2007. 35. Meyer 1917. 82. Tinbergen 1952. 36. Victoria Transport Policy Institute 2007. 37. Pucher and others 2005. Chapter 8 38. World Bank 2002a. 1. Azzoni, Menezes-Filho, Menezes, and Silveira-Neto (2000), 39. Metro Bits (http://mic-ro.com/metro). World Bank 1998. 40. Booth 1901. 2. World Bank (2007b), figure 8.2 comes from this Report. 41. Abreu 2008, a note prepared for this Report. 3. World Bank 2007a. 42. Nesslein 2003. 4. Estimate based on figures reported in Maddison (2008). In 43. Borgegård and Kemeny 2004. 1700 India's GDP was $90,750 million compared with a world GDP 44. Hall and Vidén 2005. of $371,369 million in 1990 international dollars (Maddison 2008). 45. Swedish Council for Building Research 1990; Borgegård and 5. In landlord systems, the landlord was responsible for collect- Kemeny 2004. ing revenues from villages after retaining part of the revenue he col- 46. Hoffman 1996. lected; in individual systems, British government officers collected 47. Hall 2002; Chandler 1992. revenue directly from cultivators; in village systems, a village com- 48. Chandler 1992. munity body bore the responsibility for revenue collection (Baner- 48. Imparato and Ruster 2003; Abreu 2008. jee and Iyer 2005). 50. World Bank 2007j. 6. Lall, Wang, and Deichmann 2008b. 51. Hall 1984. 7. Government of India 2006. 52. Trackman, Fisher, and Salas 1999. 8. Hunt 2000. 53. Mackay 1999; Jones and Datta 2000, based on Abreu 2008. 9. Gorzelak 2007. 54. Mackay 1999, based on Abreu 2008. 10. European Commission 2002. 55. Goodlad 1996; Mackay 1999; Jones and Datta 2000. 11. Between 1994 and 1999 structural funds allocated resources 56. These comparisons are based on data from Maddison 2006. to areas on the basis of five "objectives": supporting development 302 WORLD DEVELOPMENT REPORT 2009 and structural adjustment of areas whose development is lagging, 46. World Bank 2008e. helping frontier areas or parts of areas seriously affected by indus- 47. Econometric analysis based on panel data (Lei and Haynes trial decline, combating long-term unemployment and facilitating 2004). labor market integration of persons excluded from the labor mar- 48. Kloeppinger-Todd 2007. Overall, there are more than 20 ket, speeding adjustment of agricultural structures as part of com- million mobile phone subscribers in the Philippines. mon agriculture policy reform and structural adjustment of rural 49. Jensen 2007. areas, and promoting development and structural adjustment in 50. Luo 2004. areas with low population density. 51. Yepes and Lall 2008. 12. Rodríguez-Pose and Fratesi 2004. 52. Baldwin, Forslid, Martin, Ottaviano, and Robert-Nicoud 13. European Commission 2007. 2003. The predictions of new economic geography models are as 14. Objective One is to help regions (called "areas" in this follows: Infrastructure policies that facilitate interregional trade Report) whose development is lagging to catch up. Some 50 regions, between leading and lagging areas will increase spatial concentra- representing 22 percent of the EU's population, are included. tions of economic activity in leading areas. These policies will also 15. Dall'Erba 2003. increase growth in the whole economy while reducing nominal 16. Estimates suggest that resettlement costs were US$7,000 per income inequalities between areas and between workers and capital family in the mid-1980s (Adhiati and Bobsien 2001). owners. By contrast, infrastructure policies that improve connec- 17. Davezies (2001), Martin (2005). This figure shows that the tivity within lagging areas may enhance local economic growth but coefficient of variation of per capita GDP (orange line, left axis) can slow the growth of the whole economy. across NUTS 2 areas in France rises between 1982 and 2002; whereas 53. Faini 1983. the coefficient of variation of per capita disposable incomes falls 54. Combes and Lafourcade 2001. (blue line, right axis). 55. Jacoby 2000. 18. Morrisson and Murtin 2005. 56. World Bank 2008b. 19. United States Census Bureau 2002. 57. Lall, Schroeder, and Schmidt 2008. 20. About $8 an hour in 2007 U.S. dollars. 58. Government of Malaysia 2001. 21. Engel, Galetovic, and Raddatz 1998. 59. Malaysia Industrial Development Authority (http://www 22. Gibson and Rozelle 2003. .mida.gov.my/). 23. Escobal and Torero 2000. 60. Al-Hassan and Diao 2007. 24. Baer 1995. 61. Park forthcoming. 25. Stewart 2008. 62. Lee (2008). 26. Hill and Gaddy 2003. 63. The following examples are based on background research 27. Rolfes 2002. for this report. 28. Whalley and Zhang 2007. 64. Donoso-Clark and Leninhan (2008) call these centrally 29. Hewings, Feser, and Poole 2008; population figure for 1996. administered policies to disperse economic activity"first-generation" 30. Rephann and Isserman (1994). One of the first ex post eval- approaches to territorial development. uations of regional development programs ever conducted in the 65. Donoso-Clark and Leninhan 2008. United States using an experimental design. 66. Drabenstott 2005. 31. Specific examples are discussed later in this chapter. 67. Hewings, Feser, and Poole 2008. 32. Andrienko and Guriev 2003. 68. Carvalho, Lall, and Timmins 2006. Constitutional funds 33. Based on recent research using census micro data for Brazil were created in 1989 to finance economic activities in the north and between 1960 and 2000. northeast regions. 34 . Margo 1988. 69. World Bank 1977; Scott 1982. 35. Borsch-Supan 1987. 70. Chakravorty and Lall 2005. 36. Duarte, Ferreira, and Salvato Jr. 2004. The labor force in the 71. Fay, Felkner, and Lall 2008. northeast has 4.6 years of schooling compared with the average of 72. Based on research done by the Secretary of Economic Policy, 6.4 years nationwide, and 7.3 years in the southeast. reported in Calmon (2003) and World Bank (2005a). 37. Shah and Shen 2006. 73. Details are provided in Calmon (2003). 38. They will be accountable to the higher levels of government 74. Haussman, Rodrik, and Velasco 2005. providing these transfers. 75. Donoso-Clark and Leninhan 2008 point out that a feature 39. Numbers are for 2000 (Chakraborty 2003). of the failed "first-generation" territorial development programs 40. Euractiv (2008). of the 1960s and 1970s was that they depended on external inputs 41. World Bank (2003b) provides a comprehensive treatment of and expertise, rather than exploiting local assets and comparative land management practices and policies in different parts of the advantage. world, and how specific policies promote development and reduce 76. A recent paper on firm location in developing countries poverty. summarizes the findings from empirical work on the main factors 42. Palacin and Shelburne 2005. influencing location choices (see Deichmann, Lall, Redding, and 43. Bayes 2007. Venables forthcoming). 44. World Bank 2008b. 77. Uvalic 1993. 45. World Bank 2008a. 78. Bolton and Roland 1997. Endnotes 303 Chapter 9 29. Both examples are from a World Bank Independent Evalua- 1. Woronoff (1972), p. 141. The Entente was created in 1959 tion Group report on regional programs. and designed to promote the economic development of members 30. Leigland and Roberts 2007. by raising funds, guaranteeing loans, and encouraging trade and 31. World Bank 2006n. investment. It operates through the Mutual Aid and Loan Guar- 32. N'Guessan and Chitou 2006. antee Fund headquartered in Abidjan, Côte d'Ivoire. The original 33. African Development Bank 2007. member states were Dahomey (now Benin), Côte d'Ivoire, Niger, 34. Bougheas, Demetriades, and Morgenroth 1999; Limão and and Upper Volta (now Burkina Faso); Togo joined in 1966. Venables 2001. 2. Collier 2007. 35. For instance, Foroutan and Pritchett (1993) make this argu- 3. See Brülhart (2008) for a detailed analysis on intermediate ment for Sub-Saharan Africa. goods. 36. Coulibaly and Fontagné 2006. 4. Coulibaly 2008. 37. Buys, Deichmann, and Wheeler 2006. 5. See Yoshino (2008), using both the World Bank enterprise 38. Shepherd and Wilson 2006. survey database and the Doing Business database. 39. Vargas-Hidalgo 1979; Wionczek 1970. 6. Melitz 2003. 40. Hazelwood 1979; Kasekende and Ng'eno 1999. 7. See Schiff and Winters (2003) for a detailed development on 41. Brandts and Cooper (2007) show that effective communi- these points. cation is a more powerful coordination mechanism than purely 8. Viner 1950; Meade 1956; Balassa 1967; Aitken 1973; among financial incentives. others. 42. Sandler 2002. 9. Baier and Bergstrand 2004; Bond 2005; Evenett 2005; Berg- 43. Arce 2001; 2004; Arce and Sandler 2002. strand 2006. 44. Schiff and Winters 2002. 10. Bhagwati 1995; Bhagwati, Greenaway, and Panagariya 1998; 45. In Kraske and others (1996), p. 95. Krishna 1998. 46. See the International Development Association and World 11. Krugman 1991; Frankel, Stein, and Wei 1996; Carrere 2005. Bank (2008) for a thorough discussion of these issues. 12. Martin, Mayer, and Thoenig (2008) estimate that countries 47. Schiff and Winters 2002. trading a lot with their neighbors are less likely to have an armed 48. Mayer 2008. conflict with them than countries trading mainly with distant part- 49. Note that Northeast Asia is an exceptional case, because of ners. its proximity to Japan. Its main markets are still North America and 13. World Bank 2004a; United Nations 2007. the European Union. For purposes of market access, given its trade 14. Also in the 1970s, Sub-Saharan Africa was made up of vast links with China, Southeast Asia should be considered as part of the countries, mostly rural with a few isolated, densely populated areas same neighborhood as China. (mostly on the coast) and no interconnecting infrastructure-- 50. Hill and Gaddy 2003. hardly conducive to regional integration (CSAO/ECOWAS 2005). 51. International Monetary Fund (1998) for the EU enlarge- 15. The current WTO rules specify that regional trade agree- ment; Schiff and Wang (2003) for NAFTA. ments should cover substantially all sectors, which excludes sector- 52. World Bank 2002b. specific trade deals for today's developing countries. But the prin- 53. Jessen 2002. ciple of starting with focused areas of cooperation is still valid. 54. World Bank (2003a) report on trade and investment. 16. Venables (2003) shows that regional integration between 55. Assessment made using the methodology proposed by low-income countries tends to lead to income divergence between Coulibaly (2007). The trade impact of the Gulf Cooperation Coun- the least developed and the relatively more advanced member coun- cil (GCC) is not faring better from this assessment, even if its mem- tries. Goyal and Staal (2004) show that small countries are more in bers took the important step to launch a common market on Janu- favor of integration, whereas large countries prefer integrating with ary 1, 2008. countries of equal size. 56. World Bank 2007g. 17. Coulibaly 2006. 57. Antweiler and Trefler 2002. 18. World Bank 2004b. 58. See for instance Moreno and Trehan (1997). 19. Laeven and Woodruff 2007. 59. Arora and Vamvakidis (2005) show the impact of South 20. Box 3.3 of the report on food safety and agricultural health Africa on its neighbors, while Behar and Collier emphasize on standards (World Bank 2005c). the impact of resource-rich countries on their neighbors in Sub- 21. Gibbon and Ponte 2005; Chandra 2006; Broadman 2006; Saharan Africa. Czubala, Shepherd, and Wilson 2007. 60. Nabi and Nasim 2001. 22. Fink and Mattoo 2004; Hoekman 2006. 61. Naqvi and Schuler 2007. 23. Chow and others 2005; World Bank 2007g. 62. World Bank 2007h. 24. World Bank 2007f. 63. Söderbaum 2001. 25. Arvis, Raballand, and Marteau 2007. 64. Behar and Collier 2008. 26. Djankov, Freund, and Pham 2006. 65. United Nations (2004a) and United Nations and African 27. Behar and Manners 2008. Union (2006) report on regional integration. 28. World Bank 2004b. 66. N'Dulu 2001. 304 WORLD DEVELOPMENT REPORT 2009 67. Gregoire and Labazee 1993. GIM 4 68. Ladman 1979; Asiwaju 2005. 1. Alesina, Easterly, and Matuszeski 2006. 69. Dorosh, Haggblade, and Dradri 2007; Jayne and others 2005; 2. de Blij 2005. Tschirley and others 2004; Negri and Porto 2007. 3. Naudé 2007. 70. Little 2007. 4. Ndulu and others (2007), p. 101. 71. Gregoire and Labazee 1993; Yade and others 1999. 5. Satterthwaite 2007. 72. Adebusoye 2006. 6. Naudé and Matthee 2007. 73. Jourdan and NEPAD 2006. 7. Ramos 2007. 74. Hoekman and Njinkeu 2007. 8. Martínez-Zarzoso, García-Menéndez, and Suárez-Burguet 75. Independent Evaluation Group (IEG) 2007. 2003. 76. Collier and Venables 2007. References es The word "processed" describesinformally reproduced works that may not bebecommonly available through libraries. processed describes informally reproduced works that may not commonly available through libraries. Abreu, María. 2008."Effectively Dealing with Slums." Background Aghion, Philippe, and Jeffrey G. Williamson. 1998. Growth, note for the WDR 2009. Inequality and Globalization: Theory, History, and Policy. New Acemoglu, Daron, Simon Johnson, and James A. Robinson. 2001. York: Cambridge University Press. "The Colonial Origins of Comparative Development: An Aitken, Norman D. 1973."The Effects of the EEC and the EFTA on Empirical Investigation." American Economic Review 91 (5): ew European Trade: A Temporal Cross-Section Analysis." American 1369­401. Economic Review 63 (5): 881­92. ------. 2002."Reversal of Fortune: Geography and Institutions in Al-Hassan, Ramatu M., and Xinshen Diao. 2007."Regional Dis- the Making of the Modern World Income Distribution." Quar- parities in Ghana: Policy Options and Public Investment Impli- terly ly Journal of Economics 117 (4): 1231­94. cations." IFPRI Working Paper 693, Internationhal Food Policy Acosta, Pablo, Pablo Fajnzylber, and J. Humberto Lopez. 2007. Research Institute, Washington, DC. "The Impact of Remittances on Poverty and Human Capital: Alcalá, Francisco, and Antonio Ciccone. 2004."Trade and Produc- Evidence from Latin America Household Surveys." In Interna- tivity." Quarterlyly Journal of Economics 119 (2): 613­46. tional Migration, Economic Development, and Policy, ed. Caglar Alderman, Harold, Miriam Babita, Gabriel Demombynes, Ntha- Özden and Maurice Schiff. Washington, DC, and New York: biseng Makhatha, and Berk Özler. 2002."How Low Can You Go? World Bank and Palgrave Macmillan. Combining Census and Survey Data for Mapping Poverty in Adams, David, and E. M. Hastings. 2001."Urban Renewal in Hong South Africa." Journal of African Economies 11 (2): 169­200. Kong: Transition From Development Corporation to Renewal Alesina, Alberto, Arnaud Devleeschauwer, William Easterly, Sergio Authority." Land Use Policy 18(3):245­58. it Kurlat, and Romain Wacziarg. 2003."Fractionalization." Journal Adams, Richard H. Jr. 2006."Remittances, Poverty, and Investment of Economic Growth 8 (2): 155­94. in Guatemala." In International Migration, Remittances, and the Alesina, Alberto, William Easterly, and Janina Matuszeski. 2006. Brain Drain, ed. Caglar Özden and Maurice Schiff. Washington, "Artificial States." Working Paper 2115, Harvard Institute of DC, and New York: World Bank and Palgrave Macmillan. Economic Research, Cambridge, MA. Adebusoye, Paulina Makinwa. 2006."Geographic Labour Mobil- Alesina, Alberto, and Enrico Spolaore. 2003. The Size of Nations. ity in Sub-Saharan Africa." Globalization, Growth, and Poverty Cambridge, MA, and London: MIT Press. Working Paper Series 1, IDRC, Ottawa. Allen, Robert, Jean-Pascal Bassino, Debin Ma, Christine Moll- Adelman, Irma, and Sherman Robinson. 1978."Migration, Demo- Murata, and Jan Luiten Van Zanden. 2007."Wages, Prices, and graphic Change, and Income Distribution in a Model of a Living Standards in China, 1738­1925: In Comparison with Developing Country." In Research in Population Economics, vol. Europe, Japan, and India." Economics Working Paper 316, Uni- I, ed. Julian Simon. Greenwich, CT: JAI Press. versity of Oxford, Oxford, U.K. Adhiati, M. Adriana Sri, and Armin Bobsien. 2001. Indonesia's Alston, Lee, and T. J. Hatton. 1991."The Earnings Gap betweeneen Transmigration Programme--An Update: A Report Prepared for Agricultural and Manufacturing Laborers, 1925­1941." Journal Dow to Down to Earth th. London: International Campaign for Ecological of Economic History 51 (1): 83­99. Justice in Indonesia. Amiti, Mary, and Shang-Jin Wei. 2005."Fear of Service Outsourc- Africa Development Bank (ADB). 2007. Report of the High Level ing: Is It Justified?" Economic Policy 20 (42): 308­47. Panel: Investing in Africa's Future. Abidjan, Côte d'Ivoire: ADB. Anderson, James E., and Eric van Wincoop. 2003."Gravity with Aghion, Philippe, Robin Burgess, Stephen J. Redding, and Fabrizio Gravitas: A Solution to the Border Puzzle." American Economic Zilibotti. 2006."The Unequal Effects of Liberalization: Evidence Review 93 (1): 170­92. from Dismantling the License Raj in India." NBER Working ------. 2004."Trade Costs." Journal of Economic Literature 42 (3): atur Paper 12031, National Bureau of Economic Research, Cam- 691­751. bridge, MA. Anderson, Kathryn H., and Richard Pomfret. 2005."Spatial Aghion, Philippe, and Peter Howitt. 1992."A Model of Growth Inequality and Development in Central Asia." In Spatial Dis- Through Creative Destruction." Econometrica 60(2):323­51. parities in Human Development: Perspectives from Asia, ed. Ravi ------. 2005."Growth With Quality-Improving Innovations: An Kanbur, Anthony J. Venables, and Guanghua Wan. Tokyo and Integrated Framework." In Handbook of Economic Growth, Vol- or New York: United Nations University Press. ume 1A, ed. Philippe Aghion and Steven Durlauf. Amsterdam, Andrews, Martyn, Kenneth Clark, and William Whittaker. 2007. Netherlands: North-Holland. "The Employment and Earnings of Migrants in Great Britain." 305 306 WORLD DEVELOPMENT REPORT 2009 Working Paper 3068, Institute for the Study of Labor, Bonn, Atkinson, Michael M., and Richard A. Powers. 1987."Inside the Germany. Industrial Policy Garbage Can: Selective Subsidies to Business in Andrienko, Yuri, and Sergei M. Guriev. 2003."Determinants of Canada." Canadian Public Policy 13 (2): 208­17. Interregional Mobility in Russia: Evidence from Panel Data." Au, Chun-Chung, and J. Vernon Henderson. 2006a."Are Chinese Working Paper 3835, Centre for Economic Policy Research, Cities Too Small?" Review of Economic Studies 73 (3): 549­76. London. ------. 2006b."How Migration Restrictions Limit Agglomeration Angel, Shlomo. 2008."Preparing for Urban Expansion: A Proposed and Productivity in China." Journal of Development Economics Strategy for Intermediate Cities in Ecuador." In The New Global 80 (2): 350­88. Frontier: Cities, Poverty and Environment in the 21st Century, ed. Audretsch, David B., and Maryann P. Feldman. 1996."R&D Spill- G. Martine, G. McGranahan, M. Montgomery, and R. Castilla- overs and the Geography of Innovation and Production." Ame- Fernandez. London: IIED/UNFPA and Earthscan. rican Economic Review 86 (3): 630­40. Angeriz, Alvaro, John McCombie, and Mark Roberts. 2008."New Azzoni, Carlos, Naercio Menezes-Filho, Tatiane A. De Menezes, Estimates of Returns to Scale and Spatial Spillovers for EU and Raul Silveira-Neto. 2000."Geography and Income Con- Regional Manufacturing, 1986­2002." International Regional vergence among Brazilian States." Research Network Working Science Review 31 (1): 62­87. Paper R-395, Inter-American Development Bank, Washington, Anh, Dang Nguyen. 2003."Migration and Poverty in Asia, with DC. Reference to Bangladesh, China, the Philippines, and Viet Bae, Chang-Hee Christine, and Yaourai Suthiranart. 2003."Policy Nam." Paper presented at the Ad Hoc Expert Group Meeting on Options towards a Sustainable Urban Transportation Strategy Migration and Development, Bangkok, Thailand, August 27. for Bangkok." International Development Planning Review 25 Anh, Dang Nguyen, Cecilia Tacoli, and Hoang Xuan Thanh. 2003. (1): 31­51. "Migration in Vietnam: A Review of Information on Current Baer, Werner. 1995. The Brazilian Economy: Growth and Develop- Trends and Patterns, and Their Policy Implications." Paper pre- ment, 4th ed. Westport, CT: Praeger Publishers. sented at the Regional Conference on Migration, Development Baier, Scott L., and Jeffrey H. Bergstrand. 2004."Economic Deter- and Pro-poor Policy Choices in Asia, Dhaka, Bangladesh. minants of Free Trade Agreements." Journal of International Eco- Antràs, Pol, Luis Garicano, and Esteban Rossi-Hansberg. 2006. nomics 64 (1): 29­63. "Offshoring in a Knowledge Economy." Quarterly Journal of Bairoch, Paul. 1982."International Industrialization Levels from Economics 121 (1): 31­77. 1750 to 1980." Journal of European Economic History 11 (2): Antweiler, Werner, and Daniel Trefler. 2002."Increasing Returns 269­333. and All That: A View from Trade." American Economic Review 92 ------. 1988. Cities and Economic Development: From the Dawn of (1): 93­119. History to the Present. Chicago: University of Chicago Press. Arce, Daniel M. 2001."Leadership and the Aggregation of Interna- Bairoch, Paul, and Gary Goertz. 1986."Factors of Urbanization in tional Collective Action." Oxford Economic Papers 53 (1): 114­37. the Nineteenth-Century Developed Countries: A Descriptive ------. 2004."Asymmetric Leadership and International Public and Econometric Analysis." Urban Studies 23 (4): 285­305. Goods." Public Finance Review 32 (5): 528­58. Baker, Jim. 1999. Crossroads: A Popular History of Malaysia and Armstrong, Harvey, and Jim Taylor. 2000. Regional Economics and Singapore. Singapore: Times Books International. Policy. Oxford, U.K.: Blackwell. Baker, Judy, Rakhi Basu, Maureen Cropper, Somik V. Lall, and Arnold, John. 2006."Best Practices in Management of Interna- Akie Takeuchi. 2005."Urban Poverty and Transport: The Case tional Trade Corridors." Transport Working Paper TM-13, of Mumbai." Policy Research Working Paper 3693, World Bank, World Bank, Washington, DC. Washington, DC. Arora, Vivek, and Athanasios Vamvakidis. 2005."The Implica- Baker, Kevin. 2001."The First Slum in America." New York Times, tions of South African Economic Growth for the Rest of Africa." September 30. South African Journal of Economics 73 (2): 229­42. Balassa, Bela. 1967."Trade Creation and Trade Diversion in Euro- Arrow, Kenneth J. 1962."Economic Welfare and the Allocation of pean Common Market." Economic Journal 77 (305): 1­21. Resources for Invention." In The Rate and Direction of Inven- Baldwin, John R., and Paul K. Gorecki. 1986. Role of Scale in tive Activity: Economic and Social Factors, ed. Conference of the Canada­U.S. Productivity Differences in the Manufacturing Sec- Universities and the National Bureau Committee for Economic tor 1970­1979. Buffalo, NY, and Toronto, Canada: University of Research. Princeton, NJ: Princeton University Press. Toronto Press. Arvis, Jean-François, Gael Raballand, and Jean-François Marteau. Baldwin, Richard E. 1995."A Domino Theory of Regionalism." In 2007."The Cost of Being Landlocked: Logistics Cost and Supply Expanding Membership of the European Union, ed. Richard E. Chain Reliability." Policy Research Working Paper 4258, World Baldwin, Pertti Haaparanta, and Jaakko Kiander. Cambridge, Bank, Washington, DC. MA, and New York: Cambridge University Press. Asiwaju, Anthony I. 2005."Transfrontier Regionalism: The Euro- ------. 2008."Sequencing and Depth of Regional Economic Inte- pean Union Perspective on Postcolonial Africa, with Special Ref- gration: Lessons for the Americas from Europe." World Economy erence to Borgu." In Holding the Line. Borders in a Global World, 31 (1): 5­30. ed. Heather N. Nicol and Ian Townsend-Gault. Vancouver, Can- Baldwin, Richard E., Rikard Forslid, Philippe Martin, Gianmarco ada: University of British Columbia Press. I. P. Ottaviano, and Frederic Robert-Nicoud. 2003. Economic References 307 Geography and Public Policy. Princeton, NJ: Princeton University Behar, Alberto, and Phil Manners. 2008."Logistics and Exports." Press. Department of Economics and CSAE Working Paper 2008-13, Baldwin, Richard E., and Philippe Martin. 1999."Two Waves of University of Oxford, Oxford, U.K. Globalization: Superficial Similarities, Fundamental Differ- Bekaert, Geert, Campbell R. Harvey, and Christian Lundblad. ences." In Globalization and Labor, ed. Horst Siebert. Tuebingen, 2005."Does Financial Liberalization Spur Growth?" Journal of Germany: Mohr Siebeck. Financial Economics 77 (1): 3­55. Baldwin, Richard E., Philippe Martin, and Gianmarco I. P. Ottavi- Bell, Martin. 2003."Comparing Internal Migration between ano. 2001."Global Income Divergence, Trade, and Industrializa- Countries: Measures, Data Sources, and Results." Working tion: The Geography of Growth Take-Offs." Journal of Economic Paper 2003/02, University of Queensland Centre for Population Growth 6 (1): 5­37. Research, Brisbane, Australia. Baldwin, Richard E., and Charles Wyplosz. 2006. The Economics of Bento, Antonio M., Maureen L. Cropper, Ahmed Mushfiq Moba- European Integration, 2nd ed. London: McGraw-Hill Education. rak, and Katja Vinha. 2003."The Impact of Urban Spatial Struc- Balisacan, Arsenio, Hal Hill, and Sharon Faye Piza. Forthcoming. ture on Travel Demand in the United States." Policy Research "Spatial Disparities and Development Policy in the Philippines." Working Paper 3007, World Bank, Washington, DC. In Reshaping Economic Geography in East Asia, ed. Yukon Huang Bergstrand, Jeffrey H. 2006."Regional Integration as a Develop- and Alessandro Magnoli Bocchi. Washington, DC: World Bank, ment Strategy." Paper presented at The New Regionalism: Prog- EAP Companion Volume to the WDR 2009. ress, Setbacks, and Challenges, Washington, DC. Balk, Deborah L., Uwe Deichmann, Gregory Yetman, Francesca Bernstein, Jeffrey I., and M. Ishaq Nadiri. 1988."Inter-industry Pozzi, Simon I. Hay, and Andrew Nelson. 2006."Determining R&D Spillovers, Rates of Return, and Production in High-Tech Global Population Distribution: Methods, Applications, and Industries." American Economic Review 78 (2): 429­34. Data." Advances in Parasitology 62: 119­56. Bertaud, Alain. 2003."Order without Design." World Bank, Wash- Banerjee, Abhijit, and Lakshmi Iyer. 2005."History, Institutions, ington, DC. Processed. and Economic Performance: The Legacy of Colonial Land ------. 2004."The Spatial Organization of Cities: Deliber- Tenure Systems in India." American Economic Review 95 (4): ate Outcome or Unforeseen Consequence?" Working Paper 1190­213. WP-2004-01, Institute of Urban and Regional Development, Barro, Robert J., and Xavier Sala-I-Martin. 1992."Convergence." Berkeley, CA. Journal of Political Economy 100 (2): 223­51. Bertram, Geoffrey. 2004."On the Convergence of Small Island ------. 2004. Economic Growth, 2nd ed. Cambridge, MA: MIT Economies with their Metropolitan Patrons." World Develop- Press. ment 32 (2): 343­64. Barro, Robert J., Xavier Sala-I-Martin, Olivier J. Blanchard, and Bhagwati, Jagdish. 1995."U.S. Trade Policy: The Infatuation with Robert E. Hall. 1991."Convergence across States and Cities." Free Trade Areas." In The Dangerous Drift to Preferential Trade Brookings Papers on Economic Activity 1991 (1): 107­82. Agreements, ed. Jagdish Bhagwati and Anne O. Krueger. Wash- Bartelsman, Eric J., Ricardo J. Caballero, and Richard K. Lyons. ington, DC: American Enterprise Institute. 1994."Customer and Supplier-Driven Externalities." American Bhagwati, Jagdish, David Greenaway, and Arvind Panagariya. 1998. Economic Review 84 (4): 1075­84. "Trading Preferentially: Theory and Policy." Economic Journal Basu, Susanto, and John G. Fernald. 1997."Returns to Scale in U.S. 108 (449): 1128­48. Production: Estimates and Implications." Journal of Political Bhagwati, Jagdish, and T. N. Srinivasan. 2002."Trade and Pov- Economy 105 (2): 249­83. erty in the Poor Countries." American Economic Review 92 (2): Baumert, Kevin A., Timothy Herzog, and Jonathan Pershing. 2005. 180­3. Navigating the Numbers: Greenhouse Gas Data and International Biedermann, Rob. 2007."Sinvin Realty Handles Meatpacking Dis- Climate Policy. Washington, DC: World Resources Institute. trict Deal in NYC." crefeed.com, October 29. Baumol, William J. 1986."Productivity Growth, Convergence, and Bikhchandani, Sushil, David Hirshleifer, and Ivo Welch. 1998. Welfare: What the Long-Run Data Show." American Economic "Learning from the Behavior of Others: Conformity, Fads, and Review 76 (5): 1072­85. Informational Cascades." Journal of Economic Perspectives 12 (3): Bayes, Abdul. 2007. Impact Assessment of Jamuna Multipurpose 151­70. Bridge Project (JMBP) on Poverty. Dhaka: Japan Bank for Inter- Bird, Kate, and Kate Higgins. 2008."Regional Inequality and Pri- national Cooperation. mary Education in Northern Uganda, Policy Brief 2." Back- Bazoglu, Nefise. 2008. Cities in Transition: Demographics and the ground paper for the WDR 2009. Development of Cities. Philadelphia: Pennsylvania State University. Black, Duncan, and J. Vernon Henderson. 1999."A Theory of Beaumier, Guy. 1998. Regional Development in Canada. Ottawa: Urban Growth." Journal of Political Economy 107 (2): 252­84. Government of Canada. ------. 2003."Urban Evolution in the USA." Journal of Economic Beegle, Kathleen, Joachim De Weerdt, and Stefan Dercon. 2008. Geography 3 (4): 343­72. "Migration and Economic Mobility in Tanzania: Evidence from Blake, Gerald. 2005."Boundary Permeability in Perspective." In a Tracking Survey." World Bank, Washington, DC. Processed. Holding the Line: Borders in a Global World, ed. Heather N. Nicol Behar, Alberto, and Paul Collier. 2008."Does Proximity Matters and Ian Townsend-Gault. Vancouver, Canada: University of More Than Wealth?" CSAE, Oxford, U.K. British Columbia Press. 308 WORLD DEVELOPMENT REPORT 2009 Blanchard, Olivier J., Lawrence F. Katz, Robert E. Hall, and Barry Brandts, Jordi, and David J. Cooper. 2007."It's What You Say, Not Eichengreen. 1992."Regional Evolutions." Brookings Papers on What You Pay: An Experimental Study of Manager-Employee Economic Activity 1992 (1): 1­75. Relationships in Overcoming Coordination Failure." Journal of Boltho, Andrea, Wendy Carlin, and Pasquale Scaramozzino. 1997. the European Economic Association 5 (6): 1223­68. "Will East Germany Become a New Mezzogiorno?" Journal of Brezis, Elise S., and Paul R. Krugman. 1997."Technology and the Comparative Economics 24 (3): 241­64. Life Cycle of Cities." Journal of Economic Growth 2 (4): 369­83. Bolton, Patrick, and Gerard Roland. 1997."The Breakup of Bristow, R. 1984. Land-Use Planning in Hong Kong: History, Poli- Nations: A Political Economy Analysis." Quarterly Journal of cies, and Procedures. Hong Kong: Oxford University Press. Economics 112 (4): 1057­90. Broadman, Harry G. 2005. From Disintegration to Reintegration: Bond, Eric W. 2005."The Sequencing of Trade Liberalization Eastern Europe and the Former Soviet Union in International in the Presence of Adjustment Costs." Paper presented at the Trade. Washington, DC: World Bank. Sequencing of Regional Economic Integration: Issues in the ------. 2006. Africa's Silk Road: China and India's New Economic Breadth and Depth of Economic Integration in the Americas, Frontier. Washington, DC: World Bank. Notre Dame, IN. Brockerhoff, M., and P. Hewett. 2000."Inequality of Child Mortal- Booth, Charles. 1901. Improved Means of Locomotion as a First Step ity among Ethnic Groups in Sub-Saharan Africa." Bulletin of the Towards the Cure of the Housing Difficulties of London. London: World Health Organization 78 (1): 30­41. Macmillan. Brown, Annette N. 1997."The Economic Determinants of the Bordo, Michael D., Barry Eichengreen, and Douglas A. Irwin. 1999. Internal Migration Flows in Russia During Transition." Working "Is Globalization Today Really Different Than Globalization Paper 89, William Davidson Institute, Ann Arbor, MI. a Hundred Years Ago?" NBER Working Paper 7195, National Brown, David, Marianne Fay, John Felkner, Somik V. Lall, and Bureau of Economic Research, Cambridge, MA. Hyoung Gun Wang. 2008."The Death of Distance? Economic Boreak, Sik. 2000."Land Ownership, Sales, and Concentration Implications of Infrastructure Improvement in Russia." World in Cambodia. A Preliminary Review of Secondary Data and Bank, Office of the Chief Economist, Europe and Central Asia Primary Data from Four Recent Surveys." Working Paper 16, Region, Washington, DC. Development Resource Institute, Phnom Penh. Brueckner, Jan K. 2007."Government Land-Use Interventions: Borgegård, Lars-Erik, and Jim Kemeny. 2004."Sweden: High-Rise An Economic Analysis." Paper presented at the World Bank 4th Housing for a Low-Density Country." In High-Rise Housing in Urban Research Symposium, Washington, DC. Europe: Current Trends and Future Prospects, ed. R. Turkington, Brülhart, Marius. 2008."An Account of Global Intra-Industry R. van Kempen, and E. Wassenberg. Delft, Netherlands: Delft Trade, 1962­2006." Working Paper 2008-08, University of Not- University Press. tingham, Nottingham, U.K. Borjas, George J. 1990. Friends or Strangers: The Impact of Immi- Brunetti, Aymo, and Beatrice Weder. 2003."A Free Press Is Bad gration on the U.S. Economy. New York: Basic Books. News for Corruption." Journal of Public Economics 87 (7­8): Borjas, George J., Stephen G. Bronars, and Stephen J. Trejo. 1992. 1801­24. "Self Selection and Internal Migration in the United States." Brushett, Stephen. 2005."Management and Financing of Road Journal of Urban Economics 32 (2): 159­85. Transport Infrastructure in Africa." Sub-Saharan Africa Trans- Borsch-Supan, Axel. 1987."The Role of Education: Mobility port Policy Working Paper 4, World Bank, Washington, DC. Increasing or Mobility Impeding?" NBER Working Paper 2329, Bryceson, Deborah Fahy. 1999."Sub-Saharan Africa Betwixt and National Bureau of Economic Research, Cambridge, MA. Between: Rural Livelihood Practices and Policies." Working Bossone, Biagio, and Jong-Kun Lee. 2004."In Finance, Size Mat- Paper 43, African Studies Centre, Leiden, Netherlands. ters: The `Systemic Scale Economies' Hypothesis." IMF Staff Buckley, Robert, Alain Bertaud, and V. K. Phatak. 2005."Property Papers 51 (1): 19­46. Rights and Interlocking Policy Constraints Urban Land Mar- Bougheas, Spiros, Panicos O. Demetriades, and Edgar L. W. Mor- kets: Reforming Mumbai's Real Estate Raj." Paper presented at genroth. 1999."Infrastructure, Transport Costs, and Trade." the Land Policies and Administration for Accelerated Growth Journal of International Economics 47 (1): 169­89. and Poverty Reduction, New Delhi. Bourguignon, François, and Christian Morrisson. 1998."Inequal- Buckley, Robert, and Jerry Kalarickal, eds. 2006. Thirty Years of ity and Development: The Role of Dualism." Journal of Develop- World Bank Shelter Lending: What Have We Learned? Washing- ment Economics 57 (2): 233­57. ton, DC: World Bank. ------. 2002."Inequality among World Citizens: 1820­1992." Burchfield, Marcy, Henry G. Overman, Diego Puga, and Matthew American Economic Review 92 (4): 727­44. Turner. 2006."Causes of Sprawl: A Portrait from Space." Quar- Brahmbhatt, Milan. 2004. Strong Fundamentals to the Fore: terly Journal of Economics 121 (2): 587­633. Regional Overview. Washington, DC: World Bank, East and Burkey, Mark L. 2006."Gini Coefficients for the 2000 Census." Pacific Region. North Carolina A&T State University, Department of Econom- Brakman, Steven, Harry Garretsen, and Charles van Marrewijk. ics and Transportation-Logistics, Greensboro, NC. 2001. An Introduction to Geographical Economics: Trade, Loca- Button, Kenneth. 1999."Shipping Alliances: Are They at the Core tion, and Growth. New York: Cambridge University Press. of Solving Instability Problems in Shipping?" Paper presented at References 309 the 1999 Meeting of the International Association of Maritime Catin, Maurice, and Christophe Van Huffel. 2003."Inégalités Economists, Halifax, Canada. Régionales et Développement Economique: le Cas Français Buys, Piet, Susmita Dasgupta, Timothy S. Thomas, and David (1850­2000)." La Revue d'Economie Régionale e Urbaine Part 5: Wheeler. 2008."Determinants of a Digital Divide in Sub- 1­18. Saharan Africa: A Spatial Econometric Analysis of Cell Phone Chakraborty, Pinaki. 2003."Unequal Fiscal Capacities across Coverage." Policy Research Working Paper 4516, World Bank, Indian States: How Corrective Is the Fiscal Transfer Mecha- Washington, DC. nism?" Paper presented at the WIDER project conference on Buys, Piet, Uwe Deichmann, and David Wheeler. 2006."Road Net- Spatial Inequality in Asia at United Nations University Head- work Upgrading and Overland Trade Expansion in Sub-Saharan quarters, Tokyo. Africa." Policy Research Working Paper 4097, World Bank, Chakravorty, Sanjoy, and Somik V. Lall. 2005."Industrial Loca- Washington, DC. tion and Spatial Inequality: Theory and Evidence from India." Cabinet Council. 1972. New Integrated Spatial Development Plan Review of Development Economics 9 (1): 47­68. (Japanese). Tokyo: Cabinet Council. ------. 2007. Made in India: The Economic Geography and Politi- Calem, Paul S. 1994."The Impact of Geographic Deregulation on cal Economy of Industrialization. New Delhi and New York: Small Banks." Business Review (November/December): 17­31. Oxford University Press. Calem, Paul S., and Gerald A. Carlino. 1991."Urban Agglomera- Chandler, Alfred Dupont Jr. 1977. The Visible Hand: The Manage- tion Economies in the Presence of Technical Change." Journal of rial Revolution in American Business. Cambridge, MA: Belknap Urban Economics 29 (1): 82­95. Press. Cali, Massimiliano. 2008."Urbanization, Inequality, and Economic Chandler, Mittie Olion. 1992."Public Housing Desegregation: Growth: Evidence from Indian States." Background paper for What Are the Options?" Housing Policy Debate 3 (2): 509­34. the WDR 2009. Chandler, Tertius. 1987. Four Thousand Years of Urban Growth: An Calmon, Paulo Carlos Du Pin. 2003."Evaluation of Subsidies in Historical Census. Lampeter, U.K.: Edwin Mellen Press. Brazil: An Overview." World Bank, Washington, DC. Chandler, Tertius, and Gerald Fox. 1974. Three Thousand Years of Canning, David, and Esra Bennathan. 2007."The Rate of Return Urban Growth. New York and London: Academic Press. to Transportation Infrastructure." In Transport, Infrastructure Chandra, Vandana, ed. 2006. Technology Adaptation and Exports: Investment and Economic Productivity, ed. OECD/ECMT. Paris: How Some Developing Countries Got It Right. Washington, DC: Organisation for Economic Co-operation and Development. World Bank. Cárdenas, Mauricio, and Adriana Pontón. 1995."Growth and Con- Chang, Ha-Joon. 2005."Policy Space in Historical Perspective-- vergence in Colombia: 1950­1990." Journal of Development Eco- with Special Reference to Trade and Industrial Policies." Paper nomics 47 (1): 5­37. presented at the Queen Elizabeth House 50th Anniversary Con- Carlino, Gerald A. 1979."Increasing Returns to Scale in Metropoli- ference: The Development of Threats and Promises, Oxford, tan Manufacturing." Journal of Regional Science 19 (3): 369­72. U.K. Carlino, Gerald A., and Satyajit Chatterjee. 2001."Employment Charlot, Sylvie, and Gilles Duranton. 2004."Communication Deconcentration: A New Perspective on America's Postwar Externalities in Cities." Journal of Urban Economics 56 (3): Urban Evolution." Working Paper 01-4, Federal Reserve Bank of 581­613. Philadelphia, Philadelphia. Chase-Dunn, Christopher, Yukio Kawano, and Benjamin D. Carlton, Dennis W. 1983."The Location and Employment Choices Brewer. 2000."Trade Globalization since 1795: Waves of Inte- of New Firms: An Econometric Model with Discrete and Con- gration in the World System." American Sociological Review 65 tinuous Endogenous Variables." Review of Economics and Statis- (1): 77­95. tics 65 (3): 440­49. Chen, Yang. 2008."China: A Case Study of 1D-2D-3D Areas." Carrère, Céline. 2007."Regional Agreement and the Welfare in the Background note for the WDR 2009. South: When Scale Economies in Transport Matters." Working ------. 2008."Urban Agglomeration Policy in China." Back- Paper 2007-26, CERDI, Clermont-Ferrand, France. ground note for the WDR 2009. Carvalho, Alexandre S., Somik V. Lall, and Christopher Timmins. Chinitz, Benjamin J. 1961."Contrasts in Agglomeration: New York 2006."Regional Subsidies and Industrial Prospects of Lagging and Pittsburgh." American Economic Review 51 (2): 279­89. Regions." Policy Research Working Paper 3843, World Bank, Chinn, Menzie D., and Hiro Ito. 2006."What Matters for Financial Washington, DC. Development? Capital Controls, Institutions, and Interactions." Caselli, Francesco. 2005."Accounting for Cross-Country Income Journal of Development Economics 81 (1): 163­92. Differences." In Handbook of Economic Growth, vol. I, ed. Philippe Chiquier, Loic, Olivier Hassler, and Michael Lea. 2004."Mortgage Aghion and Steven N. Durlauf. Amsterdam: North-Holland. Securities in Emerging Markets." Policy Research Working Paper Caselli, Francesco, and James Feyrer. 2007."The Marginal Product 3370, World Bank, Washington, DC. of Capital." Quarterly Journal of Economics 122 (2): 535­68. Chomitz, Kenneth M., Piet Buys, and Timothy S. Thomas. 2005. Cashin, Paul A. 1995."Economic Growth and Convergence across "Quantifying the Rural-Urban Gradient in Latin America and the Seven Colonies of Australasia: 1861­1991." Economic Record the Caribbean." Policy Research Working Paper 3634, World 71 (2): 132­44. Bank, Washington, DC. 310 WORLD DEVELOPMENT REPORT 2009 Chow, Hwee K., Peter N. Kriz, Roberto S. Mariano, and Augustine Working Paper 2894, Centre for Economic Policy Research, H. H. Tan. 2005. Trade, Investment, and Financial Integration London. in East Asia. Tanglin, Singapore: Association of Southeasrtern ------. 2005."Transport Costs: Measures, Determinants, and Asian Nations + 3 Secretariat. Regional Policy Implications for France." Journal of Economic Christaller, Walter. 1933. Central Places in Southern Germany. Geography 5 (3): 319­49. Englewood Cliffs, NJ: Prentice Hall. Commission for Africa (CFA). 2005. Our Common Interest: Report Christensen, Dan Charly. 2004."Physiocracy: The Missing Link." of the Commission for Africa. London: CFA. In Modernisation and Tradition European Local and Manorial Connell, J., B. Dasgupta, R. Laishley, and M. Lipton. 1976. Migra- Societies 1500­1900, ed. K. Sundberg, T. Germundsson, and K. tion from Rural Areas: The Evidence from Village Studies. Delhi: Hansen. Lancaster, U.K.: Nordic Academic Press. Oxford University Press. Ciccone, Antonio. 2002."Agglomeration Effects in Europe." Euro- Costa, Dora L., and Matthew E. Kahn. 2000."Power Couples: pean Economic Review 46(2):213­27. Changes in the Locational Choice of the College Educated, Ciccone, Antonio, and Robert E. Hall. 1996."Productivity and 1940­1990." Quarterly Journal of Economics 115 (4): 1287­315. the Density of Economic Activity." American Economic Review Coulibaly, Souleymane. 2006."Regional Integration and the Per- 86(1):54­70. sistent Uneven Spread of Economic Activities in Developing Ciccone, Antonio, and Giovanni Peri. 2006."Identifying Human- Areas." Economie Internationale 106 (2): 5­23. Capital Externalities: Theory with Applications." Review of Eco- ------. 2007."Evaluating the Trade Impact of Developing RTAs: nomic Studies 73 (2): 381­412. A Semi-Parametric Approach." Policy Research Working Paper Clark, Colin. 1957. The Conditions of Economic Progress, 3rd ed. 4220, World Bank, Washington, DC. London: Macmillan. ------. 2008."On the Complementarity of Regional and Global Clemens, Michael A., Claudio E. Montenegro, and Lant Pritch- Trade." Background paper for the WDR 2009. ett. 2008."The Place Premium: Wage Differences for Identical Coulibaly, Souleymane, and Lionel Fontagné. 2006."South-South Workers across the U.S. Border." Policy Research Working Paper Trade: Geography Matters." Journal of African Economies 15 (2): 4671, World Bank, Washington, DC. Background paper for the 313­41. WDR 2009. Crafts, Nicholas. 2005."Regional GDP in Britain, 1871­1911: Cockrem, Tom. 2007."Singapore's Slum Deal." The Sun-Herald, Some Estimates." Scottish Journal of Political Economy 52 (1): April 8. 54­64. Coe, Patrick J., and J. C. Herbert Emery. 2004."The Disintegrating ------. 2008."Spatial Disparities in 19th-Century British Indus- Canadian Labour Market? The Extent of the Market Then and trialization." Background note for the WDR 2009. Now." Canadian Journal of Economics 37 (4): 879­97. Crafts, Nicholas, Terence C. Mills, and Abay Mulatu. 2007."Total Cohen, Daniel, and Marcelo Soto. 2007."Growth and Human Factor Productivity Growth on Britain's Railways, 1852­1912: A Capital: Good Data, Good Results." Journal of Economic Growth Reappraisal of the Evidence." Exploration in Economic History 44 12 (1): 51­76. (4): 608­34. Cole, William E., and Richard D. Sanders. 1985."Internal Migra- Crafts, Nicholas, and Gianni Toniolo. 1996. Economic Growth in tion and Urban Employment in the Third World." American Europe since 1945. Cambridge, U.K., and New York: Cambridge Economic Review 75 (3): 481­94. University Press. Collier, Paul. 2003. Breaking the Conflict Trap: Civil War and Devel- Crafts, Nicholas F. R. 1989."British Industrialization in an Inter- opment Policy. Washington, DC, and New York: World Bank and national Context." Journal of Interdisciplinary History 19 (3): Oxford University Press. 415­28. ------. 2007. The Bottom Billion: Why the Poorest Countries Are Criscuolo, Chiara, Jonathan E. Haskel, and Mattew J. Slaughter. Failing and What Can Be Done about It. New York and Oxford, 2004."Why Are Some Firms More Innovative? Knowledge U.K.: Oxford University Press. Inputs, Knowledge Stocks, and the Role of Global Engagement." Collier, Paul, and Stephen A. O'Connel. 2008."Opportunities and Tuck School of Business, Hanover, NH. Choices." In The Political Economy of Economic Growth in Africa, CSAO-ECOWAS. 2005. Atlas des Transports et des Telecommunica- 1960­2000, vol. 1, ed. Benno J. Ndulu, Stephen A. O'Connel, tions dans la West African States. Paris and Abuja: Club Du Sahel Robert H. Bates, Paul Collier, and Charles C. Soludo. Cam- et de L'Afrique de L'Ouest and Economic Community of West bridge, U.K.: Cambridge University Press. African States. Collier, Paul, and Anthony J. Venables. 2007."Rethinking Trade Cullinane, Sharon. 2002."The Relationship between Car Owner- Preferences: How Africa Can Diversify Its Exports." World Econ- ship and Public Transport Provision: A Case Study of Hong omy 30 (8): 1326­45. Kong." Transport Policy 9 (1): 29­39. Combes, Pierre-Philippe. 2000."Marshall-Arrow-Romer Exter- Cutler, David, Angus Deaton, and Adriana Lleras-Muney. 2006. nalities and City Growth." Working Paper 99-06, Centre d'Étude "The Determinants of Mortality." Journal of Economic Perspec- et de Recherche en Action Sociale, Paris. tives 20 (3): 97­120. Combes, Pierre-Philippe, and Miren Lafourcade. 2001."Transport Czubala, Witold, Ben Shepherd, and John S. Wilson. 2007."Help Cost Decline and Regional Inequalities: Evidence from France." or Hindrance? The Impact of Harmonized Standards on Afri- References 311 can Exports." Policy Research Working Paper 4400, World Bank, Deininger, Klaus W., and S. Jin. 2003."Land Sales and Rental Washington, DC. Markets in Transition. Evidence from Rural Vietnam." Policy Da Mata, Daniel, Uwe Deichmann, J. Vernon Henderson, Somik V. Research Working Paper 3013, World Bank, Washington, DC. Lall, and Howard G Wang. 2007."Determinants of City Growth Dekle, Robert, and Jonathan Eaton. 1999."Agglomeration and in Brazil." Journal of Urban Economics 62 (2): 252­72. Land Rents: Evidence from the Prefectures." Journal of Urban Dall'Erba, Sandy. 2003."The Trade-off Efficiency-Equity as an Economics 46 (2): 200­14. Explanation of the Mitigated Success of the European Regional Demurger, Sylvie, Jeffrey D. Sachs, Wing Thye Woo, Shuming Bao, Development Policies." Working Paper 2 (3), EUC, Urbana, IL. Gene Chang, and Andrew Mellinger. 2002."Geography, Eco- Davezies, Laurent. 2001. Policies for Sound Development. Paris: nomic Policy and Regional Development in China." Asian Eco- Organisation for Economic Co-operation and Development. nomic Papers 1 (1): 146­97. Davis, Donald R., and David E. Weinstein. 1999."Economic Geog- Deshingkar, Priya, and Sven Grimm. 2004. Voluntary Internal raphy and Regional Production Structure: An Empirical Investi- Migration: An Update. London: Overseas Development Institute. gation." European Economic Review 43 (2): 379­407. Desmet, Klaus, and Marcel Fafchamps. 2003."What Are Falling Davis, Steven J., and John C. Haltiwanger. 1992."Gross Job Cre- Transport Costs Doing to Spatial Concentration across U.S. ation, Gross Job Destruction, and Employment Reallocation." Counties?" Working Paper 3853, Centre for Economic Policy Quarterly Journal of Economics 107 (3): 819­63. Research, London. de Blij, Harm. 2005. Why Geography Matters: Three Challenges Fac- ------. 2006."Employment Concentration across U.S. Counties." ing America: Climate Change, the Rise of China, and Global Ter- Regional Science and Urban Economics 36 (4): 482­509. rorism. New York: Oxford University Press. Deutsch, Robert. 2006."Beneficiary Assessment of Land Titles De Brauw, Alan, and John Giles. 2008."Migrant Labor Markets Recipients under Land Management and Administrative Project and the Welfare of Rural Households in the Developing World: (LMAP)." Evidence from China." Policy Research Working Paper 4585, Devereux, Michael P., Rachel Griffith, and Helen Simpson. 2004. World Bank, Washington, DC. "The Geographic Distribution of Production Activity in the de la Fuente, Angel. 2000."Convergence across Countries and U.K." Regional Science and Urban Economics 34 (5): 533­64. Regions: Theory and Empirics." Working Paper 2465, Centre for Dillinger, William. 2007."Poverty and Regional Development in Economic Policy Research, London. Eastern Europe and Central Asia." Working Paper 118, World ------. 2004."Second-Best Redistribution through Public Invest- Bank, Washington, DC. ment: A Characterization, an Empirical Test and an Application Disdier, Anne-Celia, and Keith Head. 2008."The Puzzling Persis- to the Case of Spain." Regional Science and Urban Economics 34 tence of the Distance Effect on Bilateral Trade." Review of Eco- (5): 489­503. nomics and Statistics 90 (1): 37­48. de la Fuente, Angel, Xavier Vives, Juan J. Dolado, and Riccardo Faini. Dixie, G. 2002. Review of Current and Future Airfreight Situa- 1995."Infrastructure and Education as Instruments of Regional tion from Bangladesh. Dhaka: Accord Associates for the Hortex Policy: Evidence from Spain." Economic Policy 10 (20): 11­51. Foundation. De Long, J. Bradford. 1988."Productivity Growth, Convergence, Dixit, Avinash K., and Joseph E. Stiglitz. 1977."Monopolistic and Welfare: Comment." American Economic Review 78 (5): Competition and Optimum Product Diversity." American Eco- 1138­54. nomic Review 67 (3): 297­308. De Long, J. Bradford, and Andrei Shleifer. 1993."Princes and Mer- Djankov, Simeon, Caroline L. Freund, and Cong S. Pham. 2006. chants: European City Growth before the Industrial Revolu- "Trading on Time." Policy Research Working Paper 3909, World tion." Journal of Law and Economics 36 (2): 671­702. Bank, Washington, DC. de Soto, Hernando. 2000. The Mystery of Capital: Why Capitalism Dobkins, Linda Harris, and Yannis M. Ioannides. 2001."Spatial Triumphs in the West and Fails Everywhere Else. London: Bantan Interactions among U.S. Cities, 1900­1990." Regional Science Press. and Urban Economics 31 (6): 701­31. Decressin, Jörg, and Antonio Fatás. 1994."Regional Labour Mar- Docquier, Frédéric. 2006."Brain Drain and Inequality across ket Dynamics in Europe." Working Paper DP1085, Centre for Nations." Paper presented at the EUDN-AFD Conference on Economic Policy Research, London. Migration and Development, Paris. Deichmann, Uwe, Marianne Fay, Jun Koo, and Somik V. Lall. 2004. Docquier, Frédéric, Michel Beine, and Hillel Rapoport. 2006. "Economic Structure, Productivity, and Infrastructure Quality "Brain Drain and Human Capital Formation in Developing in Southern Mexico." Annals of Regional Science 38 (3): 361­85. Countries: Winner and Losers." Working Paper 2006-23, Uni- Deichmann, Uwe, Kai Kaiser, Somik V. Lall, and Zmarak Shalizi. versité Catholique de Louvain, Département des Sciences Eco- 2005."Agglomeration, Transport, and Regional Development in nomiques, Louvain-la-Neuve, Belgium. Indonesia." Policy Research Working Paper 3477, World Bank, Dongguan Government. 2005. Guangdong Statistical Yearbook Washington, DC. 2005. Beijing, China: China Statistics Press. Deichmann, Uwe, Somik V. Lall, Stephen J. Redding, and Anthony Donoso Clark, M., and M. Leninhan. 2008."A Space-Based Devel- J. Venables. 2008."Industrial Location in Developing Countries." opment Framework: A Bottom-Up Approach to Economic Inte- World Bank Research Observer 23(2): 219­46. gration." World Bank, Environmentally and Socially Sustainable 312 WORLD DEVELOPMENT REPORT 2009 Development Unit, Europe and Central Asia Regional Office, Easterly, William, and Ross Levine. 2001."What Have We Learned Washington, DC. from a Decade of Empirical Research on Growth? It's Not Fac- Donsky, Martin. 1990."Putting a Spin on the Yarn Business: Mac- tor Accumulation: Stylized Facts and Growth Models." World field Tries to Get a Leg Up on Its Competition." Business North Bank Economic Review 15 (2): 177­219. Carolina, September 1. Easterly, William, and Yaw Nyarko. 2008."Is the Brain Drain Good Dorosh, Paul A., Steven Haggblade, and Simon Dradri. 2007. for Africa?" Working Paper 19, Brookings Global Economy and "Cotton in Zambia: Alternative Instruments for Ensuring Food Development, Washington, DC. Security and Price Stability in Zambia." Collaborative Working Eaton, Jonathan, and Zvi Eckstein. 1997."Cities and Growth: The- Paper ZM-FSRP-WP-29, International Development, East Lan- ory and Evidence from France and Japan." Regional Science and sing, MI. Urban Economics 27 (4­5): 443­74. Douthat, Ross, Marshall Poe, and Abigail Cutler. 2006."The Poor The Economist. 2007a."A Flourishing Slum." The Economist, get Richer." Atlantic Monthly. December 9. Drabenstott, Mark. 2005. A Review of the Federal Role in Regional ------. 2007b."Chile and Peru. Neighbours, But Not Yet Friends." Economic Development. Kansas City, MO: Federal Reserve Bank The Economist, June 28. of Kansas City. The Economist Economic Intelligence Unit. 2007. China/India Tech- Duarte, Angelo José Mont Alverne, Pedro Cavalcanti Gomes Fer- nology: Leapfrogging or Piggybacking? London: The Economist. reira, and Marcio Antonio Salvato Jr. 2004."Regional or Educa- Eeckhout, Jan. 2004."Gibrat's Law for (All) Cities." American Eco- tional Disparities? A Counterfactual Exercise." Working Paper nomic Review 94 (5): 1429­51. 532, Fundação Getulio Vargas Department of Economics, Sâo Ellison, Glenn, and Edward L. Glaeser. 1997."Geographic Con- Paulo. centration in U.S. Manufacturing Industries: A Dartboard Dumais, Guy, Glenn Ellison, and Edward L. Glaeser. 2002."Geo- Approach." Journal of Political Economy 105 (5): 889­927. graphic Concentration as a Dynamic Process." Review of Eco- Engel, Eduardo, Alexander Galetovic, and Claudio Raddatz. 1998. nomics and Statistics 84 (2): 193­204. "Taxes and Income Distribution in Chile: Some Unpleasant Dumke, Rolf H. 1994."Urban Inequality in Wilhemine Germany." Redistributive Arithmetic." Documentos de Trabajo 41, Centro Institut für Volkswirtscaftslehre, Universitat der Bundeswehr de Economía Aplicada, Universidad de Chile, Santiago, Chile. Munchen, Munich, Germany. Engels, Friedrich. 1987. The Condition of the Working Class in Eng- Dunlap, David W. 1992."Some Land-Use History Highlights." New land. New York: Penguin Books (original published in German York Times, April 12. in 1945). Dunlevy, James A., and Don Bellante. 1983."Net Migration, Escobal, Javier, and Maximo Torero. 2000."Does Geography Endogenous Incomes and the Speed of Adjustment to the Explain Differences in Economic Growth in Peru?" Research North-South Differential." Review of Economics and Statistics 65 Network Working Paper R-404, Inter-American Development (1): 66­75. Bank, Washington, DC. Dunne, Timothy, Mark J. Roberts, and Larry Samuelson. 1989a. Estache, Antonio, and Marianne Fay. 2007."Current Debates on "Plant Turnover and Gross Employment Flows in the U.S. Infrastructure Policy." Policy Research Working Paper 4410, Manufacturing Sector." Journal of Labor Economics 7 (1): World Bank, Washington, DC. 48­71. Estache, Antonio, Marianela González, and Lourdes Trujillo. 2002. ------. 1989b."The Growth and Failure of U.S. Manufacturing "Efficiency Gains from Port Reform and the Potential for Yard- Plants." Quarterly Journal of Economics 104 (4): 671­98. stick Competition: Lessons from Mexico." World Development Duranton, Gilles, and Henry G. Overman. 2002."Testing for 30 (4): 545­60. Localization Using Micro-Geographic Data." Working Paper Ethier, Wilfred J. 1982."National and International Returns to 540, Centre for Economic Performance, London. Scale in the Modern Theory of International Trade." American ------. 2007."Testing for Localization Using Micro-Geographic Economic Review 72(3):389­405. Data." Review of Economic Studies 72(4):1077­106. Euractiv. 2008. Commission Defends Proposal on Pension `Portabil- Duranton, Gilles, and Diego Puga. 2000."Diversity and Specializa- ity.' Brussels: Euractiv. tion in Cities: Why, Where, and When Does It Matter?" Urban European Commission. 2001. Unity, Solidarity, Diversity for Studies 37 (3): 533­55. Europe, its People and its Territory: Second Report on Economic ------. 2001."Nursery Cities: Urban Diversity, Process Innova- and Social Cohesion. Brussels: European Commission. tion, and the Life Cycle of Products." American Economic Review ------. 2002. European Union Public Finance. Luxembourg: Euro- 91 (5): 1454­77. pean Commission, Office for Official Publications of the Euro- ------. 2004."Micro-Foundations of Urban Agglomeration pean Communities. Economies." In Handbook of Urban and Regional Economies, ------. 2007. Growing Regions, Growing Europe. Brussels: Euro- vol. 4, ed. J. Vernon Henderson and Jacque Thisse. Amsterdam: pean Communities. North-Holland. Evans, Carolyn L., and James E. Harrigan. 2005."Distance, Time, ------. 2005."From Sectoral to Functional Urban Specialization." and Specialization: Lean Retailing in General Equilibrium." Journal of Urban Economics 57 (2): 343­70. American Economic Review 95 (1): 292­313. References 313 Evans, Peter. 2002. Livable Cities? Urban Struggles for Livelihood Fink, Carsten, and Aaditya Mattoo. 2004."Regional Agreements and Sustainability. Berkeley, CA: University of California Press. and Trade in Services: Policy Issues." Journal of Economic Inte- Evenett, Simon J. 2005."Competitive Liberalization: A Tourna- gration 19 (4): 742­79. ment Theory-Based Interpretation." Paper presented at the Fink, Carsten, Aaditya Mattoo, and Ileana Cristina Neagu. 2005. Sequencing of Regional Economic Integration: Issues in the "Assessing the Impact of Communication Costs on Interna- Breadth and Depth of Economic Integration in the Americas, tional Trade." Journal of International Economics 67 (2): 428­45. Notre Dame, IN. Fischer, Carolyn, Winston Harrington, and Ian W. H. Parry. 2007. Ezcurra, Roberto, and Pedro Pascual. 2007."Spatial Disparities in "Should Automobile Fuel Economy Standards Be Tightened?" Productivity in Central and Eastern Europe." Eastern European Energy Journal 28 (4): 1­30. Economics 45 (3): 5­32. Fogel, Robert W. 1979."Notes on the Social Savings Controversy." Fafchamps, Marcel, and Klaus Desmet. 2000."The Changing Spa- Journal of Economic History 39 (1): 1­54. tial Distribution of Economic Activity across U.S. Counties." Fontagné, Lionel, Thierry Mayer, and Soledad Zignago. 2005. Economic Working Paper 43, Oxford, U.K. "Trade in the Triad: How Easy Is the Access to Large Markets?" Faini, Riccardo. 1983."Cumulative Process of Deindustrialization Canadian Journal of Economics 38 (4): 1401­30. in an Open Region: the Case of Southern Italy." Journal of Devel- Foroutan, Faezeh, and Lant Pritchett. 1993."Intra-Sub-Saharan opment Economics 12 (3): 277­301. African Trade: Is It Too Little?" Journal of African Economies 2 ------. 1996."Increasing Returns, Migration, and Convergence." (1): 74­105. Journal of Development Economics 49 (1): 121­36. Frankel, Jeffrey A., Ernesto Stein, and Shang-Jin Wei. 1996. Faini, Riccardo, Curzio Giannini, and Giampaolo Galli. 1993. "Regional Trade Arrangements: Natural or Supernatural." Amer- "Finance and Development: The Case of Southern Italy." In ican Economic Review 86 (2): 52­6. Finance and Development: Issues and Experience, ed. Alberto Freund, Caroline L., and Diana Weinhold. 2004."The Effect of the Giovannini. Cambridge, U.K.: Cambridge University Press. Internet on International Trade." Journal of International Eco- Fan, Shenggen, Xiaobo Zhang, and Neetha Rao. 2004."Public nomics 62 (1): 171­89. Expenditure, Growth, and Poverty Reduction in Rural Uganda." Fuentes, Ricardo, and Andrés Montes. 2004."Mexico and the Mil- IFPRI Working Paper 4, International Food Policy Research lennium Development Goals in a Subnational Level." Journal of Institute, Washington, DC. Human Development 5 (1): 97­120. Farvacque-Vitkovic, Catherine, Alicia Casalis, Christian Eghoff, Fujita, Masahisa, ed. 2007. Regional Integration in East Asia: From and Mahine Diop. 2007."Development of the Cities of Mali: the Viewpoint of Spatial Economics. New York: Palgrave Mac- Challenges and Priorities." Africa Region Working Paper 104, millan. World Bank, Washington, DC. Fujita, Masahisa, and Ryoichi Ishii. 1999."Global Location Behav- Fay, Marianne, John Felkner, and Somik V. Lall. 2008."Market ior and Organizational Dynamics of Japanese Electronics Firms Access and Firm Productivity in the Russian Federation." World and Their Impact on Regional Economies." In The Dynamic Bank, Washington, DC. Processed. Firm: The Role of Technology, Strategy, Organization, and Fay, Marianne, and Charlotte Opal. 2000."Urbanization without Regions, ed. Alfred D. Chandler Jr., Peter Hagstrom, and Orjan Growth: A Not-So-Uncommon Phenomenon." Policy Research Solvell. Oxford, U.K.: Oxford University Press. Working Paper 2412, World Bank, Washington, DC. Fujita, Masahisa, and Takatoshi Tabuchi. 1997."Regional Growth Fearon, James D., and David D. Laitin. 2003."Additional Tables for in Post War Japan." Regional Science and Urban Economics 27 Ethnicity, Insurgency, and Civil War." Stanford University, Palo (6): 643­70. Alto, CA. Fujita, Masahisa, Paul Krugman, and Anthony J. Venables. 2001. Feder, Gershon, and Peter Lanjouw. 2001."Rural Non-farm Activi- The Spatial Economy. Cities, Regions, and International Trade. ties and Rural Development: From Experience towards Strat- Cambridge, MA: MIT Press. egy." Rural Development Strategy Background Paper 4, World Fuss, Melvyn A., and Vinod K. Gupta. 1981."A Cost Function Bank, Washington, DC. Approach to the Estimation of Minimum Efficient Scale, Feenstra, Robert C., and Gordon H. Hanson. 1997."Foreign Direct Returns to Scale, and Suboptimal Capacity: With an Application Investment and Relative Wages: Evidence from Mexico's Maqui- to Canadian Manufacturing." European Economic Review 15 (2): ladoras." Journal of International Economics 42 (3­4): 371­93. 123­35. Feldman, Maryann P. 1994. The Geography of Innovation. Boston, Gabaix, Xavier, and Yannis M. Ioannides. 2004."Evolution of City MA: Kluwer. Size Distributions." In Handbook of Urban and Regional Eco- Feldman, Maryann P., and David B. Audretsch. 1999."Innovation nomics, vol. 4, ed. J. Vernon Henderson and Jacque-Francois in Cities: Science-Based Diversity, Specialization, and Localized Thisse. Amsterdam: Elsevier. Competition." European Economic Review 43 (2): 409­29. Gakenheimer, Ralph. 1999."Urban Mobility in the Developing Ferreira, Afonso. 2000."Convergence in Brazil: Recent Trends and World." Transportation Research Part A: Policy and Practice 33 Long-Run Prospects." Journal of Applied Economics 3 (4): 479­89. (7): 671­89. Fingleton, Bernard. 2003."Increasing Returns: Evidence from Local Gallaway, Lowell E., and Richard K. Vedder. 1971."Mobility of Wages in Great Britain." Oxford Economic Papers 55 (4): 716­39. Native Americans." Journal of Economic History 31 (3): 613­49. 314 WORLD DEVELOPMENT REPORT 2009 Gallup, John L., Jeffrey D. Sachs, and Andrew D. Mellinger. 1999. Goodlad, R. 1996."The Housing Challenge in South Africa." "Geography and Economic Development." International Urban Studies 33 (9): 1629­45. Regional Science Review 22 (2): 179­232. Gordon, Robert J. 1990. The Measurement of Durable Goods Prices. Garau, Pietro. 2008. Notes for the State of the World's Cities Report Chicago: University of Chicago Press. 2008: Policy Analysis and Recommendations. Nairobi: UN- Gorzelak, Grzegorz. 2007."Cohesion and Convergence: Synonyms HABITAT. or Two Different Notions?" Paper presented at the Policy Work- García-Verdú, Rodrigo. 2005."Income, Mortality, and Literacy shop "Spatial Disparities and Development Policy," Berlin. Distribution Dynamics across States in Mexico: 1940­2000." Gottman, Jean. 1977."Megalopolis and Antipolis: The Telephone Cuadernos de Economía 42 (125): 165­92. and the Structure of the City." In Ithiel de Sola Pool, (eds.), The Garrett, James. 2005. Mobility, Migration, and Rural-Urban Social Impact of the Telephone. Cambridge, MA: Massachusetts Changes, Brief 2. Washington, DC: International Food Policy Institute of Technology. Research Institute. Government of India Planning Commission. 2002a. 10th Five Year Garrett, Power. 2002."Slums of Baltimore." In From Mobtown to Plan (2002­2007). New Delhi: Government of India, Planning Charm City: New Perspectives on Baltimore's Past, ed. Jessica L. Commission. Elfenbein, John R. Breihan, and Thomas L. Hollowak. Balti- ------. 2002b. National Human Development Report 2001. New more, MD: Maryland Historical Society. Delhi: Government of India, Planning Commission. Gaspar, Jess, and Edward L. Glaeser. 1998."Information Tech- ------. 2006. Towards Faster and More Inclusive Growth: An nology and the Future of Cities." Journal of Urban Economics approach to the Eleventh Five Year Plan 2007­2012. New Delhi: 43(1):136­56. Government of India. Gastner, Michael T. and Mark E. J. Newman. 2004."Diffusion- Government of Malaysia. 2001. Eight Malaysia Plan: 2001­2005. based method for producing density-equalizing maps." Proceed- Kuala Lumpur: Malaysia Economic Planning Unit. ings of the National Academy of Sciences, 101, 20:7499­7504. Government of Nigeria, Federal Ministry of Science and Technol- Gauthier, Jean-Paul. 2007. Special Economic Zones: Performance, ogy. 2003. Multi-Disciplinary Committee Report of the Techno- Lessons Learned, and Implications for Zone Development. Wash- Economic Survey on Wood and Wood Products Sector, 2003, 4th ington, DC: World Bank, Foreign Investment Advisory Service. Update. Lagos: Government of Nigeria, Federal Ministry of Sci- Gibbon, Peter, and Stefano Ponte. 2005. Trading Down: Africa, ence and Technology. Value Chains, and the Global Economy. Philadelphia: Temple Government of Sri Lanka. 2008. Poverty Indicators Household Income University Press. and Expenditure Survey, 2006­07. Colombo: Sri Lanka Depart- Gibson, John, and Scott Rozelle. 2003."Poverty and Access to ment of Census and Statistics, Ministry of Planning and Finance. Roads in Papua New Guinea." Economic Development and Cul- Goyal, Sanjeev, and Klaas Staal. 2004."The Political Economy of tural Change 52 (1): 159­85. Regionalism." European Economic Review 48 (3): 563­93. Gill, Indermit, and Homi Kharas. 2007. An East Asia Renaissance: Green, Alan G. 1969."Regional Inequality, Structural Change, and Ideas for Economic Growth. Washington, DC: World Bank. Economic Growth in Canada, 1890­1956." Economic Develop- Gilman, Sidney. 1984. The Competitive Dynamics of Container ment and Cultural Change 17 (4): 567­83. Shipping. Liverpool, U.K.: Gower Publishing. Gregoire, Emmanuel, and Pascal Labazee. 1993. Grands Commer- Glaeser, Edward L., and Matthew E. Kahn. 2001."Decentralized çants d'Afrique de l'Ouest. Paris: Karthala-Orstom. Employment and the Transformation of the American City." Greytak, David, and Paul Blackley. 1985."Labor Productivity and Working Paper 1912, Harvard Institute of Economic Research, Local Industry Size: Further Issues in Assessing Agglomeration Cambridge, MA. Economies." Southern Economic Journal 51 (4): 1121­29. Glaeser, Edward L., Hedi D. Kallal, José A. Scheinkman, and Andrei Gribbin, August. 2000."Overpopulated Megacities Face Frighten- Shleifer. 1992."Growth in Cities." Journal of Political Economy ing Future: Overpopulation in Developing Countries Impact on 100 (6): 1126­52. the West." Insight on the News, August 21. Glaeser, Edward L., Jed Kolko, and Albert Saiz. 2001."Consumer Griliches, Zvi, and Vidar Ringstad. 1971. Economies of Scale and the City." Journal of Economic Geography 1 (1): 27­50. Form of the Production Function: An Econometric Study of Nor- Glaeser, Edward L., and David C. Maré. 2001."Cities and Skills." wegian Manufacturing Establishment Data. Amsterdam: North- Journal of Labor Economics 19 (2): 316­42. Holland. Glaeser, Edward L., José A. Scheinkman, and Andrei Schleifer. Grossman, Gene M., and Elhanan Helpman. 1991."Quality Lad- 1995."Economic Growth in a Cross-Section of Cities." Journal ders in the Theory of Growth." Review of Economic Studies of Monetary Economics 36 (1): 117­43. 58(1):43­61. Golob, Thomas F., and David Brownstone. 2005."The Impact of ------. 1995."Technology and Trade." In Gene M. Grossman and Residential Density on Vehicle Usage and Energy Consump- Kenneth Rogoff, (eds.), Handbook of International Economics, tion." Energy Institute Policy and Economics Working Paper Volume III. Amsterdam, Netherlands: North-Holland. EPE-011, University of California, Berkeley, CA. Groupe Recherche--Actions pour le Développement. 2001. Good, David F. 1986."Uneven Development in the Nineteenth "Potentialités et Conflits dans les Zones Péri-urbaines: Le Cas Century: A Comparison of the Habsburg Empire and the de Bamako au Mali." Rural-Urban Interactions and Livelihood United States." Journal of Economic History 46 (1): 137­51. Strategies Working Paper 5, IIED, London. References 315 Guasch, Juan Luis. 2008. Exporting by Mail: An Innovative Trade Harley, C. Knick. 1980."Transportation, the World Wheat Trade, Facilitation Tool. Washington, DC: World Bank, IFC Smart Les- and the Kuznets Cycle, 1850­1913." Explorations in Economic sons. History 17 (3): 218­50. Guerra, Carlos A., Priscilla W. Gikandi, Andrew J. Tatem, Abdis- Harrigan, James E. 2005."Airplanes and Comparative Advantage." alan M. Noor, Dave L. Smith, Simon I. Hay, and Robert W. NBER Working Paper 11688, National Bureau of Economic Snow. 2008."The Limits and Intensity of Plasmodium Falci- Research, Cambridge, MA. parum Transmission: Implications for Malaria Control and Harrigan, James E., and Anthony J. Venables. 2006. "Timeli- Elimination Worldwide." Plos Medicine 5 (2): e38­e38. ness and Agglomeration." Journal of Urban Economics 59 (2): Gupta, Vinod K. 1983."Labor Productivity, Establishment Size, 300­16. and Scale Economies." Southern Economic Journal 49(3):853­9. Harris, John R., and Michael P. Todaro. 1970."Migration, Unem- Gwatkin, Davidson R., Shea Rutstein, Kiersten Johnson, Eldaw Suli- ployment, and Development: A Two-Sector Analysis." American man,Adam Wagstaff, and Agbessi Amouzou. 2007. Socio-Economic Economic Review 60 (1): 126­42. Differences in Health, Nutrition, and Population within Developing Hau, Timothy. 1990."Electronic Road Pricing: Developments in Countries: An Overview.Washington, DC: World Bank. Hong Kong, 1983­1989." Journal of Transport Economics and Hall, Carolyn. 1984."Regional Inequalities in Well-Being in Costa Policy 24 (2): 203­14. Rica." Geographical Review 74 (1): 48­62. Haussmann, Ricardo, and Dani Rodrik. 2003."Economic Develop- Hall, Peter. 2002. Cities of Tomorrow. Oxford, U.K.: Blackwell Pub- ment as Self-Discovery." Journal of Development Economics 72 lishing. (2): 603­33. Hall, Robert E. 1988."The Relation between Price and Marginal Haussmann, Ricardo, Dani Rodrik, and Andrés Velasco. 2005. Cost in U.S. Industry." Journal of Political Economy 96 (5): "Growth Diagnostics." Center for International Development, 921­47. Cambridge, MA. ------. 1990."Invariance Properties of Solow's Productivity Hay, Donald Andrew. 1979."The Location of Industry in a Devel- Residual." In Growth, Productivity, Unemployment: Essays to oping Country: The Case of Brazil." Oxford Economic Papers 31 Celebrate Bob Solow's Birthday, ed. Peter Diamond. Cambridge, (1): 93­120. MA: MIT Press. Hay, Simon I., Carlos A. Guerra, Andrew J. Tatem, Abdisalan M. Hall, Thomas, and Sonja Vidén. 2005."The Million Homes Pro- Noor, and Robert W. Snow. 2004."The Global Distribution and gramme: A Review of the Great Swedish Planning Model." Plan- Population at Risk of Malaria: Past, Present, and Future." Lancet ning Perspectives 20 (3): 301­28. Infectious Diseases 4 (6): 327­36. Hamaguchi, Nobuaki. Forthcoming."Regional Integration, Hay, Simon I., David L. Smith, and Robert W. Snow. 2008."Mea- Agglomeration, and Income Distribution in East Asia." In suring Malaria Endemicity from Intense to Interrupted Trans- Reshaping Economic Geography in East Asia, ed. Yukon Huang mission." Lancet Infectious Diseases 8 (6): 369­78. and Alessandro Magnoli Bocchi. Washington, DC: World Bank, Hayashi, Takehisa. 2003. Readings on Local Public Finance. Tokyo: EAP Companion Volume to the WDR 2009. Toyo Keizai Shinposya. Hammond, George William, and Jürgen von Hagen. 1994."Indus- Hazlewood, Arthur. 1979."The End of the East African Commu- trial Localization: An Empirical Test for Marshallian Localiza- nity: What Are the Lessons for Regional Integration Schemes?" tion Economies." Working Paper 917, Centre for Economic Journal of Common Market Studies 18 (1): 40­58. Policy Research, London. He, Canfei. Forthcoming."Industrial Agglomeration and Eco- Hancock, Diana, David B. Humphrey, and James A. Wilcox. 1999. nomic Performance in the Transitional China." In Reshaping "Cost Reductions in Electronic Payments: The Roles of Consoli- Economic Geography in East Asia, ed. Yukon Huang and Ales- dation, Economies of Scale, and Technical Change." Journal of sandro Magnoli Bocchi. Washington, DC: World Bank, EAP Banking and Finance 23 (2­4): 391­421. Companion Volume to the WDR 2009. Hansen, Eric R. 1990."Agglomeration Economics and Industrial Head, Charles Keith, and Thierry Mayer. 2006."Regional Wage and Decentralization: The Wage-Productivity Trade-Offs." Journal of Employment Responses to Market Potential in the EU." Regional Urban Economics 28(2):140­59. Science and Urban Economics 36 (5): 573­94. Hanson, Gordon H. 1996."Agglomeration, Dispersion, and the Head, Charles Keith, John Reis, and Deborah Swenson. 1995. Pioneer Firm." Journal of Urban Economics 39 (3): 255­81. "Agglomeration Benefits and Location Choice: Evidence from ------. 1998a."North American Economic Integration and Japanese Manufacturing Investment in the United States." Jour- Industry Location." Oxford Review of Economic Policy 14 (2): nal of International Economics 38 (3­4): 223­47. 30­44. Heideloff, Christel, and Manfred Zachcial. 2006. Shipping Statisti- ------. 1998b."Regional Adjustment to Trade Liberalization." cal Yearbook 2006. Bremen, Germany: Institute of Shipping Eco- Regional Science and Urban Economics 28 (4): 419­44. nomics and Logistics. Hardoy, Jorge E., and David Satterthwaite. 1989. Squatter Citizen: Helliwell, John F. 2002."Measuring the Width of National Bor- Life in the Urban Third World. London: Earthscan. ders." Review of International Economics 10 (3): 517­24. Hargan, Jim. 2007."Prince Charles Poundbury." Brittish Heritage. Helpman, Elhanan, and Paul Krugman. 1987. Market Structure and May, 2007. Foreign Trade. Cambridge, MA: MIT Press. 316 WORLD DEVELOPMENT REPORT 2009 Henderson, J. Vernon. 1974."The Sizes and Types of Cities." Amer- Hill, Hal, Budy Resosudarmo, and Yogi Vidyattama. 2007."Indo- ican Economic Review 64 (4): 640­56. nesia's Changing Economic Geography." Working Papers in Eco- ------. 1986."Efficiency of Resources Usage and City Size." Jour- nomics and Development Studies 2007-13, Bandung, Indonesia. nal of Urban Economics 19 (1): 47­70. Hoekman, Bernard. 2006."Liberalizing Trade in Services: A Sur- ------. 1994."Where Does an Industry Locate?" Journal of Urban vey." Policy Research Working Paper 4030, World Bank, Wash- Economics 35(1):83­104. ington, DC. ------. 1997a."Externalities and Industrial Development." Jour- Hoekman, Bernard, and Dominique Njinkeu. 2007."Aid for Trade nal of Urban Economics 42 (3): 449­70. and Export Competitiveness: New Opportunities for Africa." ------. 1997b."Medium Size Cities." Regional Science and Urban World Bank, Washington, DC. Economics 27 (6): 583­612. Holmes, Thomas J. 1999."Localization of Industry and Verti- ------. 2003a."Marshall's Scale Economies." Journal of Urban cal Disintegration." Review of Economics and Statistics 81 (2): Economics 53 (1): 1­28. 314­25. ------. 2003b."The Urbanization Process and Economic Growth: Holmes, Thomas J., and John J. Stevens. 2002."Geographic Con- The So-What Question." Journal of Economic Growth 8 (1): centration and Establishment Scale." Review of Economics and 47­71. Statistics 84 (4): 682­91. ------. 2005a."Development and Growth." In Handbook of Eco- Holzmann, Hajo, and Sebastian Vollmer. 2008."The Distribution nomic Growth, vol. 1, part B, ed. Philippe Aghion and Steven N. of Income across German Counties, 1992­2005." University of Durlauf. Amsterdam: North-Holland. Goettingen, Goettingen, Germany. ------. 2005b."Urbanization and Growth." In Handbook of Eco- Hoover, Edgar M., and Raymond Vernon. 1959. Anatomy of a nomic Growth, ed. Philippe Aghion and Steven N. Durlauf. Metropolis: The Changing Distribution of People and Jobs within Amsterdam: North-Holland. the New York Metropolitan Area. Cambridge, MA: Harvard Uni- versity Press. ------. 2007. Urbanization in China: Policy Issues and Options. New York: China Economic Research and Advisory Programme. Huang, Yukon, and Xubei Luo. Forthcoming."Reshaping Eco- nomic Geography in China." In Reshaping Economic Geography Henderson, J. Vernon, Ari Kuncoro, and Damhuri Nasution. 1996. in East Asia, ed. Yukon Huang and Alessandro Magnoli Bocchi. "The Dynamics of Jabotabek Development." Bulletin of Indone- Washington, DC: World Bank. sian Economic Studies 32 (1): 71­95. Huber, Peter W. 1995."New York, Capital of the Information Age." Henderson, J. Vernon, Ari Kuncoro, and Matthew Turner. 1995. New York City Journal 5(1):12­22. "Industrial Development in Cities." Journal of Political Economy 103 (5): 1067­90. Hughes, Joseph P., William W. Lang, Loretta J. Mester, and Choon- Geol Moon. 1999."The Dollars and Sense of Bank Consolida- Henderson, J.Vernon, Todd Lee, and Yung Joon Lee. 2001."Scale tion." Journal of Banking and Finance 23 (2­4): 291­324. Externalities in Korea."Journal of Urban Economics 49 (3): 479­504. Hughes, Joseph P., and Loretta J. Mester. 1998."Bank Capitaliza- Henderson, J. Vernon, Zmarak Shalizi, and Anthony J. Venables. tion and Cost: Evidence of Scale Economies in Risk Manage- 2001."Geography and Development." Journal of Economic Geog- ment and Signaling." Review of Economics and Statistics 80 (2): raphy 1 (1): 81­105. 314­25. Henriksen, Ingrid. 2003."Freehold Tenure in Late Eighteenth Cen- Hulten, Charles R. 2007."Transportation Infrastructure, Produc- tury Denmark." In Advances in Agricultural Economic History, tivity and Externalities." In ECMT Round Tables No. 132: Trans- ed. Kyle D. Kaufman. New York: JAI Press. port, Infrastructure Investment and Economic Productivity, ed. Henry, Peter Blair. 2007."Capital Account Liberalization: Theory, OECD/ECMT. Paris: Organisation for Economic Co-operation Evidence, and Speculation." Journal of Economic Literature 45 and Development. (4): 887­935. Hummels, David. 2001. Time as a Trade Barrier. West Lafayette, Hering, Laura, and Sandra Poncet. 2006."Market Access and Indi- IN: Purdue University Press. vidual Wages: Evidence from China." Working Paper 2006-23, ------. 2007."Transportation Costs and International Trade in Centre d'Études Prospectives et d'Informations Internationales, the Second Era of Globalization." Journal of Economic Perspec- Paris. tives 21 (3): 131­54. Herzog, Henry W. Jr., and Alan M. Schlottmann, eds. 1991. Industry Hummels, David, Volodymyr Lugovskyy, and Alexandre Skiba. Location and Public Policy. Knoxville, TN: University of Tennessee. 2007."The Trade Reducing Effect of Market Power in Interna- Hewings, G., E. Feser, and K. Poole. 2008."Spatial/Territorial tional Shipping." NBER Working Paper 12914, National Bureau Development Policies in the United States." Background paper of Economic Research, Cambridge, MA. for the WDR 2009. Hunt, Edward H. 1986."Industrialization and Regional Inequality: Hidalgo, C. A., B. Klinger, A. L. Barabási, and R. Hausmann. Wages in Britain, 1760­1914." Journal of Economic History 46 2007."The Product Space Conditions for the Development of (4): 935­66. Nations." Science 317 (5837): 482­87. Hunt, Jennifer. 2000."Why Do People Still Live in East Germany?" Hill, Fiona, and Clifford Gaddy. 2003. The Siberia Curse: How NBER Working Paper 7564, National Bureau of Economic Communist Planners Left Russia out in the Cold. Washington, Research, Cambridge, MA. DC: Brookings Institution Press. References 317 Hurd, John II. 1975."Railways and the Expansion of Markets in Jensen, Robert. 2007."The Digital Provide: Information (Tech- India, 1861­1921." Explorations in Economic History 12 (3): nology), Market Performance, and Welfare in the South Indian 263­88. Fisheries Sector." Quarterly Journal of Economics 122 (3): 879­ Iliffe, John. 1995. Africans: The History of a Continent. Cambridge, 924. U.K.: Cambridge University Press. Jessen, Anneje. 2002."Regional Public Goods and Small Econo- Independent Evaluation Group (IEG). 2007."The Development mies: The Caribbean Regional Negotiating Machinery." In Potential of Regional Programs: An Evaluation of World Bank Regional Public Goods: From Theory to Practice, ed. Antoni Este- Support of Multicountry Operations." Paper presented at the vadeordal, Brian Frantz, and Tam R. Nguyen. Washington, DC: IEG Conference on Unlocking the Potential of Regional Devel- Inter-American Development Bank and Asian Development opment Programs, Washington, DC. Bank. Ingram, Gregory. 1997."Patterns of Metropolitan Development: Jones, Gareth A., and Kavita Datta. 2000."Enabling Markets to What Have We Learned?" Policy Research Working Paper 1841, Work? Housing Policy in the `New' South Africa." International World Bank, Washington, DC. Planning Studies 5 (3): 393­416. Ingram, Gregory, and John Whitehead. 2008."The Distribu- Jourdan, Paul, and NEPAD. 2006."Regional Strategies. The Case tion and Concentration of Population in the U.S., 1900­2000." for a Resource-Based Spatial Development Programme." Paper Working Paper WP08GI1, Lincoln Institute of Land Policy, presented at the U.S.­Africa Infrastructure Conference, Wash- Cambridge, MA. ington, DC. Instituto Brasileiro de Geografia e Estatística (IBGE). 2004. Perfil Junius, Karsten. 1997."Economies of Scale: A Survey of Empirical Dos Municipios Brasileiros. Rio de Janeiro: IBGE. Literature." Working Paper 813, Kiel Institute of World Eco- Inter-American Development Bank. Forthcoming. Unclogging nomics, Kiel, Germany. the Arteries: A Report on the Impact of Transport Costs on Latin Justman, Moshe. 1994."The Effect of Local Demand on Industry American and Caribbean Trade. Cambridge, MA: Harvard Uni- Location." Review of Economics and Statistics 76 (4): 742­53. versity Press. Kaldor, Nicholas. 1972."The Irrelevance of Equilibrium Econom- International Air Transport Association. 2007a. Aviation Economic ics." Economic Journal 82 (328): 1237­55. Benefits: Issues Briefing No. 8. Montreal, Canada: International Kamada, Kimiyoshi, Nobuhiro Okuno, and Ritsuko Futagami. Air Transport Association. 1998."Decisions on Regional Allocation of Public Investment: ------. 2007b. Economics Briefing: Air Freight Market Outlook. the Case of Japan." Applied Economic Letters 5 (8): 503­6. Montreal, Canada: International Air Transport Association. Kanbur, Ravi, and Anthony J. Venables. 2005. Rising Spatial Dis- International Development Association, and World Bank. 2008. parities and Development: Why Do they Matter? Oxford, U.K.: Regional Integration Assistance Strategy. Washington, DC: World Oxford University Press, UNU-WIDER Studies in Development Bank. Economics. International Monetary Fund. 1998. World Economic Outlook. Kanbur, Ravi, and Xiaobo Zhang. 2005."Fifty Years of Regional Financial Crises: Causes and Indicators. Washington, DC: Inter- Inequality in China: A Journey through Central Planning, national Monetary Fund. Reform, and Openness." Review of Development Economics 9 (1): ------. 2007. Directions of Trade. Washington, DC: International 87­106. Monetary Fund. Kasekende, Louis A., and Nehemiah Ng'eno. 1999."Regional Inte- Ivaldi, Marc, and Gerard McCullough. 2007."Railroad Pricing and gration and Economic Liberalization in Eastern and Southern Revenue-to-Cost Margins in the Post-Staggers Era." In Research Africa." In Regional Integration and Trade Liberalization in Sub- in Transportation Economics, ed. Scott Dennis and Wayne Talley. Saharan Africa, Volume 3: Regional Case-Studies, ed. Ademola Amsterdam: Elsevier. Oyejide, Ibrahim Elbadawi, and Stephen Yeo. London: Palgrave Macmillan Press. Jacobs, Jane. 1970. The Economy of Cities. New York: Vintage. Kawashima, Tatsuhiko. 1975."Urban Agglomeration Economies Jacoby, Hanan G. 2000."Access to Markets and the Benefits of in Manufacturing Industries." Papers in Regional Science 34 (1): Rural Roads." Economic Journal 110 (465): 713­37. 157­72. Jaffe, Adam B. 1986."Technological Opportunity and Spillover of Kearns, Kevin C. 2006. Dublin's Lost Heroines: Mammies and Gran- R&D: Evidence from Firms' Patents, Profits, and Market Value." nies in a Vanished Dublin. Park West, Dublin: Gill and Macmillan. American Economic Review 76 (5): 984­1001. Kee, Hiau Looi, Alessandro Nicita, and Marcelo Olarreaga. 2006. Jaffe, Adam B., Manuel Trajtenberg, and Rebecca Henderson. 1993. "Estimating Trade Restiveness Indices." Policy Research Working "Geographic Localization of Knowledge Spillovers as Evidence by Paper 3840, World Bank, Washington, DC. Patent Citations." Quarterly Journal of Economics 108 (3): 577­98. Keefer, Philip, and Stuti Khemani. 2005."Democracy, Public Jayne, Thomas S., B. Zulu, D. Mather, E. Mghenyi, E. Chirwa, and Expenditures, and the Poor: Understanding Political Incentives David L. Tschirley. 2005."Maize Marketing and Trade Policy in for Providing Public Services." World Bank Research Observer a Pro-Poor Agricultural Growth Strategy: Insights from House- Advance Access 20 (1): 1­27. hold Surveys in Eastern and Southern Africa." Paper presented at Toward Improved Maize Marketing and Trade Policies in the Kenworthy, Jeff. 2005."Transport Energy Use and Greenhouse Southern Africa Region, Centurion, South Africa. Gases in Urban Passenger Transport Systems: A Study of 84 318 WORLD DEVELOPMENT REPORT 2009 Global Cities." Institute for Sustainability and Technology Pol- Kraske, Jochen, William H. Becker, William Diamond, and Louis icy, Murdoch University, Murdoch, Australia. Galambos. 1996. Bankers with a Mission: The Presidents of the Kenya Ministry of Planning and National Development and World Bank, 1946­91. Oxford, U.K.: Oxford University Press. UNDP. 2005. Millennium Development Goals Country Report Krishna, Pravin. 1998."Regionalism and Multilateralism: A Politi- 2005: Kenya. Nairobi: United Nations and Kenya Ministry of cal Economy Approach." Quarterly Journal of Economics 113 (1): Planning and National Development. 227­50. Khanna, Tarun. 2008. Billions of Entrepreneurs: How China and Kroehnert, Steffen, Franziska Medicus, and Reiner Kinglolz. 2006. India Are Reshaping Their Futures--and Yours. Cambridge, MA: Die Demografische Lage der Nation. Wie Zukunftsfaehig sind Harvard Business School Press. Deutschlands Regionen? Daten, Fakten, Analysen. Berlin: Berlin Kilkenny, Maureen. 1998."Economies of Scale. Lecture for Eco- Institute for Population and Development. nomics 376: Rural, Urban and Regional Economics." Iowa State Kroehnert, Steffen, and Sebastian Vollmer. 2008."Where Have All University. Ames, IA. Young Women Gone?" Background paper for the WDR 2009. Kilroy, Austin. 2008."Intra-Urban Spatial Inequalities: Cities as Krueger, Anne O. 2006."Globalization and International Loca- `Urban Regions.'" Background paper for the WDR 2009. tional Competition." Paper presented at the Symposium in Kim, Sukkoo. 1995."Expansion of Markets and the Geographic Honor of Herbert Giersch, Kiel, Germany. Distribution of Economic Activities: The Trends in U.S. Krugman, Paul R. 1980."Scale Economies, Product Differentiation, Regional Manufacturing Structure, 1860­1987." Quarterly Jour- and the Pattern of Trade."American Economic Review 70(5):950­9. nal of Economics 110 (4): 881­908. ------. 1981."Intraindustry Specialization and the Gains from ------. 1999."Regions, Resources, and Economics Geography: Trade." Journal of Political Economy 89(5):959­73. Sources of U.S. Regional Comparative Advantage, 1880­1987." ------. 1991a. Geography and Trade. Cambridge, MA: MIT Press. Regional Science and Urban Economics 29 (1): 1­32. ------. 1991b."Increasing Returns and Economic Geography." Kim, Sukkoo, and Robert A. Margo. 2004."Historical Perspectives Journal of Political Economy 99 (3): 483­99. on U.S. Economic Geography." In Handbook of Regional and ------. 1991c."Is Bilateralism Bad?" In International Trade and Urban Economics, Volume 4: Cities and Geography, ed. J. Vernon Trade Policy, ed. Elhanan Helpman and Assaf Razin. Cambridge, Henderson and Jacque-Francois Thisse. Amsterdam: North- MA: MIT Press. Holland. ------. 1993."First Nature, Second Nature, and Metropolitan Kim, Won Bae. 1998."Korea's Policy of Rural-Urban Integrated Location." Journal of Regional Science 33 (2): 129­44. Cities." Paper presented at the International Workshop on Rural ------. 1995."Innovation and Agglomeration: Two Parables Sug- and Urban Linkages, Curitiba, Brazil. gested by City-Size Distributions." Japan and the World Economy Kloeppinger-Todd, Renate. 2007."ICT and Rural Access to 7 (4): 371­90. Finance." Paper presented at the infoDev/ARD Workshop on ------. 2007."The `New' Economic Geography: Where Are We?" Using ICT to Support Rural Livelihoods, Washington, DC. In Regional Integration in East Asia, ed. Masahisa Fujita. New Knoke, William. 1996. Bold New World: The Essential Road Map to York: Palgrave Macmillan. the Twenty-First Century. New York: Kodansha International. Kuncoro, Ari. Forthcoming."Spatial Agglomeration, Site Pro- Koh, Tommy, and Li Lin Chang, eds. 2005. The Little Red Dot: ductivity and Local Governance: Indonesian Experience from Reflections by Singapore's Diplomats. Singapore: World Scientific 1980 to 2003." In Yukon Huang and Alessandro Magnoli Bocchi Publishing. (eds.) Reshaping Economic Geography in East Asia. Washington, Kolko, Jed. 1998."New England at Your Service: The New Geogra- DC: World Bank, EAP Companion Volume to the WDR 2009. phy of Service Industries." Regional Review 1998 (Q4): 6­11. Kuroda, Haruhiko. 2007."Toward an Integrated, Poverty-Free and ------. 1999."Can I Get Some Service Here? Information Tech- Peaceful East Asia." In East Asia Visions: Perspectives on Economic nology, Service Industries, and the Future of Cities." Ph.D. the- Development, ed. Indermit Gill, Yukon Huang, and Homi Kha- sis, Harvard University, Cambridge, MA. ras. Washington, DC, and Singapore: World Bank. ------. 2000."The Death of Cities? The Death of Distance? Evi- Ladman, Jerry R. 1979."The Economic Interdependence of Con- dence from the Geography of Commercial Internet Usage." In tiguous Border Cities: The Twin City Multiplier." Annals of The Internet Upheaval, ed. Ingo Vogelsang and Benjamin Com- Regional Science 13 (1): 23­28. paine. Cambridge, MA: MIT. Laeven, Luc, and Christopher Woodruff. 2007."The Quality of the ------. 2007."Agglomeration and Co-Agglomeration of Services Legal System, Firm Ownership, and Firm Size." Review of Eco- Industries." Munich Personal RePEc Archive (MPRA) Working nomics and Statistics 89 (4): 601­14. Paper 3362, Munich, Germany. Lall, Somik V. 2005."City Performance and Policy Actions." In Koola, J., and Caglar Özden. 2008."Making the Move: The Effect Inputs for a Strategy of Cities in Brazil, ed. World Bank. Washing- of Migration on Welfare in Uganda." World Bank, Washington, ton, DC: World Bank. DC. Lall, Somik V., Uwe Deichmann, Mattias K. A. Lundberg, and Kose, M. Ayhan, Eswar Prasad, Kenneth Rogoff, and Shang-Jin Nazmul Chaudhury. 2004."Tenure, Diversity, and Commit- Wei. 2006."Financial Globalization: A Reappraisal." Working ment: Community Participation for Urban Service Provision." Paper 06/189, International Monetary Fund, Washington, DC. Journal of Development Studies 40 (3): 1­26. References 319 Lall, Somik V., Elizabeth Schroeder, and Emily Schmidt. 2008. Case of the Apple's iPod." Personal Computing Industry Cen- "Geographically Prioritizing Infrastructure Improvements to ter, Irvine, CA. Accelerate Growth in Uganda." World Bank, Washington, DC. Linn, F. Johannes, and David Tiomkin. 2006."The New Impetus Lall, Somik V., Harris Selod, and Zmarak Shalizi. 2006."Rural- toward Economic Integration between Europe and Asia." Asia Urban Migration in Developing Countries: A Survey of Theo- Europe Journal 4 (1): 31­41. retical Predictions and Empirical Findings." Policy Research Lipsey, Richard G., and Alice Nakamura, eds. 2006. Services Indus- Working Paper 3915, World Bank, Washington, DC. tries and the Knowledge-Based Economy. Calgary, Canada: Uni- Lall, Somik V., Hyoung Gun Wang, and Uwe Deichmann. 2008. versity of Calgary Press. "Infrastructure and City Competitiveness in India." Policy Little, Peter D. 2007."Unofficial Cross-Border Trade in Eastern Research Working Paper, World Bank, Washington, DC. Africa." Paper presented at the Staple Food Trade and Market Land Enquiry Commission. 1914. The Land, vol. II. London: Hod- Policy Options for Promoting Development in Eastern and der and Stoughton. Southern Africa, Rome. Landes, David S. 1998. The Wealth and Poverty of Nations: Why Lösch, Auguste. 1940. The Economics of Location. New Haven, CT: Some Are So Rich and Some So Poor. New York: W. W. Norton. Yale University Press. Lane, C. B. 1860. Railway Communication in London and the Lucas, Robert E. Jr. 1988."On the Mechanics of Economic Devel- Thames Embankment. London: James Ridgway. opment." Journal of Monetary Economics 22 (1): 3­42. Lanne, Markku, and Matti Liski. 2003."Trends and Breaks in Per- ------. 2004."Life Earnings and Rural-Urban Migration." Journal Capita Carbon Dioxide Emissions, 1870­2028." Working Paper of Political Economy 112 (S1): S29­S59. 0302, Center for Energy and Environmental Policy Research, ------. 2007."Trade and the Diffusion of the Industrial Revolu- Cambridge, MA. tion." NBER Working Paper 13286, National Bureau of Eco- Layke, Christian, and Stephen Adam. 2008."Spatial Allocation nomic Research, Cambridge, MA. of Public Expenditures in Nigeria." Background paper for the Lucas, Robert E. B. 1997."Internal Migration in Developing Coun- WDR 2009. tries." In Handbook of Population and Family Economics, ed. Leamer, Edward E. 2007."A Flat World, a Level Playing Field, Mark Rosenzweig and Oded Stark. Amsterdam: North-Holland. a Small World After All, or None of the Above? A Review of ------. 2003."Migration and Lagging Regions." Boston Univer- Thomas L. Friedman's The World Is Flat." Journal of Economic sity, Boston. Processed. Literature 45 (1): 83­126. ------. 2006."Migration and Economic Development in Africa: Leamer, Edward E., and Michael Storper. 2001."The Economic A Review of Evidence." Journal of African Economies 15 (2): Geography of the Internet Age." Journal of International Business 337­95. Studies 32 (4): 641­65. Luo, Xubei. 2004."The Role of Infrastructure Investment Loca- Lebergott, Stanley. 1964. Manpower in Economic Growth: The tion in China's Western Development." Policy Research Working American Record since 1800. New York: McGraw Hill. Paper 3345, World Bank, Washington, DC. Lederman, Daniel, William F. Maloney, and Luis Servén. 2005. Lutz, Vera. 1962. Italy: A Study in Economic Development. London: Lessons from NAFTA for Latin America and the Caribbean. Oxford University Press. Washington, DC, and Palo Alto, CA: World Bank and Stanford Lynch, Sarah N., and Eugene Mulero. 2007."Williamsburg's Most University Press. Recent Reinvention: The Makeover in Brooklyn." The Coopera- Lee, Annabel. 2008."A History of Korea's Spatial Transformation tor. Nov., 2007. and Economic Growth." Washington, DC. Lysenko, A. J., and I. N. Semashko. 1968."Geography of Malaria: A Lei, Ding, and Kingsley E. Haynes. 2004."The Role of Telecom- Medico-Geographic Profile of an Ancient Disease [in Russian]." munications Infrastructure in Regional Economic Growth of In Medicinskaja Geografija, ed. A. W. Lebedew. Moscow: Russian China." Paper presented at the Telecommunications Policy Academy of Sciences. Research Conference, Washington, DC. Mackay, C. J. 1999."Housing Policy in South Africa: The Challenge Leigland, James, and Andrew Roberts. 2007."The African Project of Delivery." Housing Studies 14 (3): 387­99. Preparation Gap." Note 18. MacMillan, Margaret. 2002. Paris 1919: Six Months That Changed Levinsohn, James, and Amil Petrin. 1999."When Industries the World. New York: Random House. Become More Productive, Do Firms?" NBER Working Paper Maddison, Angus. 2006. The World Economy. Paris: Organisation 6893, National Bureau of Economic Research, Cambridge, MA. for Economic Co-operation and Development. Levinson, Marc. 2006. The Box: How the Shipping Container Made ------. 2008a. The World Economy: Volume 1: A Millennial Per- the World Smaller and the World Economy Bigger. Princeton, NJ: spective. Paris: OECD. Princeton University Press. ------. 2008b."World Population, GDP, and Per Capita GDP, Limão, Nuno, and Anthony J. Venables. 2001."Infrastructure, 1­2003 AD." University of Gröningen Growth and Development Geographical Disadvantage, Transport Costs, and Trade." World Centre, Gröningen, Netherlands. Bank Economic Review 15 (3): 451­79. Malaysia Economic Planning Unit. 2008. Malaysia Achieving the Linden, Greg, Kenneth L. Kraemer, and Jason Dedrick. 2007. MDGs. Kuala Lumpur: Malaysia Economic Planning Unit, "Who Captures Value in a Global Innovation System? The Primer Minister's Office. 320 WORLD DEVELOPMENT REPORT 2009 Malinowski, Matthew. 2007."Cold Snap Prompts Chile to Seek Mayer, Thierry. 2008."Market Potential and Development." Back- Gas Deal with Old Foe Bolivia." Christian Science Monitor, ground paper for the WDR 2009. August 8. McCallum, John. 1995."National Borders Matter: Canada-U.S. Margo, Robert A. 1988."Schooling and the Great Migration." Regional Trade Patterns." American Economic Review 85 (3): NBER Working Paper 2697, National Bureau of Economic 615­23. Research, Cambridge, MA. McCann, Philip. 2001. Urban and Regional Economics. Oxford, ------. 2000. Wages and Labor Markets in the United States, 1820­ U.K.: Oxford University Press. 1860. Chicago: University of Chicago Press. McCrae, Ian. 2006."Uncertainties in Road Transport Emissions in ------. 2004."The North-South Wage Gap, Before and After the Developing Countries." HM Treasury, London. Civil War." In Slavery in the Development of the Americas, ed. McInnis, R. Marvin. 1966."Regional Income Differentials in Can- David Eltis, Frank D. Lewis, and Kenneth L. Sokoloff. New York: ada, 1911­1961." Journal of Economic History 26 (4): 586­88. Cambridge University Press. McKenzie, David J. 2001."The Impact of Capital Controls on Markussen, T. Forthcoming."Property Rights, Productivity, and Growth Convergence." Journal of Economic Development 26 (1): Common Property Resources: Insights from Rural Cambodia." 1­25. World Development. ------. 2007."Paper Walls Are Easier to Tear Down: Passport Marland, Gregg, Tom Boden, and Robert Andres. 2007."National Costs and Legal Barriers to Emigrations." World Development 35 CO2 Emissions from Fossil-Fuel Burning, Cement Manufac- (11): 2026­39. ture, and Gas Flaring: 1751­2004." Carbon Dioxide Information McKenzie, David J., and Hillel Rapoport. 2007."Self-Selection Analysis Center, Oakridge National Laboratory, Oak Ridge, TN. Patterns in Mexico-U.S. Migration: The Role of Migration Marshall, Alfred. 1890. Principles of Economics. London: Mac- Networks." Policy Research Working Paper 4118, World Bank, millan. Washington, DC. Martin, Philippe. 2005."The Geography of Inequalities in Europe." Meade, James E. 1956."The Theory of Customs Union." American Swedish Economic Policy Review 12: 83­108. Economic Review 46 (4): 724­26. Martin, Philippe, Thierry Mayer, and Mathias Thoenig. 2008. Melitz, J. Marc. 2003."The Impact of Trade on Intra-industry "Civil Wars and International Trade." Journal of the European Reallocations and Aggregate Industry Productivity." Economet- Economic Association 6 (2­3): 541­50. rica 71 (6): 1695­725. Martin, Ronald L. 2005."Centralized Versus Decentralized Finan- Mera, Koichi. 1973."Regional Production Function and Social cial Systems: Is There a Case for Local Capital Markets?" Paper Overhead Capital: An Analysis of the Japanese Case." Regional presented at the Cambridge-MIT Institute Conference on Ven- and Urban Economics 3 (2): 157­86. ture Capital, Finance and Regional Development, Cambridge, Metzer, Jacob. 1974."Railroad Development and Market Integra- U.K., June 17. tion: The Case of Tsarist Russia." Journal of Economic History 34 Martinez-Galarraga, Julio. 2007."New Estimates on Regional GDP (3): 529­50. in Spain, 1860­1930." Working Paper 177, Universitat de Barce- Meybeck, Michel, Pamela Green, and Charles Vorosmarty. 2001."A lona Espai de Recerca en Economia, Barcelona. New Typology for Mountains and Other Relief Classes." Moun- Martínez-Zarzoso, Inmaculada, L. García-Menéndez, and C. Suá- tain Research and Development 21 (1): 34­45. rez-Burguet. 2003."Impact of Transport Costs on International Meyer, Balthasar H. 1917. History of Transportation in the United Trade: The Case of Spanish Ceramic Exports." Maritime Eco- States before 1860. Washington, DC: Carnegie Institution of nomics and Logistics 5 (2): 179­98. Washington. Martínez-Zarzoso, Inmaculada, and Laura Marquez-Ramos. 2007. Miguel, Edward, Shanker Satyanath, and Ernest Sergenti. 2004. "The Effect of Trade Facilitation on Sectoral Trade." Working "Economic Shocks and Civil Conflict: An Instrumental Vari- Paper 167, Ibero-America Institute for Economic Research, Göt- ables Approach." Journal of Political Economy 112 (4): 725­53. tingen, Germany. Milanovic, Branko. 2005. Worlds Apart. Measuring International Martins, Joaquim Oliveira, Stefano Scarpetta, and Dirk Pilat. 1996. and Global Inequality. Princeton, NJ: Princeton University Press. "Mark-up Ratios in Manufacturing Industries: Estimates for 14 Miles, Alice. 2007."What's That Whimpering from the Fields?" OECD Countries." Working Paper 162, Organisation for Eco- London Times, July 17. nomic Co-operation and Development, Paris. Mills, Edwin S., and James MacKinnon. 1973."Notes on the New Massey, Douglas S. 1988."International Migration and Economic Urban Economics." Bell Journal of Economics and Management Development in Comparative Perspective." Population and Science 4(2):593­601. Development Review 14 (3): 383­413. Minot, Nicholas, Bob Baulch, and Michael Epprecht. 2003. Pov- ------. 2003."Patterns and Processes of International Migration erty and Inequality in Vietnam: Spatial Patterns and Geographic in the 21st Century." Paper presented at the Conference on Afri- Determinants. Washington, DC: International Food Policy can Migration in Comparative Perspective, Johannesburg. Research Institute and Inter-ministerial Poverty Mapping Task Maurel, Françoise, and Béatrice Sédillot. 1999."A Measure of the Force. Geographic Concentration in French Manufacturing Indus- Mints, A. A. 1974. Prognoznaya Hypoteza Razvitiya Narodnogo tries." Regional Science and Urban Economics 29 (5): 575­604. Khozyastva Evropeyskoy Chasti SSSR [A Forecast Hypothesis of References 321 the Development of the European USSR Economy]. Moscow: Naqvi, Zareen F., and Philip Schuler. 2007. The Challenges and Resursy, Sreda, Rasselenie, Nauka. Potential of Pakistan-India Trade. Washington, DC: World Bank. Mitchell, Brian R. 1964."The Coming of the Railway and the Naudé, Wim A. 2007."Geography and Development in Africa: United Kingdom Economic Growth." Journal of Economic His- Overview and Implications for Regional Cooperation." WIDER tory 24 (3): 315­36. Working Paper Series 2007/03, United Nations, Helsinki. Mohammed, Saif I., and Jeffrey G. Williamson. 2004."Freight Naudé, Wim A., and W. F. Krugell. 2006."Sub-national Growth Rates and Productivity Gains in British Tramp Shipping, 1869­ Rate Differentials in South Africa: An Econometric Analysis." 1950." Explorations in Economic History 41 (2): 172­203. Papers in Regional Science 85 (3): 443­57. Montenegro, Claudio E., and Maximilian L. Hirn. 2008."A New Naudé, Wim A., and Marianne Matthee. 2007."The Significance Disaggregated Set of Labor Market Indicators Using Standard- of Transport Costs in Africa." United Nations University, Tokyo. ized Household Surveys from Around the World." Background Processed. paper for the WDR 2009. Ndulu, Benno J., Lopamudra Chakraborty, Lebohang Lijane, Moomaw, Ronald L. 1981."Productivity and City Size: A Critique Vijaya Ramachandran, and Jerome Wolgin. 2007. Challenges of of the Evidence." Quarterly Journal of Economics 96 (4): 675­88. African Growth: Opportunities, Constraints and Strategic Direc- ------. 1983."Is Population Scale a Worthless Surrogate for Busi- tions. Washington, DC: World Bank. ness Agglomeration Economies?" Regional Science and Urban Negri, Mariano, and Guido Porto. 2007."Burley Tobacco Clubs Economics 13 (4): 525­45. in Malawi: Non-market Institutions for Export." World Bank, Moreno, Ramón, and Bharat Trehan. 1997."Location and the Washington, DC. Growth of Nations." Journal of Economic Growth 2 (4): 399­418. Negroponte, Nicholas. 1995. Being Digital. New York: Vintage Moretti, Enrico. 2004a."Estimating the Social Return to Higher Books. Education: Evidence from Longitudinal and Repeated Cross- Nelson, Andrew. 2008."Accessibility Model and Population Esti- Sectional Data." Journal of Econometrics 121 (1­2): 175­212. mates." Background paper for the WDR 2009. ------. 2004b."Workers' Education, Spillovers, and Productivity: Nesslein, Thomas S. 2003."Markets versus Planning: An Assess- Evidence from Plant-Level Production Functions." American ment of the Swedish Housing Model in the Post-War Period." Economic Review 94 (3): 656­90. Urban Studies 40 (7): 1259­82. Morrisson, Christian, and Fabrice Murtin. 2005."The World Dis- Neumayer, Eric. 2006."Unequal Access to Foreign Spaces: How tribution of Human Capital, Life Expectancy and Income: A States Use Visa Restrictions to Regulate Mobility in a Globalised Multi-dimensional Approach." London School of Economics, World." Paper presented at the Annual Meeting of the Interna- London. tional Studies Association, San Diego, CA. Mukherjee, Andy. 2007."India's Answer to Shenzhen Needs Politi- Neuwirth, Robert. 2007."Security of Tenure in Istanbul: The Tri- cal Will." Bloomberg.com, March 15. umph of the `Self-Service' City." United Nations Habitat. Case Mumford, Lewis. 1963. The Highway and the City. New York: Har- study prepared for Enhancing Urban Safety and Security, Global court Brace and World Bank. Report on Human Settlements 2007. Mutlu, Servet. 1991."Regional Disparities, Industry and Govern- Nigeria National Bureau of Statistics. 2004. National Nigeria Liv- ment Policy in Japan." Development and Change 22 (3): 547­86. ing Standard Survey (NLSS). Abuja: Nigeria National Bureau of N'Dulu, Benno J. 2001."From Vision to Reality of African Eco- Statistics. nomic Integration: Priority Actions and the Institutional ------. 2006. Poverty Profile for Nigeria. Abuja: National Bureau Framework for the Way Forward." World Bank, Washington, of Statistics, Federal Republic of Nigeria. DC. Nikomborirak, Deunden. 2007."Liberalization of Air Transport N'Guessan, Koffi, and Bassirou Chitou. 2006."ENSEA of Abidjan: and Competition Concerns in ASEAN." In Market Access, Trade in the Heart of Statistical Training in Francophone Africa." Afri- in Transport Services and Trade Facilitation, ed. OECD/ECMT. can Statistical Journal 2: 151­56. Paris: Organisation for Economic Co-operation and Develop- Nabi, Ijaz, and Anjum Nasim. 2001."Trading with the Enemy: A ment. Case for Liberalizing Pakistan-India Trade." In Regionalism and Nordas, Hildegunn Kyvik, Enrico Pinali, and Massimo Geloso Globalization: Theory and Practice, ed. Sajal Lahiri. New York: Grosso. 2006."Logistics and Time as a Trade Barrier." Trade Pol- Routledge. icy Working Paper 35, Organisation for Economic Co-operation Naisbitt, John. 1995. The Global Paradox. New York: Avon Books. and Development, Paris. Nakajima, Tomio. 1982. Public Investment: Theory and Practice. Nordhaus, William D. 2006."Geography and Macroeconomics: Tokyo: Gyosei. New Data and New Findings." Proceedings of the National Acad- emy of Science 103 (10): 3510­17. Nakamura, Ryohei. 1985."Agglomeration Economies in Urban Manufacturing Industries: A Case of Japanese Cities." Journal of Nordhaus, William D., and Joseph Boyer. 2000. Warming the Urban Economics 17 (1): 108­24. World: Economic Models of Global Warming. Cambridge, MA: MIT Press. Nam, Charles B., William J. Serow, and David F. Sly. 1990. Interna- tional Handbook on Internal Migration. New York: Greenwood North, Douglass C. 1971."Institutional Change and Economic Press. Growth." Journal of Economic History 31 (1): 118­25. 322 WORLD DEVELOPMENT REPORT 2009 O'Gorman, Marvin. 1939a."Square Deals for London Traffic: A Paci, Pierella, Erwin R. Tiongson, Mateusz Walewski, Jacek Liwin- Study of London's Traffic Problem with Suggested Solutions. ski, and Maria M. Stoilkova. 2007."Internal Labor Mobility in Part 1." Highways and Bridges, March 29. Central Europe and the Baltic Region." Policy Research Working ------. 1939b."Square Deals for London Traffic: A Study of Lon- Paper 105, World Bank, Washington, DC. don's Traffic Problem with Suggested Solutions. Part 2." High- Packard, Truman. 1997."Adjustment, Migration and Economic ways and Bridges, April 5. Informality in Latin America." M.S. Economic for Development O'Rourke, Kevin H., and Jeffrey Williamson. 1994."Late Nine- thesis, Queen Elizabeth House, University of Oxford. teenth-Century Anglo-American Factor-Price Convergence: Paillacar, Rodrigo. 2007."Market Potential and Worker Heteroge- Were Heckscher and Ohlin Right?" Journal of Economic History neity as Determinants of Brazilian Wages." Université de Paris I 54 (4): 892­916. Panthéon-Sorbonne. Paris. Obstfeld, Maurice, Jay C. Shambaugh, and Alan M. Taylor. 2004. Pakenham, Thomas. 1992. The Scramble for Africa. New York: "Monetary Sovereignty, Exchange Rates, and Capital Controls: Harper Collins. The Trilemma in the Interwar Period." NBER Working Paper Palacin, Jose, and Robert C. Shelburne. 2005."The Private Housing 10393, National Bureau of Economic Research, Cambridge, MA. Market in Eastern Europe and the CIS." Working Paper 2005-5, Obstfeld, Maurice, and Alan M. Taylor. 2002."Globalization and United Nations Commission for Europe, Geneva. Capital Markets." NBER Working Paper 8846, National Bureau Pallage, Stéphane, and Michel A. Robe. 2003."On the Welfare Cost of Economic Research, Cambridge, MA. of Economic Fluctuation in Developing Countries." Interna- Offer, A. 1981. Property and Politics, 1870­1914. Cambridge, U.K.: tional Economic Review 44 (2): 677­98. Cambridge University Press. Park, Sam Ock. Forthcoming."A History of Korea's Spatial Trans- Office of the Deputy Prime Minister. 2003. Cities, Regions, and formation and Economic Growth." In Reshaping Economic Competitiveness. London: Office of the Deputy Prime Minister. Geography in East Asia, ed. Yukon Huang and Alessandro Mag- Oh, Jinhwan. 2008."Korea: A Case Study of 1D-2D-3D Areas." noli Bocchi. Washington, DC: World Bank, EAP Companion Background note for the WDR 2009. Volume to the WDR 2009. ------. 2008."Spatially Blind and Connective Policies in a 1D Parsons, Christopher R., Ronald Skeldon, Terrie L. Walmsley, and Region: The Case of Eumseong County in South Korea." Back- L. Alan Winters. 2007."Quantifying International Migration: A ground note for the WDR 2009. Database of Bilateral Migrant Stocks." In International Migra- Ohmae, Kenichi. 1990. The Borderless World: Power and Strategies tion, Economic Development and Policy, ed. Caglar Özden and in the Interlinked Economy. New York: Harper Business. Maurice Schiff. Washington, DC: World Bank. Okuma, Ichiro, ed. 1980. Readings in Public Finance. Tokyo: Toyo Paul, Catherine J. Morrison, and Donald S. Siegel. 1999."Scale Kenzai Shinposya. Economies and Industry Agglomeration Externalities: A Dynamic Cost Function Approach." American Economic Review Openshaw, Stan, and Paul Taylor. 1979."A Million or So Correla- 89 (1): 272­90. tion Coefficients: Three Experiments on the Modifiable Areal Unit Problem." In Statistical Applications in the Spatial Sciences, Pedersen, P. O. 1990."The Role of Small Rural Towns in Develop- ed. Neil Wrigley. London: Pion. ment." In Small Town Africa, Seminar Proceedings No. 23, ed. Jonathan Baker. Uppsalla: Scandinavian Institute of African Oppenheimer, Andrés. 2008."Latin America Is Lagging. Someone Studies. Tells Its Leaders." Washington Post, January 13. Peng, Zhong-Ren, Yi Zhu, and Shunfeng Song. 2008."Mobility of Organisation for Economic Co-operation and Development the Chinese Urban Poor: A Case Study of Hefei City." Chinese (OECD). 2003. Place-Based Policies for Rural Development: The Economy 41 (1): 36­57. Micro-Regions Strategy, Mexico (Case Study). Background paper for the OECD's Horizontal Review of Place-Based Policies for Phillips, Truman P., Daphne S. Taylor, Lateef Sanni, and Malachy Development, Paris. O. Akoroda. 2004. The Global Cassava Development Strat- egy. Rome: Food and Agriculture Organization of the United ------. 2006. Territorial Reviews: Competitive Cities in the Global Nations. Economy. Paris: Organisation for Economic Co-operation and Development. Policy Research Institute for Land, Infrastructure Transport, and Tourism Research. 2001. Discussion on Equity and Social Capital ------. 2007. OECD: Regions at a Glance. Paris: Organisation for Improvement. Tokyo: Policy Research Institute for Land, Infra- Economic Co-operation and Development. structure, Transport, and Tourism. Orwell, George. 1937. The Road to Wigan Pier. London: Victor Poncet, Sandra. 2005."A Fragmented China: Measures and Deter- Gollancz. minants of Chinese Domestic Market Integration." Review of Overseas Development Institute. 2006. What's Next in Interna- International Economics 13 (3): 409­30. tional Development? London: Overseas Development Institute. ------. 2006."Provincial Migration Dynamics in China: Borders, Owen, Nicholas. 1983. Economies of Scale, Competitiveness, and Costs, and Economic Motivations." Regional Science and Urban Trade Patterns within the European Community. Oxford, U.K.: Economics 36 (3): 385­98. Clarendon Press. Prateen, Cliff. 1988. "Survey of the Economies of Scale." Eco- Özden, Caglar, and Maurice Schiff. 2007. International Migration, nomic Papers 67, Commission of the European Communities, Economic Development and Policy. Washington, DC: World Bank. Brussels. References 323 PricewaterhouseCoopers. 2007. Which Are the Largest City Econo- Revenue Mobilization Allocation and Fiscal Commission. 2003. mies in the World and How Might This Change by 2020? London States and Local Governments 2003 Revenue Allocation Indices. PricewaterhouseCoopers: UK Economic Outlook. Abuja, Nigeria: Revenue Mobilization Allocation and Fiscal Pritchett, Lant. 2006. Let Their People Come: Breaking the Gridlock Commission. on Global Labor Mobility. Washington, DC: Center for Global Richter, K., R. van der Weide, and Phonesaly Souksavath. 2005. Development. Lao PDR Poverty Trends 1992/93­2002/03. Washington, DC: Public Administration Research and Consultation Centre. 2004. National Statistics Centre and World Bank. Egypt Millennium Development Goals: Second Country Report. Roberts, Leslie, and Martin Enserink. 2007."María: Did They Cairo: United Nations and Egypt Ministry of Planning. Really Say ... Eradication?" Science 318 (5856): 1544­45. Pucher, John, Nisha Korattyswaropam, Neha Mittal, and Neenu Roberts, Mark. 2004."The Growth Performances of the GB Coun- Ittyerah. 2005."Urban Transport Crisis in India." Transport Pol- ties: Some New Empirical Evidence for 1977­1993." Regional icy 12 (3): 185­98. Studies 38 (2): 149­65. Puga, Diego. 2002."European Regional Policies in Light of Recent Roberts, Mark, and Uwe Deichmann. 2008."Regional Spillover Location Theories." Journal of Economic Geography 2 (4): 373­ Estimation." Background paper for the WDR 2009. 406. Robles, Marcos. 2003. Estimación de Indicadores de pobreza y Des- Puga, Diego, and Anthony J. Venables. 1999."Agglomeration and igualdad a Nivel Municipal en Honduras. Tegucigalpa: BID-ME- Economic Development: Import Substitution vs. Trade Liberal- COVI and Instituto Nacional de Estadísticas de Honduras. ization." Economic Journal 109 (455): 292­311. Rodríguez, Francisco, and Dani Rodrik. 2000."Trade Policy and Quah, Danny T. 1996."Twin Peaks: Growth and Convergence in Economic Growth: A Skeptic's Guide to the Cross-National Evi- Models of Distribution Dynamics." Economic Journal 106 (437): dence." In NBER Macroeconomics Annual 2000, ed. Ben S. Ber- 1045­55. nanke and Kenneth Rogoff. Cambridge, MA: National Bureau Quinn, Dennis P., and A. María Toyoda. 2006."Does Capital of Economic Research. Account Liberalization Lead to Growth?" Georgetown Univer- Rodríguez, Jorge. 2004."Migración Interna en América Latina y el sity, Washington, DC. Caribe: Estudio Regional del Período 1980­2000." Serie Pobla- Rajan, Raghuram, and Shang-Jin Wei. 2004."The Non-Threat ción y Desarrollo 50, ECLAC, Santiago. That Is Outsourcing." Business Times (Singapore). Rodríguez-Pose, Andrés, and Ugo Fratesi. 2004."Between Devel- Ratha, Dilip, and William Shaw. 2007."South-South Migration opment and Social Policies: The Impact of European Structural and Remittances." Working Paper 102, World Bank, Washington, Funds in Objective 1 Regions." Regional Studies 38 (1): 97­113. DC. Rodrik, Dani, Arvind Subramanian, and Francesco Trebbi. 2004. Ratha, Dilip, and Zhimei Xu. 2008. Migration and Remittances "Institutions Rule: The Primacy of Institutions over Geography Factbook 2008. Washington, DC: World Bank. and Integration in Economic Development." Journal of Eco- Rauch, James E. 1993."Productivity Gains from Geographic Con- nomic Growth 9 (2): 131­65. centration of Human Capital: Evidence from the Cities." Journal Roeger, Werner. 1995."Can Imperfect Competition Explain the of Urban Economics 34 (3): 380­400. Difference between Primal and Dual Productivity Measures? ------. 1999."Networks versus Markets in International Trade." Estimates for U.S. Manufacturing." Journal of Political Economy Journal of International Economics 48 (1): 7­35. 103 (2): 316­30. Ravallion, Martin. 2007."Inequality Is Bad for the Poor." In Rohwedder, Cecilie, and Keith Johnson. 2008."Pace-Setting Zara Inequality and Poverty Re-Examined, ed. Steven Jenkins and Seeks More Speed to Fight Its Rising Cheap-Chic Rivals." Wall John Micklewright. New York: Oxford University Press. Street Journal, February 20. Ravallion, Martin, and Dominique van de Walle. 2006a."Does Ris- Rolfes, L. Jr. 2002."Making the Legal Basis for Private Land Rights ing Landlessness Signal Success or Failure for Vietnam's Agricul- Operational and Effective." Paper presented at the World Bank tural Transition?" Policy Research Working Paper 3871, World Regional Land Policy Workshop, Budapest, April 6. Bank, Washington, DC. Romer, Paul M. 1986."Increasing Returns in Long-Run Growth." ------. 2006b."Land Reallocation in an Agrarian Transition." Journal of Political Economy 94 (5): 1002­37. Economic Journal 116 (514): 924­42. ------. 1990."Endogenous Technological Change." Journal of Redding, Stephen J., and Anthony J. Venables. 2004."Economic Political Economy 98(5):71­102. Geography and International Inequality." Journal of Interna- ------. 1994."The Origins of Endogenous Growth." Journal of tional Economics 62 (1): 53­82. Economic Perspectives 8 (1): 3­22. Rephann, Terance, and Andrew Isserman. 1994."New Highways Rose, Andrew K. 2005."Cities and Countries." Discussion Paper as Economic Development Tools: An Evaluation Using Quasi- 5235, Centre for Economic Policy Research, London. Experimental Matching Methods." Regional Science and Urban Rosenbloom, Joshua L., and William A. Sundstrom. 2003. "The Economics 24 (6): 723­51. Decline and Rise of Interstate Migration in the United States: Reporters without Borders for Press Freedom. 2007. Freedom of the Evidence from the IPUMS, 1850­1990." NBER Working Press Worldwide in 2007. Paris: Reporters without Borders for Paper 9857, National Bureau of Economic Research, Cam- Press Freedom. bridge, MA. 324 WORLD DEVELOPMENT REPORT 2009 Rosenthal, Stuart S., and William C. Strange. 2001."The Deter- Middle Income Nations. London: International Institute for minants of Agglomeration." Journal of Urban Economics 50 (2): Environment and Development. 191­229. Schacter, Jason P. 2004."Geographical Mobility: 2002 to 2003." ------. 2003."Geography, Industrial Organization, and Agglom- Current Population Report, U.S. Census Bureau, Washington, eration." Review of Economics and Statistics 85 (2): 377­93. DC. March. ------. 2004."Evidence on the Nature and Sources of Agglomera- Scherer, Frederic M. 1980. Industrial Market Structure and Eco- tion Economies." In Handbook of Regional and Urban Economics, nomic Performance. Boston: Houghton Mifflin. vol. IV, ed. J. Vernon Henderson and Jacque Thisse. Amsterdam: Schiff, Maurice, and Caglar Özden. 2006. International Migration, North-Holland. Remittances and the Brain Drain. Washington, DC: World Bank Rossi-Hansberg, Esteban, and Mark L. J. Wright. 2007."Urban and Palgrave Macmillan. Structure and Growth." Review of Economic Studies 74(2):597­ Schiff, Maurice, and Yanling Wang. 2003."Regional Integration 624. and Technology Diffusion: The Case of North American Free Rostow, Walt Whitman. 1960. The Stages of Economic Growth. Trade Agreement." Policy Research Working Paper 3132, World Cambridge, U.K.: Cambridge University Press. Bank, Washington, DC. Roy, Tirthankar. 2002."Economic History and Modern India: Schiff, Maurice, and L. Alain Winters. 2002."Regional Coopera- Redefining the Link." Journal of Economic Perspectives 16 (3): tion and the Role of International Organizations and Regional 109­30. Integration." Policy Research Working Paper 2872, World Bank, Ruster, Jeff, and Ivo Imperato. 2003. Slum Upgrading and Participa- Washington, DC. tion: Lessons from Latin America. Washington, DC: World Bank. ------. 2003. Regional Integration and Development. New York: Sachs, Jeffrey D., and Pia Malaney. 2002."The Economic and Oxford University Press. Social Burden of Malaria." Nature 415 (6872): 680­5. Schmertmann, Carl P. 1992."Estimation of Historical Migration Sahn, David E., and David C. Stifel. 2003."Urban-Rural Inequality Rates from a Single Census: Interregional Migration in Brazil, in Living Standards in Africa." Journal of African Economies 12 1900­1980." Population Studies 46 (1): 103­20. (4): 564­97. Schneider, A., M. A. Friedl, and D. Potere. Forthcoming."A New Sakamaki, Tetsuro. 2006."Regional Disparities and National Land Map of Global Urban Extent from MODIS Data." Geophysical Development Policy in East Asian Countries." Journal of Japan Research Letters. Bank for International Cooperation Institute 29: 84­122. Scholz, Stephan. 2006."The POETICS of Industrial Carbon Diox- Sánchez, Ricardo J., Jan Hoffmann, Alejandro Micco, Georgina V. ide Emissions in Japan: An Urban and Institutional Extension Pizzolitto, Martín Sgut, and Gordon Wilmsmeier. 2003."Port of the IPAT Identity." Carbon Balance and Management 1 (11): Efficiency and International Trade: Port Efficiency as a Deter- 1­10. minant of Maritime Transport Costs." Maritime Economics and Schwenning, Gustav. 1927."An Attack on Shanghai Slums." Social Logistics 5 (2): 199­218. Forces 6 (1): 125­31. Sandler, Todd. 2002."Demand and Institutions for Regional Pub- Scott, Ian. 1982. Urban and Spatial Development in Mexico. Balti- lic Goods." In Regional Public Goods: From Theory to Practice, more, MD: Johns Hopkins University. ed. Antoni Estevadeordal, Brian Frantz, and Tam R. Nguyen. Seeman, Helene Zuckerman, and Alanna Siegfried. 1978."The Washington, DC: Inter-American Development Bank and Asian Evolution of SoHo." In Helene Zuckerman Seeman and Alanna Development Bank. Siegfried, (eds.), SoHo: A Guide. New York, NY: Neal-Schuman Sandler, Todd, and Daniel M. Arce. 2002."A Conceptual Frame- Publishers, Inc. work for Understanding Global and Transnational Public Goods Segal, David. 1976."Are There Returns to Scale in City Size?" for Health." Fiscal Studies 23 (2): 195­222. Review of Economics and Statistics 58 (3): 339­50. Santos, Georgina, and Blake Shaffer. 2004."Preliminary Results of Selod, Harris. 2007."Land Policies and Urban Segregation." World the London Congestion Charging Scheme." Public Work Man- Bank, Washington, DC. agement and Policy 9 (2): 164­81. Shah, Anwar, and Chunli Shen. 2006."The Reform of the Intergov- Sassen, Saskia. 1991. The Global City: New York, London, Tokyo. ernmental Transfer System to Achieve a Harmonious Society and Princeton, NJ: Princeton University Press. a Level Playing Field for Regional Development in China." Policy Satterthwaite, David. 2007."The Transition to a Predominantly Research Working Paper 4100, World Bank, Washington, DC. Urban World and Its Underpinnings." Paper presented at the Shatz, Howard J., and Anthony J. Venables. 2005."The Geography UNU-WIDER International Workshop on Beyond the Tipping of International Investment." In Oxford Handbook of Economic Point: Development in an Urban World, London, October 19. Geography, ed. Gordon L. Clark, Meric S. Gertler, and Maryann ------. 2008."Expanding the Supply and Reducing the Cost of P. Feldman. Oxford, U.K.: Oxford University Press. Land for Housing in Urban Areas in Low and Middle Income Shaw, Dan. 2007."Meatpacking District: What a Difference 38 Nations." Background note for the WDR 2009. Years Make." The New York Times, October 14. Satterthwaite, David, Saleemul Hug, Mark Pelling, Hannah Reid, Shefer, Daniel. 1973."Localization Economies in SMSA's: A Pro- and Patricia Romero-Lankao. 2007. Adapting to Climate Change duction Function Analysis." Journal of Regional Science 13 (1): in Urban Areas: The Possibilities and Constraints in Low and 55­64. References 325 Shepherd, Ben, and John S. Wilson. 2006."Road Infrastructure in ------. 2000b."Institutions, Factor Endowments, and Paths of Europe and Central Asia: Does Network Quality Affects Trade?" Development in the New World." Journal of Economic Perspec- Policy Research Working Paper 4104, World Bank, Washington, tives 14 (3): 217­32. DC. Solow, Robert M. 1956."A Contribution to the Theory of Eco- Showers, Victor. 1979. World Facts and Figures. Chichester, U.K.: nomic Growth." Quarterly Journal of Economics 70 (1): 65­94. John Wiley and Sons. Soltow, Lee. 1989."The Rich and the Destitute in Sweden, 1805­ Shukla, Vibhooti, and Oded Stark. 1986."Urban External Econo- 1855: A Test of Tocqueville's Inequality Hypotheses." Economic mies and Optimal Migration." In Migration, Human Capital, History Review 42 (1): 43­63. and Development, ed. Oded Stark. Greenwich, CT: JAI Press. Son, Dang Kim. Forthcoming."Rural Development and Issues in Sicsic, Pierre. 1992."City-Farm Wage Gaps in Late Nineteenth- Vietnam: Spatial Disparities and Some Recommendations." In Century France." Journal of Economic History 52 (3): 675­95. Reshaping Economic Geography in East Asia, ed. Yukon Huang Siegel, Fred. 1997. The Future Once Happened Here: New York, and Alessandro Magnoli Bocchi. Washington, DC: World Bank, D.C., L.A., and the Fate of America's Big Cities. Florence, MA: EAP Companion Volume to the WDR 2009. Free Press. Sonobe, Tetsushi, and Keijiro Otsuka. 2006. Cluster-Based Indus- Simmons, Beth A. 2005."Rules over Real Estate: Trade, Territorial trial Development: An East Asian Model. Hampshire, U.K.: Pal- Conflict, and International Borders as Institutions." Journal of grave Macmillan. Conflict Resolution 49 (6): 823­48. Sophal, Chan, Tep Saravi, and Sarthi Acharya. 2001."Land Tenure Sinai, Todd, and Joel Waldfogel. 2004."Geography and the Inter- in Cambodia: A Data Update." Working Paper 19, Development net: Is the Internet a Substitute or Complement for Cities?" Resource Institute, Phnom Penh. Journal of Urban Economics 56 (1): 1­24. Soto, Raimundo, and Arístides Torche. 2004."Spatial Inequal- Sinn, Hans-Werner, and Frank Westermann. 2001."Two Mezzo- ity, Migration, and Economic Growth in Chile." Cuadernos de giornos." NBER Working Paper 8125, National Bureau of Eco- Economía 41 (124): 401­24. nomic Research, Cambridge, MA. Sovannarith, So, Real Sopheap, Uch Utey, Sy Rathmony, Brett Bal- Skeldon, Ronald. 1986."On Migration Patterns in India during the lard, and Sarthi Acharya. 2001."Social Assessment of Land in 1970s." Population and Development Review 12 (4): 759­79. Cambodia: A Field Study." Prepared for the Ministry of Land Slaughter, Mattew J. 2001."Trade Liberalization and Per Capita Management, Urban Planning, and Construction, Royal Gov- Income Convergence: A Difference-in-Differences Analysis." ernment of Cambodia, Cambodia Development Resource Insti- Journal of International Economics 55 (1): 203­28. tute, Phnom Penh. Smith, Adam. 1976 [1776]. An Inquiry into the Nature and Causes SPARTECA. 1996. Rules of Origin Requirements of the South Pacific of the Wealth of Nations. Chicago: University of Chicago Press Trade and Economic Co-operation Agreement. Suva: South Pacific (Cannan's edition of the Wealth of Nations was originally pub- Regional Trade and Economic Co-operation Agreement, Forum lished in 1904 by Methuen & Co. Ltd. First Edition in 1776). Secretariat. Smith, Donald F. Jr., and Richard Florida. 1994."Agglomeration Spence, A. Michael. 1976."Product Selection, Fixed Costs, and and Industrial Location: An Econometric Analysis of Japanese- Monopolistic Competition." Review of Economic Studies Affiliated Manufacturing Establishments in Automotive-Related 43(2):217­35. Industries." Journal of Urban Economics 36 (1): 23­41. Spolaore, Enrico, and Romain Wacziarg. 2005."Borders and Smith, P. J. 1989."The Rehousing/Relocation Issue in an Early Growth." Journal of Economic Growth 10 (4): 331­86. Slum Clearance Scheme: Edinburgh 1865­1885." Urban Studies Squire, Lyn. 1981. Employment Policy in Developing Countries. 26 (1): 100­14. Oxford, U.K.: Oxford University Press. Snow, Robert W., Marlies Craig, Uwe Deichmann, and Karen Sri Lanka National Council for Economic Development, and Marsh. 1999."Estimating Mortality, Morbidity, and Disability UNDP. 2005. Millennium Development Goals Country Report Due to Malaria among Africa's Non-Pregnant Population." Bul- 2005: Sri Lanka. Colombo: Sri Lanka National Council for Eco- letin of the World Health Organization 77 (8): 624­40. nomic Development. Sobel, Robert. 2000. How a Generation of Americans Created the Stewart, Dona J. 1996."Cities in the Desert: The Egyptian New World's Most Prosperous Society. New York: St. Martin's Press. Town-Program." Annals of the Association of American Geogra- Söderbaum, Fredrik. 2001."Institutional Aspects of the Maputo phers 86 (3): 459­80. Development Corridor." Development Policy Research Unit Stewart, Frances. 2008."Horizontal Inequalities: A Neglected Working Paper 9675, University of Cape Town, Cape Town, Dimension of Development." Queen Elizabeth House Working South Africa. Paper 81, Oxford University, Oxford, U.K. Sohn, Jeongmok. 2003. Seoul Dosi Hoek Iyagi [Seoul: The Story of Stinnett, Douglas M., Jaroslav Tir, Philip Schafer, Paul F. Diehl, and Urban Planning]. Seoul: Hanul. Charles Gochman. 2002."The Correlates of War Project Direct Sokoloff, Kenneth L., and Stanley L. Engerman. 2000a."History Contiguity Data, Version 3." Conflict Management and Peace Sci- Lessons: Institutions, Factor Endowments, and Paths of Devel- ence 19 (2): 58­66. opment in the New World." Journal of Economic Perspectives 14 Stockel, Sigrid. 2002."Infant Mortality and Concepts of Hygiene: (3): 217­32. Strategies and Consequences in the Kaiserreich and the Weimar 326 WORLD DEVELOPMENT REPORT 2009 Republic--The Example of Berlin." History of the Family 7 (4): Tol, Richard S. J. 2005."The Marginal Damage Costs of Carbon 601­16. Dioxide Emissions: An Assessment of the Uncertainties." Energy Storper, Michael, and Anthony J. Venables. 2007."Buzz: Face- Policy 33 (16): 2064­74. to-Face Contact and the Urban Economy." In Clusters, Net- Tolofari, Sonny R. 1986. Open Registry Shipping: A Comparative works and Innovation, ed. Stefano Breschi and Franco Malerba. Study of Costs Freight Rates. New York: Routledge. Oxford, U.K.: Oxford University Press. Trackman, B., W. Fisher, and L. Salas. 1999. The Reform of Property Sveikauskas, Leo A. 1975."The Productivity of Cities." Quarterly Registration Systems in Costa Rica: A Status Report. Cambridge, Journal of Economics 89 (3): 393­413. MA: Law School Harvard University. Sveikauskas, Leo A., John Gowdy, and Michael Funk. 1988."Urban Tschirley, David L., Jan J. Nijhoff, Pedro Arlindo, Billy Mwinga, Productivity: City Size or Industry Size." Journal of Regional Sci- Michael T. Weber, and Thomas S. Jayne. 2006."Anticipating and ence 28 (2): 185­202. Responding to Drought Emergencies in Southern Africa: Les- Sveikauskas, Leo A., Peter M. Townroe, and Eric R. Hansen. 1985. sons from the 2002­2003 Experience." International Develop- "Intraregional Productivity Difference in Sao Paulo State Manu- ment Working Paper 89, East Lansing, MI. facturing Plants." Weltwirtschaftliches Archiv 121(4):722­40. Turner, Michael. 1986."English Open Fields and Enclosures: Swan, Trevor W. 1956."Economic Growth and Capital Accumula- Retardation or Productivity Improvements." Journal of Economic tion." Economic Record 32 (2): 334­61. History 46 (3): 669­92. Swedish Council for Building Research. 1990. Housing Research Uchida, Hirotsugu, and Andrew Nelson. 2008.""Agglomeration and Design in Sweden. Stockholm: Swedish Council for Building Index: Towards a New Measure of Urban Concentration." Back- Research. ground paper for the WDR 2009. Tabuchi, Takatoshi. 1988."Interregional Income Differentials and United Nations. 1949. Demographic Yearbook 1948. New York: Migration: Their Interrelationships."Regional Studies 22 (1): 1­10. United Nations. Taira, Koji. 1969."Urban Poverty, Ragpickers, and the`Ants'Villa'in ------. 1952. Demographic Yearbook 1952. New York: United Tokyo."Economic Development and Cultural Change 17 (2): 155­77. Nations. Taylor, Alan M. 1996."International Capital Mobility in History: ------. 1969. Growth of the World's Urban and Rural Population, Purchasing-Power Parity in the Long Run." NBER Working Paper 1920­2000. New York: United Nations, Population Division. 5742, National Bureau of Economic Research, Cambridge, MA. ------. 1970."Unitization of Cargo." Paper presented at the Taylor, Alan M., and Jeffrey Williamson. 2006."Convergence in the United Nations Conference on Trade and Development, Age of Mass Migration." European Review of Economic History 1 Geneva. (1): 27­63. ------. 1997. China Human Development Report 1997: Human Thanh, Hoang Xuan, Dang Nguyen Anh, and Cecilia Tacoli. 2005. Development and Poverty Reduction. New York: United Nations "Livelihood Diversification and Rural-Urban Linkages in Viet- Development Program. nam's Red River Delta." IFPRI Working Paper 193, International ------. 2001."Istanbul+5: Reviewing and Appraising Progress Five Food Policy Research Institute, Washington, DC. Years after Habitat II in June 2001." United Nations, New York. Tiffen, Mary. 2003."Transitions in Sub-Saharan Africa: Agricul- ------. 2004a. Assessing Regional Integration in Africa. Addis ture, Urbanization, and Income Growth." World Development 31 Ababa: United Nations Economic Commission for Africa. (8): 1343­66. ------. 2004b. State of the World's Cities Report 2004/2005. Nai- Tilly, Richard H. 1990. Vom Zollverein zum Industriestaat: Die robi: United Nations. wirtschaftlich-soziale Entwicklung Deutschlands 1834 bis 1914. ------. 2005a. Bringing Down Barriers: Regional Cooperation for Munich, Germany: Deutscher Taschenbuch Verlag. Human Development and Human Security. Bratislava: United Timmer, Ashley S., and Jeffrey G. Williamson. 1998."Immigra- Nations Development Programme. tion Policy prior to the 1930s: Labor Markets, Policy Interac- ------. 2005b. China Human Development Report 2005: Devel- tions, and Globalization Backlash." Population and Development opment with Equity. New York: United Nations Development Review 24 (4): 739­71. Program. Timmins, Christopher. 2005."Estimable Equilibrium Models of ------. 2006a. State of the Worlds Cities 2006/7: The Millennium Locational Sorting and Their Role in Development Economics." Development Goals and Urban Sustainability: 30 Years of Shaping Journal of Economic Geography 5 (1): 83­100. the Habitat Agenda (UN Habitat). Nairobi: UN-Habitat. Tinbergen, J. 1952. On the Theory of Economic Policy. Amsterdam: ------. 2006b. The 2004 Vietnam Migration Survey: The Quality of North-Holland. Life of Migrants in Vietnam. Hanoi: United Nations Population Tobler, W. R. 1970."A Computer Movie Simulating Urban Growth Fund. in the Detroit Region." Economic Geography 46 (Supplement): ------. 2006c."World Urbanization Prospects: The 2005 Revi- 234­40. sion." United Nations, Department of Economic and Social Todaro, Michael P. 1969."A Model of Labor Migration and Urban Affairs, Population Division, New York. Unemployment in Less Developed Countries." American Eco- ------. 2007."Trade and Development Report 2007: Regional nomic Review 59 (1): 138­48. Cooperation for Development." Paper presented at the United Toffler, Alvin. 1980. The Third Wave. New York: Bantam Books. Nations Conference on Trade and Development, New York. References 327 United Nations and African Union. 2006. Assessing Regional Inte- Weber, Adna. 1899. The Growth of Cities in the Nineteenth Century: gration in Africa II: Rationalizing Regional Economic Communi- A Study in Statistics. New York: Macmillan. ties. Addis Ababa: United Nations Commission for Africa and Whalley, John, and Shunming Zhang. 2007."A Numerical Simu- African Union. lation Analysis of (Hukou) Labour Mobility Restrictions in U.S. Bureau of Transport Statistics. 2003. America on the Go: U.S. China." Journal of Development Economics 83 (2): 392­410. Holiday Travel. Washington, DC: U.S. Department of Transpor- Wheat, Leonard F. 1986."The Determinants of 1963­77 Regional tation, Research, and Innovative Technology Administration. Manufacturing Growth: Why the South and West Grow." Journal U.S. Census Bureau. 2002. Demographic Trends in the 20th Century. of Regional Science 26 (4): 635­59. Census 2000 Special Reports.Washington, DC: U.S. Census Bureau. Wheeler, David, and Ashoka Mody. 1992."International Invest- Uvalic, Milica. 1993."The Disintegration of Yugoslavia: Its Costs ment Location Decisions: The Case for U.S. Firms." Journal of and Benefits." Post-Communist Economies 5 (3): 273­93. International Economics 33 (1­2): 57­76. Van der Linde, Claas. 2003."The Demography of Clusters: Find- Whittaker, D. Hugh. 1997. Small Firms in the Japanese Economy. ings from the Cluster Meta-Study." In Innovation Clusters and Cambridge, U.K.: Cambridge University Press. Interregional Competition, ed. Johannes Brocker, Dirk Dohse, Wilbur Smith Associates and United Nations. 1980. Master Plan for and Rudiger Soltwedel. Berlin: Springer. Metropolitan Lagos. New York: United Nations. Vance, Colin, and Ralf Hedel. 2006."On the Link between Urban Williams, Naomi, and Chris Galley. 1995."Urban-Rural Differen- Form and Automobile Use: Evidence from German Survey tials in Infant Mortality in Victorian England." Population Stud- Data." Working Paper 0048, Rheinisch-Westfälisches Institut für ies 49 (3): 401­20. Wirtschaftsforschung, Essen, Germany. Williamson, Jeffrey. 1965."Regional Inequality and the Process of Vargas-Hidalgo, Rafael. 1979."The Crisis for the Andean Pact: Les- National Development: A Description of the Patterns." Economic sons for Integration among Developing Countries." Journal of Development and Cultural Change 13 (4): 1­84. Common Market Studies 17 (3): 213­26. ------. 1974. Late Nineteenth Century American Development: A Venables, Anthony J. 2001."Geography and International Inequali- General Equilibrium History. New York: Cambridge University ties: the Impact of New Technologies." Journal of Industry, Com- Press. petition and Trade 1 (2): 135­59. ------. 1982."Was the Industrial Revolution Worth It? Disameni- ------. 2003."Winners and Losers from Regional Integration ties and Death in 19th-Century British Towns." Exploration in Agreements." Economic Journal 113 (490): 747­61. Economic History 19 (3): 221­45. ------. 2006."Shifts in Economic Geography and Their Causes." ------. 1987."Did English Factor Markets Fail during the Indus- Economic Review 2006 (Q IV): 61­85. trial Revolution?" Oxford Economic Papers 39 (4): 641­78. Venables, Anthony J., and Ravi Kanbur. 2005. Spatial Inequality ------. 1990. Coping with City Growth during the British Indus- and Development, Overview of the UNU-WIDER Project. New trial Revolution. Cambridge, U.K.: Cambridge University York: Oxford University Press. Press. Victoria Transport Policy Institute. 2007. Online TDM Encyclope- Wilson, John S., Xubei Luo, and Harry J. Broadman. 2006."Trade dia. Victoria, Canada: Victoria Transport Policy Institute. and Transport Facilitation: European Accession and Capacity Vidler, Cam. 2008."Turkey: A Case Study of 1D-2D-3D Areas." Building Priorities." Paper presented at the Transport and Inter- Background note for the WDR 2009. national Trade, Report of the 130th Round Table on Transport Vigdor, Jacob L. 2006."The New Promised Land: Black-White Economics, Paris. Convergence in the American South, 1960­2000." NBER Work- Wilson, John S., Catherine L. Mann, and Tsunehiro Otsuki. 2004. ing Paper 12143, National Bureau of Economic Research, Cam- "Assessing the Potential Benefit of Trade Facilitation: A Global bridge, MA. Perspective." Policy Research Working Paper 3224, World Bank, Villarreal, Roberto. 2005."Regional Development Policies in Washington, DC. Mexico." Paper presented at the Sustainable Regional Policy and Winston, Clifford. 1993."Economic Deregulation: Days of Reck- Planning Roundtable, Beijing, December 14. oning for the Microeconomists." Journal of Economic Literature Viner, Jacob. 1950. The Customs Union Issue. New York: Carnegie 31 (3): 1263­89. Endowment of International Peace. Winters, L. Alain, and Pedro Martins. 2004."When Compara- von Hoffman, A. 1996."High Ambitions: The Past and Future of tive Advantage Is Not Enough: Business Costs in Small Remote American Low-Income Housing Policy." Housing Policy Debate Economies." World Trade Review 3 (3): 347­83. 7 (3): 423­46. Wionczek, Miguel S. 1970."The Rise and Decline of Latin Ameri- Wandschneider, Tiago, and Pravas Mishra. 2003. Rural Non- can Economic Integration." Journal of Common Market Studies Farm Economy and Livelihood Enhancement. Washington, DC: 9 (1): 49­66. National Resources Institute. Wisaweisuan, Nitinant. Forthcoming."Spatial Disparities in Thai- Ward, Doug. 2006."Urbanization Is a World Concern." Vancouver land: Does Government Policy Aggravate or Alleviate the Prob- Sun, June 14. lem?" In Reshaping Economic Geography in East Asia, ed. Yukon Warsh, David. 2006. Knowledge and the Wealth of Nations: A Story Huang and Alessandro Magnoli Bocchi. Washington, DC: World of Economic Discovery. New York: W.W. Norton and Company. Bank, EAP Companion Volume to the WDR 2009. 328 WORLD DEVELOPMENT REPORT 2009 Wolf, Martin. 2004. Why Globalization Works. New Haven, CT: Yale ------. 2006f. Trends in Average Applied Tariff Rates in Developing University Press. and Industrial Countries, 1981­2005. Washington, DC: World Woods, Robert. 2000. The Demography of Victorian England and Bank, Development Research Group (digital file). Wales. Cambridge, U.K.: Cambridge University Press. ------. 2006g. World Development Indicators 2006. Washington, Wordie, J. R. 1983."The Chronology of English Enclosure, 1500­ DC: World Bank. 1914." Economic History Review 36 (4): 483­505. ------. 2006h. World Development Report 2007: Development and World Bank. 1977. Spatial Development in Mexico. Report No. the Next Generation. Washington, DC: World Bank. 1081-ME. Washington, DC: World Bank. ------. 2007a. Accelerating Growth and Development in the Lag- ------. 1980. The Development of Regional Cities in Thailand. ging Regions of India. Washington, DC: World Bank, ESW Report No. 2900-TH. Washington, DC: World Bank. Report. ------. 1998. Public Expenditure for Poverty Alleviation in North- ------. 2007b. Brazil: Inputs for a Strategy for Cities in Brazil. east Brazil: Promoting Growth and Improving Services. Washing- Report No. 35749-BR. Washington, DC: World Bank. ton, DC: World Bank. ------. 2007c. Brazil: Measuring Poverty using Household Con- ------. 2002a. Cities on the Move: A World Bank Urban Transport sumption. Report No. 36358-BR. Washington, DC: World Bank, Strategy Review. Washington, DC: World Bank. Poverty Reduction and Economic Management Sector Unit, ------. 2002b. From Natural Resources to the Knowledge Economy: Latin America, and the Caribbean Region. Trade and Job Quality. Washington, DC: World Bank, Latin ------. 2007d. Doing Business in 2008: Making a Difference. Wash- American and Caribbean Region. ington, DC: World Bank. ------. 2003a. Trade, Investment, and Development in the Middle ------. 2007e. East Asia and Pacific Update: Will Resilience Over- East and North Africa: Engaging with the World. Washington, come Risk? Washington, DC: World Bank. DC: World Bank. ------. 2007f. Making Finance Work for Africa. Washington, DC: ------. 2003b. World Bank Policy Research Report 2003: Land Poli- World Bank, Africa Region. cies for Growth and Poverty Reduction. New York: Oxford Uni- ------. 2007g. Making the Most of Scarcity: Accountability for versity Press. Better Water Management in the Middle East and North Africa. ------. 2004a. Global Economic Prospects 2005: Trade, Regional- Washington, DC: World Bank, Middle East and North Africa ism, and Development. Washington, DC: World Bank. Region. ------. 2004b. World Development Report 2005: A Better Invest- ------. 2007h. South Asia: Growth and Regional Integration. New ment Climate for Everyone. New York: Oxford University Press. Delhi: World Bank. ------. 2005a. Brazil: Regional Economic Development--(Some) ------. 2007i. World Bank Policy Research Report 2007: At Log- Lessons from Experience. Washington, DC: World Bank. gerheads? Agricultural Expansion, Poverty Reduction, and ------. 2005b. Chile: Zonas Extremas Policies and Beyond--An Environment in the Tropical Forests. Washington, DC: World Assessment of Costs and Impact with Recommendations of Ave- Bank. nues for Policy Reform. Report No. 27357-CH. Washington, DC: ------. 2007j. World Development Indicators 2007. Washington, World Bank. DC: World Bank. ------. 2005c. Food Safety and Agricultural Health Standards: ------. 2007k. World Development Report 2008: Agriculture for Challenges and Opportunities for Developing Country Exports. Development. Washington, DC: World Bank. Washington, DC: World Bank, PREM, and ARD. ------. 2008a."Iran: Spatial Patterns of Poverty and Economic ------. 2005d. Global Economic Prospects 2006. Economic Impli- Activity." World Bank, Washington, DC. cations of Remittances and Migration. Washington, DC: World ------. 2008b."Shrinking Distance: Identifying Priorities and Bank. Assessing Trade-offs for Territorial Development Policies--SDN ------. 2005e. Growth, Poverty and Inequality: Eastern Europe and Flagship Report, Spatial and Local." World Bank, Washington, the Former Soviet Union. Washington, DC: World Bank. DC. ------. 2006a. At Home and Away: Expanding Job Opportunities ------. 2008c. Transport for Development: The World Bank's Trans- for Pacific Islanders through Labour Mobility. Washington, DC: port Business Strategy for 2008­2012. Washington, DC: World World Bank, East Asia and Pacific Region. Bank. ------. 2006b. Doing Business in 2007: How to Reform. Washing- ------. 2008d. Transport Prices and Costs in Africa: A Review of the ton, DC: World Bank. Main International Corridors. Washington, DC: World Bank. ------. 2006c. Governance, Investment Climate, and Harmonious ------. 2008e. Turkey: Country Economic Memorandum, Sustain- Society: Competitiveness Enhancements for 120 Cities in China. ing High Growth--Selected Issues. Report No. 39194. Washing- Washington, DC: World Bank. ton, DC: World Bank. ------. 2006d. Housing Finance in Emerging Economies. Washing- Woronoff, Jon. 1972. West African Wager. Metuchen, NJ: Scarecrow ton, DC: World Bank. Press. ------. 2006e. Labor Migration in the Greater Mekong Sub-Region, Wrigley, E. Anthony. 1985."Urban Growth and Agricultural Synthesis Report Phase I. Washington, DC: World Bank, Regional Change: England and the Continent in the Early Modern Office for East Asia and the Pacific. Period." Journal of Interdisciplinary History 15 (4): 683­728. References 329 Xenos, Peter. 2004."Demographic Forces Shaping Youth Popula- Yemtsov, Ruslan. 2005."Quo Vadis? Inequality and Poverty tions in Asian Cities." In Youth Poverty, and Conflict in Southeast Dynamics across Russian Regions." In Spatial Inequality and Asian Cities, ed. Lisa Hanley, Blair Ruble, and Joseph Tulchin. Development, ed. Ravi Kanbur and Anthony J. Venables. Oxford, Washington, DC: Woodrow Wilson International Center for U.K.: Oxford University Press. Scholars. Yepes, Tito. 2008."Inclusive Growth for Latin America." World Yade, Mbaye, Anne Chohin-Kuper, Valerie Kelly, John Staatz, and Bank, Washington, DC. James Tefft. 1999."The Role of Regional Trade in Agricultural Yepes, Tito, and Somik V. Lall. 2008."Evaluating the Impact of Transformation." Paper presented at the Workshops on Struc- Upgrading Informal Settlements on Land Markets: Evidence tural Transformation in Africa, Nairobi. from Bogota." World Bank, Washington, DC. Yamauchi, Futoshi, Megumi Muto, Reno Dewina, and Sony Yoshino, Yutaka. 2008."Domestic Constraints, Firm Characteris- Sumaryanto. Forthcoming."Spatial Networks, Incentives, and tics, and Geographical Diversification of Firm-Level Manufac- the Dynamics of Village Economy: Evidence from Indonesia." turing Exports in Africa." Policy Research Working Paper 4575, In Reshaping Economic Geography in East Asia, ed. Yukon Huang World Bank, Washington, DC. and Alessandro Magnoli Bocchi. Washington, DC: World Bank, Young, Allyn. 1928."Increasing Returns and Economic Progress." EAP Companion Volume to the WDR 2009. Economic Journal 38 (152): 527­42. Yang, Dean. 2008."International Migration, Remittances and Yuen, Belinda. 2004."Planning Singapore Growth for Better Liv- Household Investment: Evidence from Philippine Migrants' ing." In Enhancing Urban Management in East Asia, ed. Belinda Exchange Rate Shocks." Economic Journal 118 (528): 591­630. Yuen and Mila Freire. Hampshire, U.K.: Ashgate Publishing. Yao, Yang. Forthcoming."The Political Economy of Government Yusuf, Shahid, and Kaoru Nabeshima. 2006. Post-industrial East Policies Toward Regional Inequality in China." In Reshaping Asian Cities, Innovation for Growth. Palo Alto, CA, and Washing- Economic Geography in East Asia, ed. Yukon Huang and Ales- ton, DC: World Bank and Stanford University Press. sandro Magnoli Bocchi. Washington, DC: World Bank, EAP Zhu, Nong, and Xubei Luo. 2008."Impact of Remittances on Rural Companion Volume to the WDR 2009. Poverty and Inequality in China." World Bank, East Asia and Yasuba, Yasukichi. 1978."Freight Rates and Productivity in Ocean Pacific Regional Office, Washington, DC. Transportation for Japan, 1875­1943." Explorations in Economic History 15 (1): 11­39. Selected Indicators Table A1 Geography and access Table A2 Urbanization Table A3 Territorial development Table A4 International integration Table A5 Other indicators Sources and definitions Selected world development indicators Introduction Classification of economies by region and income, FY2009 Table 1 Key indicators of development Table 2 Millennium Development Goals: eradicating poverty and improving lives Table 3 Economic activity Table 4 Trade, aid, and finance Table 5 Key indicators for other economies Technical notes 331 332 WORLD DEVELOPMENT REPORT 2009 Table A1 Geography and access Arable land Forest land Land Airports with Ports and Rail density Road density National average Surface (% of land (% of land Coastline boundaries paved runway terminals (rail km per (road km per distance to Location area (km2) area) area) (kms) (kms) (number) (number) 100 km2) 100 km2) capital city (kms) Economy 2007 2007 2007 2007 2007 2007 2007 2007 2000­06 a 2000­06 a 2000 Afghanistan 33 00 N, 65 00 E 652,090 12.1 1.3 0 5,529 11 0 .. 5.3 418 Albania 41 00 N, 20 00 E 28,750 21.1 29.0 362 720 3 4 1.6 65.7 84 Algeria 28 00 N, 3 00 E 2,381,740 3.2 1.0 998 6,343 52 9 0.2 4.5 1,108 American Samoa 14 20 S, 170 00 W 200 10.0 90.0 116 0 2 1 .. .. 7 Andorra 42 30 N, 1 30 E 470 2.1 34.2 0 120 0 0 .. .. 10 Angola 12 30 S, 18 30 E 1,246,700 2.6 47.4 1,600 5,198 31 3 0.2 4.1 711 Antigua and Barbuda 17 03 N, 61 48 W 440 18.2 20.5 153 0 2 1 .. 264.8 30 Argentina 34 00 S, 64 00 W 2,780,400 10.2 12.1 4,989 9,861 154 8 1.2 14.6 971 Armenia 40 00 N, 45 00 E 29,800 17.6 10.0 0 1,254 11 0 3.0 27.1 87 Aruba 12 30 N, 69 58 W 180 10.5 .. 69 0 1 3 .. .. 8 Australia 27 00 S, 133 00 E 7,741,220 6.4 21.3 25,760 0 311 11 0.5 10.5 1,946 Austria 47 20 N, 13 20 E 83,870 16.8 46.8 0 2,562 25 0 7.7 162.4 204 Azerbaijan 40 30 N, 47 30 E 86,600 22.3 11.3 0 2,013 27 0 2.6 71.5 219 Bahamas, The 24 15 N, 76 00 W 13,880 0.8 51.4 3,542 0 29 3 .. .. 230 Bahrain 26 00 N, 50 33 E 710 2.8 0.0 161 0 3 2 .. 492.7 25 Bangladesh 24 00 N, 90 00 E 144,000 61.1 6.7 580 4,246 15 2 2.1 183.8 165 Barbados 13 10 N, 59 32 W 430 37.2 4.7 97 0 1 1 .. 372.1 14 Belarus 53 00 N, 28 00 E 207,600 26.3 38.0 0 2,900 41 0 2.7 45.0 185 Belgium 50 50 N, 4 00 E 30,530 27.9 22.1 67 1,385 25 6 11.7 498.1 78 Belize 17 15 N, 88 45 W 22,970 3.1 72.5 386 516 5 2 .. .. 74 Benin 9 30 N, 2 15 E 112,620 24.0 21.3 121 1,989 1 1 0.7 17.2 398 Bermuda 32 20 N, 64 45 W 50 20.0 20.0 103 0 1 2 .. .. 7 Bhutan 27 30 N, 90 30 E 47,000 3.4 68.0 0 1,075 16 0 .. 17.1 99 Bolivia 17 00 S, 65 00 W 1,098,580 2.8 54.2 0 6,940 16 0 0.3 5.8 521 Bosnia and Herzegovina 44 00 N, 18 00 E 51,210 19.5 42.7 20 1,459 8 1 1.2 .. 103 Botswana 22 00 S, 24 00 E 581,730 0.7 21.1 0 4,013 10 0 0.2 4.3 441 Brazil 10 00 S, 55 00 W 8,514,880 7.0 56.5 7,491 16,885 714 9 0.3 20.7 1,378 Brunei Darussalam 4 30 N, 114 40 E 5,770 2.7 52.8 161 381 1 3 .. .. 61 Bulgaria 43 00 N, 25 00 E 111,000 29.2 33.4 354 1,808 132 2 4.0 40.5 190 Burkina Faso 13 00 N, 2 00 W 274,000 17.7 24.8 0 3,193 2 0 0.2 5.6 220 Burundi 3 30 S, 30 00 E 27,830 38.6 5.9 0 974 1 0 .. 48.0 84 Cambodia 13 00 N, 105 00 E 181,040 21.0 59.2 443 2,752 6 2 0.3 21.7 206 Cameroon 6 00 N, 12 00 E 475,440 12.8 45.6 402 4,591 11 2 0.2 10.7 363 Canada 60 00 N, 95 00 W 9,984,670 5.0 34.1 202,080 8,893 509 8 0.5 15.5 2,449 Cape Verde 16 00 N, 24 00 W 4,030 11.4 20.8 965 0 7 3 .. 33.5 156 Cayman Islands 19 30 N, 80 30 W 260 3.8 46.2 160 0 2 2 .. .. 15 Central African Republic 7 00 N, 21 00 E 623,000 3.1 36.5 0 5,203 3 0 .. .. 450 Chad 15 00 N, 19 00 E 1,284,000 2.9 9.5 0 5,968 7 0 .. .. 671 Channel Islands 190 0.0 .. .. .. .. .. .. .. .. Chile 30 00 S, 71 00 W 756,630 2.6 21.5 6,435 6,339 73 8 0.9 10.6 1,149 China 35 00 N, 105 00 E 9,598,088 11.1 21.2 14,500 22,117 403 7 0.8 20.7 1,668 Colombia 4 00 N, 72 00 W 1,141,750 1.8 54.7 3,208 6,309 101 7 0.3 .. 456 Comoros 12 10 S, 44 15 E 1,861 35.9 2.2 340 0 4 2 .. .. 59 Congo, Dem. Rep. 0 00 N, 25 00 E 2,344,860 3.0 58.9 37 10,730 25 11 0.2 6.8 1,006 Congo, Rep. 1 00 S, 15 00 E 342,000 1.4 65.8 169 5,504 4 6 0.3 5.1 479 Costa Rica 10 00 N, 84 00 W 51,100 4.4 46.8 1,290 639 32 2 0.5 69.2 105 Côte d'Ivoire 8 00 N, 5 00 W 322,460 10.4 32.7 515 3,110 7 4 0.2 25.2 236 Croatia 45 10 N, 15 30 E 56,540 19.8 38.2 1,777 2,197 23 5 4.9 50.9 144 Cuba 21 30 N, 80 00 W 110,860 27.9 24.7 3,735 29 78 3 3.8 .. 398 Cyprus 35 00 N, 33 00 E 9,250 10.8 18.8 648 0 13 3 .. 130.5 48 Czech Republic 49 45 N, 15 30 E 78,870 39.4 34.3 0 2,290 46 0 12.4 165.2 134 Denmark 56 00 N, 10 00 E 43,090 52.7 11.8 7,314 68 28 12 6.2 169.3 195 Djibouti 11 30 N, 43 00 E 23,200 0.0 0.3 314 516 3 1 0.4 .. 85 Dominica 15 25 N, 61 20 W 750 6.7 61.3 148 0 2 2 .. .. 20 Dominican Republic 19 00 N, 70 40 W 48,730 22.7 28.4 1,288 360 14 4 1.1 119 Ecuador 2 00 S, 77 30 W 283,560 4.9 39.2 2,237 2,010 98 5 0.3 15.6 266 Egypt, Arab Rep. 27 00 N, 30 00 E 1,001,450 3.0 0.1 2,450 2,665 72 3 0.5 9.3 558 El Salvador 13 50 N, 88 55 W 21,040 31.9 14.4 307 545 4 2 2.7 .. 68 Equatorial Guinea 2 00 N, 10 00 E 28,050 4.6 58.2 296 539 3 1 .. .. 301 Eritrea 15 00 N, 39 00 E 117,600 5.6 15.4 1,151 1,626 4 2 0.3 .. 184 Estonia 59 00 N, 26 00 E 45,230 13.9 53.9 3,794 633 12 5 2.3 134.1 128 Ethiopia 8 00 N, 38 00 E 1,104,300 11.1 13.0 0 5,328 14 0 0.1 3.6 425 Faeroe Islands 62 00 N, 7 00 W 1,400 2.1 .. 1,117 0 1 1 .. .. 25 Fiji 18 00 S, 175 00 E 18,270 10.9 54.7 1,129 0 3 3 3.3 .. 113 Finland 64 00 N, 26 00 E 338,150 7.3 73.9 1,250 2,681 76 10 1.9 25.7 417 France 46 00 N, 2 00 E 551,500 33.6 28.3 3,427 2,889 292 10 5.3 172.9 365 French Polynesia 15 00 S, 140 00 W 4,000 0.8 28.7 2,525 0 39 1 .. .. 472 Gabon 1 00 S, 11 45 E 267,670 1.3 84.5 885 2,551 11 5 0.3 3.6 317 Gambia, The 13 28 N, 16 34 W 11,300 31.5 47.1 80 740 1 1 37.4 123 Georgia 42 00 N, 43 30 E 69,700 11.5 39.7 310 1,461 19 2 2.3 29.1 156 Germany 51 00 N, 9 00 E 357,050 34.1 31.8 2,389 3,621 332 10 13.8 .. 334 Ghana 8 00 N, 2 00 W 238,540 18.4 24.2 539 2,094 7 2 0.4 21.0 344 Greece 39 00 N, 22 00 E 131,960 20.4 29.1 13,676 1,228 66 6 2.0 89.2 249 Greenland 72 00 N, 40 00 W 410,450 0.0 .. 44,087 0 9 1 .. .. 1,031 Grenada 12 07 N, 61 40 W 340 5.9 11.8 121 0 3 1 .. .. 11 Guam 13 28 N, 144 47 E 540 3.6 47.3 126 0 4 1 .. .. 14 Guatemala 15 30 N, 90 15 W 108,890 13.3 36.3 400 1,687 11 2 0.8 .. 175 Guinea 11 00 N, 10 00 W 245,860 4.5 27.4 320 3,399 5 1 0.3 18.0 348 Guinea-Bissau 12 00 N, 15 00 W 36,120 10.7 73.7 350 724 3 4 .. 12.3 97 Guyana 5 00 N, 59 00 W 214,970 2.4 76.7 459 2,949 9 1 .. .. 314 Selected indicators 333 Arable land Forest land Land Airports with Ports and Rail density Road density National average Surface (% of land (% of land Coastline boundaries paved runway terminals (rail km per (road km per distance to Location area (km2) area) area) (kms) (kms) (number) (number) 100 km2) 100 km2) capital city (kms) Economy 2007 2007 2007 2007 2007 2007 2007 2007 2000­06 a 2000­06 a 2000 Haiti 19 00 N, 72 25 W 27,750 28.3 3.8 1,771 360 4 1 .. .. 99 Honduras 15 00 N, 86 30 W 112,090 9.5 41.5 820 1,520 11 4 0.6 .. 167 Hong Kong, China 22 15 N, 114 10 E 1,092 0.0 .. 733 30 3 1 .. 186.5 .. Hungary 47 00 N, 20 00 E 93,030 51.3 22.1 0 2,271 20 0 9.0 178.0 137 Iceland 65 00 N, 18 00 W 103,000 0.1 0.5 4,970 0 5 5 .. 12.9 235 India 20 00 N, 77 00 E 3,287,260 53.7 22.8 7,000 14,103 243 8 2.1 113.8 992 Indonesia 5 00 S, 120 00 E 1,904,570 12.7 48.8 54,716 2,830 159 9 0.4 20.3 1,519 Iran, Islamic Rep. 32 00 N, 53 00 E 1,745,150 9.8 6.8 2,440 5,440 129 2 0.5 11.0 654 Iraq 33 00 N, 44 00 E 438,320 13.1 1.9 58 3,650 77 3 0.5 .. 281 Ireland 53 00 N, 8 00 W 70,270 17.6 9.7 1,448 360 15 5 4.7 140.2 165 Isle of Man 54 15 N, 4 30 W 570 0.0 5.2 160 0 1 3 11.4 .. 13 Israel 31 30 N, 34 45 E 22,070 14.6 7.9 273 1,017 30 4 3.9 80.6 110 Italy 42 50 N, 12 50 E 301,340 26.3 33.9 7,600 1,932 98 8 6.6 164.8 353 Jamaica 18 15 N, 77 30 W 10,990 16.1 31.3 1,022 0 11 5 .. 193.9 67 Japan 36 00 N, 138 00 E 377,910 12.0 68.2 29,751 0 145 10 6.4 323.0 531 Jordan 31 00 N, 36 00 E 88,780 2.1 0.9 26 1,635 15 1 0.6 8.5 171 Kazakhstan 48 00 N, 68 00 E 2,724,900 8.3 1.2 0 12,012 67 0 0.5 3.3 823 Kenya 1 00 N, 38 00 E 580,370 8.2 6.2 536 3,477 15 1 0.5 11.1 372 Kiribati 1 25 N, 173 00 E 810 2.7 2.7 1,143 0 3 1 .. .. 25 Korea, Dem. Rep. 40 00 N, 127 00 E 120,540 23.3 51.4 2,495 238 36 12 4.3 .. 211 Korea, Rep. 37 00 N, 127 30 E 99,260 16.6 63.5 2,416 462 69 5 3.5 101.6 187 Kuwait 29 30 N, 45 45 E 17,820 0.8 0.3 499 462 4 6 .. 32.3 67 Kyrgyz Republic 41 00 N, 75 00 E 199,900 6.7 4.5 0 3,878 18 0 0.2 9.8 255 Lao PDR 18 00 N, 105 00 E 236,800 4.3 69.9 0 5,083 9 0 .. 13.5 311 Latvia 57 00 N, 25 00 E 64,590 17.5 47.2 531 1,368 24 2 3.7 111.6 140 Lebanon 33 50 N, 35 50 E 10,400 16.6 13.3 225 454 5 4 3.9 .. 60 Lesotho 29 30 S, 28 30 E 30,350 10.9 0.3 0 909 3 0 .. .. 98 Liberia 6 30 N, 9 30 W 111,370 4.0 32.7 579 1,585 2 2 0.5 .. 196 Libya 25 00 N, 17 00 E 1,759,540 1.0 0.1 1,770 4,348 60 6 .. .. 910 Liechtenstein 47 16 N, 9 32 E 160 25.0 43.8 0 76 0 0 5.6 .. 6 Lithuania 56 00 N, 24 00 E 65,300 30.4 33.5 90 1,613 34 1 2.8 126.6 143 Luxembourg 49 45 N, 6 10 E 2,590 23.2 33.6 0 359 1 0 10.6 201.8 25 Macao, China 22 10 N, 113 33 E 28 0.0 .. 41 0 1 1 .. 1284.0 .. Macedonia, FYR 41 50 N, 22 00 E 25,710 22.3 35.6 0 766 10 0 2.7 .. 76 Madagascar 20 00 S, 47 00 E 587,040 5.1 22.1 4,828 0 29 4 0.1 .. 385 Malawi 13 30 S, 34 00 E 118,480 26.0 36.2 0 2,881 6 0 0.8 16.4 230 Malaysia 2 30 N, 112 30 E 329,740 5.5 63.6 4,675 2,669 37 9 0.6 30.0 873 Maldives 3 15 N, 73 00 E 300 13.3 3.3 644 0 2 1 .. .. 276 Mali 17 00 N, 4 00 W 1,240,190 3.9 10.3 0 7,243 9 0 0.1 1.5 804 Malta 35 50 N, 14 35 E 320 28.1 .. 197 0 1 1 .. 704.4 9 Marshall Islands 9 00 N, 168 00 E 180 11.1 .. 370 0 4 1 .. .. .. Mauritania 20 00 N, 12 00 W 1,030,700 0.5 0.3 754 5,074 8 2 0.1 .. 686 Mauritius 20 17 S, 57 33 E 2,040 49.3 18.2 177 0 2 1 .. 99.3 51 Mayotte 12 50 S, 45 10 E 374 0.0 13.4 185 0 1 1 .. .. 13 Mexico 23 00 N, 102 00 W 1,964,380 13.0 33.7 9,330 4,353 228 7 0.9 17.7 886 Micronesia, Fed. Sts. 6 55 N, 158 15 E 700 5.7 90.0 6,112 0 6 1 .. .. 134 Moldova 47 00 N, 29 00 E 33,840 56.2 10.0 0 1,389 6 0 3.5 38.7 87 Monaco 43 44 N, 7 24 E 2 0.0 0.0 4 4 0 1 .. .. .. Mongolia 46 00 N, 105 00 E 1,566,500 0.8 6.5 0 8,220 12 0 0.1 3.1 617 Montenegro 42 30 N, 19 18 E 14,026 0.0 .. 294 625 3 1 1.8 .. .. Morocco 32 00 N, 5 00 W 446,550 19.0 9.8 1,835 2,018 26 6 0.4 12.9 369 Mozambique 18 15 S, 35 00 E 799,380 5.5 24.6 2,470 4,571 22 3 0.4 .. 1,112 Myanmar 22 00 N, 98 00 E 676,580 15.3 49.0 1,930 5,876 21 3 0.6 .. 619 Namibia 22 00 S, 17 00 E 824,290 1.0 9.3 1,572 3,936 21 2 0.3 5.1 407 Nepal 28 00 N, 84 00 E 147,180 16.5 25.4 0 2,926 10 0 0.0 12.2 236 Netherlands 52 30 N, 5 45 E 41,530 26.8 10.8 451 1,027 20 7 8.3 372.2 95 Netherlands Antilles 12 15 N, 68 45 W 800 10.0 1.3 364 15 5 4 .. .. 36 New Caledonia 21 30 S, 165 30 E 18,580 0.3 39.2 2,254 0 11 1 .. .. 157 New Zealand 41 00 S, 174 00 E 267,710 5.6 31.0 15,134 0 45 5 1.5 34.7 418 Nicaragua 13 00 N, 85 00 W 130,000 15.9 42.7 910 1,231 11 3 0.0 15.4 192 Niger 16 00 N, 8 00 E 1,267,000 11.4 1.0 0 5,697 9 0 .. 1.1 896 Nigeria 10 00 N, 8 00 E 923,770 33.5 12.2 853 4,047 36 3 0.4 21.2 380 Northern Mariana Islands 460 0.0 69.2 .. .. .. .. .. .. 16 Norway 62 00 N, 10 00 E 323,800 2.8 30.8 2,650 2,542 67 8 1.4 30.2 507 Oman 21 00 N, 57 00 E 309,500 0.1 0.0 2,092 1,374 6 2 .. 11.3 462 Pakistan 30 00 N, 70 00 E 796,100 27.6 2.5 1,046 6,774 91 2 1.1 33.5 661 Palau 7 30 N, 134 30 E 460 8.7 87.0 1,519 0 1 1 .. .. 11 Panama 9 00 N, 80 00 W 75,520 7.4 57.7 2,490 555 53 3 0.5 15.6 179 Papua New Guinea 6 00 S, 147 00 E 462,840 0.5 65.0 5,152 820 21 3 .. .. 536 Paraguay 23 00 S, 58 00 W 406,750 7.7 46.5 0 3,995 12 0 0.0 363 Peru 10 00 S, 76 00 W 1,285,220 2.9 53.7 2,414 7,461 54 2 0.2 6.2 690 Philippines 13 00 N, 122 00 E 300,000 19.1 24.0 36,289 0 83 6 0.3 67.1 555 Poland 52 00 N, 20 00 E 312,690 39.6 30.0 491 3,056 83 4 7.5 138.5 237 Portugal 39 30 N, 8 00 W 92,120 16.8 41.3 1,793 1,214 43 4 3.0 85.8 237 Puerto Rico 18 15 N, 66 30 W 8,950 8.0 46.0 501 0 17 3 1.1 289.1 60 Qatar 25 30 N, 51 15 E 11,000 1.6 .. 563 60 3 1 .. .. 55 Romania 46 00 N, 25 00 E 238,390 40.4 27.7 225 2,508 25 4 5.0 86.4 246 Russian Federation 60 00 N, 100 00 E 17,098,240 7.4 49.4 37,653 20,097 616 10 0.5 3.3 4,322 Rwanda 2 00 S, 30 00 E 26,340 48.6 19.5 0 893 4 0 .. 56.8 65 Samoa 13 35 S, 172 20 W 2,840 21.2 60.4 403 0 3 1 .. 82.6 51 334 WORLD DEVELOPMENT REPORT 2009 Table A1 Geography and access--continued Arable land Forest land Land Airports with Ports and Rail density Road density National average Surface (% of land (% of land Coastline boundaries paved runway terminals (rail km per (road km per distance to Location area (km2) area) area) (kms) (kms) (number) (number) 100 km2) 100 km2) capital city (kms) Economy 2007 2007 2007 2007 2007 2007 2007 2007 2000­06 a 2000­06 a 2000 San Marino 43 46 N, 12 25 E 60 16.7 .. 0 39 0 0 .. .. 3 Sao Tome and Princip 1 00 N, 7 00 E 960 8.3 28.1 209 0 2 1 .. .. 45 Saudi Arabia 25 00 N, 45 00 E 2,000,000 1.8 1.4 2,640 4,431 73 4 0.1 7.6 593 Senegal 14 00 N, 14 00 W 196,720 12.8 45.0 531 2,640 9 1 0.5 7.1 328 Serbia 88,361 0.0 .. .. .. .. .. 4.3 .. .. Seychelles 4 35 S, 55 40 E 460 2.2 87.0 491 0 8 1 .. 99.6 413 Sierra Leone 8 30 N, 11 30 W 71,740 8.0 38.5 402 958 1 3 .. 15.8 170 Singapore 1 22 N, 103 48 E 699 0.9 2.9 193 0 9 1 .. 462.7 11 Slovak Republic 48 40 N, 19 30 E 49,030 28.9 40.1 0 1,524 18 0 7.6 89.4 199 Slovenia 46 07 N, 14 49 E 20,270 8.7 62.8 47 1,382 6 1 6.1 190.9 65 Solomon Islands 8 00 S, 159 00 E 28,900 0.6 77.6 5,313 0 2 5 .. 0.0 214 Somalia 10 00 N, 49 00 E 637,660 1.7 11.4 3,025 2,340 7 5 .. .. 623 South Africa 29 00 S, 24 00 E 1,219,090 12.1 7.6 2,798 4,862 146 6 1.7 30.0 608 Spain 40 00 N, 4 00 W 505,370 27.4 35.9 4,964 1,918 96 8 3.0 133.5 306 Sri Lanka 7 00 N, 81 00 E 65,610 14.2 29.9 1,340 0 14 2 2.2 150.5 157 St. Kitts and Nevis 17 20 N, 62 45 W 260 19.4 13.9 135 0 2 2 13.9 .. 8 St. Lucia 13 53 N, 60 58 W 620 6.6 27.9 158 0 2 3 .. .. 15 St. Vincent and the Grenadines 13 15 N, 61 12 W 390 17.9 28.2 84 0 5 1 .. 212.6 11 Sudan 15 00 N, 30 00 E 2,505,810 7.2 28.4 853 7,687 15 1 0.3 0.0 699 Suriname 4 00 N, 56 00 W 163,270 0.4 94.7 386 1,703 5 1 .. 2.8 248 Swaziland 26 30 S, 31 30 E 17,360 10.3 31.5 0 535 1 0 1.8 20.9 64 Sweden 62 00 N, 15 00 E 450,290 6.6 67.1 3,218 2,233 155 9 2.8 103.6 439 Switzerland 47 00 N, 8 00 E 41,280 10.3 30.5 0 1,852 42 0 12.1 178.0 99 Syrian Arab Republic 35 00 N, 38 00 E 185,180 26.5 2.5 193 2,253 26 2 1.5 51.6 289 Taiwan, China 23 30 N, 121 00 E 35,980 0.0 .. 1,566 0 38 5 7.8 115.6 168 Tajikistan 39 00 N, 71 00 E 142,550 6.6 2.9 0 3,651 17 0 0.3 19.8 240 Tanzania 6 00 S, 35 00 E 947,300 4.5 39.9 1,424 3,861 11 3 0.4 8.9 395 Thailand 15 00 N, 100 00 E 513,120 27.7 28.4 3,219 4,863 66 4 0.8 11.2 428 Timor-Leste 8 50 S, 125 55 E 14,870 8.2 53.7 706 228 3 1 .. .. .. Togo 8 00 N, 1 10 E 56,790 46.1 7.1 56 1,647 2 2 1.0 .. 300 Tonga 20 00 S, 175 00 W 750 20.8 5.6 419 0 1 1 .. .. 108 Trinidad and Tobago 11 00 N, 61 00 W 5,130 14.6 44.1 362 0 3 3 .. .. 48 Tunisia 34 00 N, 9 00 E 163,610 18.0 6.8 1,148 1,424 14 4 1.4 12.4 335 Turkey 39 00 N, 35 00 E 783,560 31.0 13.2 7,200 2,648 89 8 1.1 55.5 442 Turkmenistan 40 00 N, 60 00 E 488,100 4.7 8.8 0 3,736 22 0 0.5 .. 344 Uganda 1 00 N, 32 00 E 241,040 26.4 18.4 0 2,698 5 0 0.6 35.9 223 Ukraine 49 00 N, 32 00 E 603,550 56.0 16.5 2,782 4,663 193 8 3.9 29.2 373 United Arab Emirates 24 00 N, 54 00 E 83,600 0.8 3.7 1,318 867 23 7 .. .. 141 United Kingdom 54 00 N, 2 00 W 243,610 23.7 11.8 12,429 360 334 8 6.8 160.2 361 United States 38 00 N, 97 00 W 9,632,030 19.0 33.1 19,924 12,034 5119 12 2.5 70.2 2,595 Uruguay 33 00 S, 56 00 W 176,220 7.8 8.6 660 1,648 8 5 1.2 34.3 275 Uzbekistan 41 00 N, 64 00 E 447,400 11.0 7.7 0 6,221 34 0 0.9 .. 564 Vanuatu 16 00 S, 167 00 E 12,190 1.6 36.1 2,528 0 3 3 .. .. 250 Venezuela, RB 8 00 N, 66 00 W 912,050 2.9 54.1 2,800 4,993 129 5 0.1 .. 549 Vietnam 16 00 N, 106 00 E 329,310 21.3 41.7 3,444 4,639 26 2 0.8 71.7 646 Virgin Islands (U.S.) 18 20 N, 64 50 W 350 5.7 28.6 188 0 2 2 .. .. 10 West Bank and Gaza 32 00 N, 35 15 E 6,020 17.8 1.5 0 404 3 0 .. 83.0 .. Yemen, Rep. 15 00 N, 48 00 E 527,970 2.9 1.0 1,906 1,746 16 2 .. .. 406 Zambia 15 00 S, 30 00 E 752,610 7.1 57.1 0 5,664 10 0 0.3 12.3 436 Zimbabwe 20 00 S, 30 00 E 390,760 8.3 45.3 0 3,066 17 0 0.8 25.1 285 a. Data are for the latest year available in the period shown. Selected indicators 335 Table A2 Urbanization Urbanization Population density Rural-urban disparities Population Pop. in Pop. in % of rural Urban Urban density cities > largest % of urban % of rural % of urban population Agglomeration population Urban population (number of 1 million city population population population with index, 0 (low) (% of total population (% (% of total people (% of total (% of with water with water with sanitation sanitation to 100 (high) pop.) of total pop.) pop.) per km2) pop.) urban pop.) access access services services Economy 2000 2000 2005 2015 2006 2005 2005 2004 2004 2004 2004 Afghanistan 25.0 21.3 22.9 27.1 .. .. .. 63 31 49 29 Albania 52.7 41.8 45.4 52.8 115 .. .. 99 94 99 84 Algeria 58.7 59.8 63.3 69.3 14 9.7 15.4 88 80 99 82 American Samoa .. 88.8 91.3 94.1 292 .. .. .. .. .. .. Andorra .. 92.4 90.6 87.8 141 .. .. 100 100 100 100 Angola 26.8 50.0 53.3 59.7 13 17.2 32.2 75 40 56 16 Antigua and Barbuda .. 37.3 39.1 44.7 189 .. .. 95 89 98 94 Argentina 72.1 89.2 90.1 91.6 14 39.1 35.9 98 80 92 83 Armenia 69.6 65.1 64.1 64.1 107 36.5 57.0 99 80 96 61 Aruba .. 46.7 46.6 47.6 557 .. .. 100 100 .. .. Australia 75.9 87.2 88.2 89.9 3 60.4 24.1 100 100 100 100 Austria 67.9 65.8 66.0 67.7 100 27.4 41.6 100 100 100 100 Azerbaijan 48.7 50.9 51.5 52.8 102 22.1 42.9 95 59 73 36 Bahamas, The 57.8 88.8 90.4 92.2 32 .. .. 98 86 100 100 Bahrain 94.9 94.6 96.5 98.2 1021 .. .. 100 .. 100 .. Bangladesh 48.0 23.2 25.1 29.9 1178 11.8 32.3 82 72 51 35 Barbados 91.3 49.9 52.7 58.8 679 .. .. 100 100 99 100 Belarus 60.6 70.0 72.3 76.7 47 18.2 25.2 100 100 93 61 Belgium 89.8 97.1 97.2 97.5 347 9.7 9.9 100 .. .. .. Belize 1.8 47.7 48.3 51.2 13 .. .. 100 82 71 25 Benin 37.5 38.4 40.1 44.6 77 .. .. 78 57 59 11 Bermuda .. 100.0 100.0 100.0 1271 .. .. .. .. .. .. Bhutan 3.8 9.6 11.1 14.8 14 .. .. 86 60 65 70 Bolivia 55.7 61.8 64.2 68.8 8 31.0 25.9 95 68 60 22 Bosnia and Herzegovina 37.7 43.2 45.8 51.8 76 .. .. 99 96 99 92 Botswana 27.9 53.3 57.4 64.6 3 .. .. 100 90 57 25 Brazil 63.6 81.2 84.2 88.2 22 36.9 11.7 96 57 83 37 Brunei Darussalam 63.8 71.2 73.5 77.6 71 .. .. .. .. .. .. Bulgaria 64.9 68.9 70.0 72.8 71 14.1 20.2 100 97 100 96 Burkina Faso 12.8 16.6 18.3 22.8 51 .. 36.3 94 54 42 6 Burundi 31.7 8.6 10.0 13.5 306 .. .. 92 77 47 35 Cambodia 23.8 16.9 19.7 26.1 79 9.8 49.6 64 35 53 8 Cameroon 40.2 50.0 54.6 62.7 38 18.2 18.1 86 44 58 43 Canada 70.5 79.4 80.1 81.4 4 44.5 20.5 100 99 100 99 Cape Verde 44.4 53.4 57.3 64.3 126 .. .. 86 73 61 19 Cayman Islands .. 100.0 100.0 100.0 173 .. .. .. .. .. .. Central African Republic 19.7 37.6 38.1 40.4 7 .. .. 93 61 47 12 Chad 12.1 23.4 25.3 30.5 8 .. 34.6 41 43 24 4 Channel Islands .. 30.5 30.5 31.5 782 .. .. .. .. .. .. Chile 74.8 86.0 87.6 90.1 22 34.9 39.8 100 58 95 62 China 37.2 35.8 40.4 49.2 140 17.7 2.8 93 67 69 28 Colombia 62.1 71.2 72.7 75.7 41 36.0 23.7 99 71 96 54 Comoros .. 33.8 37.0 44.0 323 .. .. 92 82 41 29 Congo, Dem. Rep. 25.6 29.8 32.1 38.6 26 16.2 32.1 82 29 42 25 Congo, Rep. 54.2 58.3 60.2 64.2 11 32.5 54.0 84 27 28 25 Costa Rica 54.0 59.0 61.7 66.9 85 28.1 45.6 100 92 89 97 Côte d'Ivoire 35.1 43.1 45.0 49.8 58 19.2 42.8 97 74 46 29 Croatia 37.3 55.6 56.5 59.5 79 .. .. 100 100 100 100 Cuba 64.2 75.6 75.5 74.7 103 19.4 25.7 95 78 99 95 Cyprus 62.1 68.7 69.3 71.5 82 .. .. 100 100 100 100 Czech Republic 73.8 74.0 73.5 74.1 132 11.4 15.6 100 100 99 97 Denmark 48.8 85.1 85.6 86.9 128 20.1 23.5 100 100 .. .. Djibouti 40.6 83.4 86.1 89.6 35 .. .. 76 59 88 50 Dominica .. 71.1 72.9 76.4 96 .. .. 100 90 86 75 Dominican Republic 71.7 62.4 66.8 73.6 196 21.4 32.0 97 91 81 73 Ecuador 49.2 60.3 62.8 67.6 47 29.9 29.1 97 89 94 82 Egypt, Arab Rep. 90.4 42.5 42.8 45.4 73 20.5 35.7 99 97 86 58 El Salvador 73.7 58.4 59.8 63.2 322 22.7 38.0 94 70 77 39 Equatorial Guinea 21.4 38.8 38.9 41.1 17 .. .. 45 42 60 46 Eritrea 21.4 17.8 19.4 24.4 45 .. .. 74 57 32 3 Estonia 45.3 69.4 69.1 70.1 32 .. .. 100 99 97 96 Ethiopia 11.9 14.9 16.0 19.1 75 3.8 24.1 81 11 44 7 Faeroe Islands .. 36.3 38.8 41.5 35 .. .. .. .. .. .. Fiji 17.7 48.3 50.8 56.1 45 .. .. 43 51 87 55 Finland 52.4 61.1 61.1 62.7 17 20.8 34.0 100 100 100 100 France 72.5 75.8 76.7 79.0 111 22.4 21.0 100 100 .. .. French Polynesia 45.8 52.4 51.7 52.3 70 .. .. 100 100 99 97 Gabon 35.9 80.2 83.6 87.7 5 .. .. 95 47 37 30 Gambia, The 44.0 49.1 53.9 61.8 162 .. .. 95 77 72 46 Georgia 50.2 52.7 52.2 53.8 64 23.4 44.8 96 67 96 91 Germany 79.6 75.1 75.2 76.3 236 7.7 5.5 100 100 100 100 Ghana 34.1 44.0 47.8 55.1 99 15.5 18.4 88 64 27 11 Greece 57.5 58.8 59.0 61.0 86 29.1 49.3 .. .. .. .. Greenland .. 81.6 82.9 85.5 0 .. .. .. .. .. .. Grenada .. 31.0 30.6 32.2 313 .. .. 97 93 96 97 336 WORLD DEVELOPMENT REPORT 2009 Table A2 Urbanization--continued Urbanization Population density Rural-urban disparities Population Pop. in Pop. in % of rural Urban Urban density cities > largest % of urban % of rural % of urban population Agglomeration population Urban population (number of 1 million city population population population with index, 0 (low) (% of total population (% (% of total people (% of total (% of with water with water with sanitation sanitation to 100 (high) pop.) of total pop.) pop.) per km2) pop.) urban pop.) access access services services Economy 2000 2000 2005 2015 2006 2005 2005 2004 2004 2004 2004 Guam 2.4 93.2 94.1 95.3 312 .. .. 100 100 99 98 Guatemala 36.6 45.1 47.2 52.0 117 .. 16.4 99 92 90 82 Guinea 15.0 31.0 33.0 38.1 37 15.8 48.0 78 35 31 11 Guinea-Bissau 20.9 29.7 29.6 31.1 57 .. .. 79 49 57 23 Guyana 36.1 28.6 28.2 29.4 4 .. .. 83 83 86 60 Haiti 33.9 35.6 38.8 45.5 337 22.9 59.0 52 56 57 14 Honduras 41.6 44.4 46.5 51.4 61 .. 29.2 95 81 87 54 Hong Kong, China 99.8 100.0 100.0 100.0 6539 103.3 103.3 .. .. .. .. Hungary 71.9 64.6 66.3 70.3 113 16.8 25.3 100 98 100 85 Iceland 57.1 92.3 92.8 93.6 3 .. .. 100 100 100 100 India 52.4 27.7 28.7 32.0 368 11.6 5.8 95 83 59 22 Indonesia 55.2 42.0 48.1 58.5 122 11.6 12.5 87 69 73 40 Iran, Islamic Rep. 60.9 64.2 66.9 71.9 42 22.9 15.8 99 84 .. .. Iraq 69.9 67.8 66.9 66.9 .. .. .. 97 50 95 48 Ireland 45.8 59.2 60.5 63.8 60 24.9 41.2 100 .. .. .. Isle of Man .. 51.8 51.8 52.8 133 .. .. .. .. .. .. Israel 81.3 91.4 91.6 91.9 320 43.5 47.5 100 100 100 .. Italy 78.0 67.2 67.6 69.5 199 17.4 8.5 100 .. .. .. Jamaica 69.4 51.8 53.1 56.7 245 .. .. 98 88 91 69 Japan 90.9 65.2 65.8 68.2 351 47.8 41.9 100 100 100 100 Jordan 77.9 80.4 82.3 85.3 61 23.9 29.0 99 91 94 87 Kazakhstan 50.6 56.3 57.3 60.3 6 7.6 13.3 97 73 87 52 Kenya 25.4 19.7 20.7 24.1 63 7.8 37.6 83 46 46 41 Kiribati .. 43.0 47.4 55.4 122 .. .. 77 53 59 22 Korea, Dem. Rep. 46.2 60.2 61.6 65.5 196 18.9 23.0 100 100 58 60 Korea, Rep. 86.4 79.6 80.8 83.1 489 50.6 24.7 97 71 .. .. Kuwait 85.2 98.2 98.3 98.5 142 71.4 72.6 .. .. .. .. Kyrgyz Republic 34.0 35.4 35.8 38.1 27 .. 43.3 98 66 75 51 Lao PDR 13.5 18.9 20.6 24.9 25 .. .. 79 43 67 20 Latvia 52.6 68.1 67.8 68.9 37 .. .. 100 96 82 71 Lebanon 79.0 86.0 86.6 87.9 392 44.3 51.2 100 100 100 87 Lesotho 23.1 17.9 18.7 22.0 65 .. .. 92 76 61 32 Liberia 17.8 54.3 58.1 64.8 36 .. 46.8 72 52 49 7 Libya 80.4 83.1 84.8 87.4 3 54.3 41.8 .. .. 97 96 Liechtenstein .. 15.1 14.6 14.7 217 .. .. .. .. .. .. Lithuania 56.1 67.0 66.6 66.8 54 .. .. .. .. .. .. Luxembourg 75.1 83.8 82.8 82.1 176 .. .. 100 100 .. .. Macao, China 58.8 100.0 100.0 100.0 16776 .. .. .. .. .. .. Macedonia, FYR 63.5 64.9 68.9 75.2 80 .. .. .. .. .. .. Madagascar 19.5 26.0 26.8 30.1 32 8.5 31.7 77 35 48 26 Malawi 23.8 15.1 17.2 22.1 141 .. .. 98 68 62 61 Malaysia 68.0 61.8 67.3 75.4 78 5.5 8.1 100 96 95 93 Maldives .. 27.5 29.6 34.8 984 .. .. 98 76 100 42 Mali 18.4 27.9 30.5 36.5 10 11.8 38.6 78 36 59 39 Malta 91.5 93.4 95.3 97.2 1261 .. .. 100 100 100 .. Marshall Islands .. 65.8 66.7 69.3 351 .. .. 82 96 93 58 Mauritania 26.3 40.0 40.4 43.1 3 .. .. 59 44 49 8 Mauritius 92.1 42.7 42.4 44.1 612 .. .. 100 100 95 94 Mayotte 0.0 .. .. .. 481 .. .. .. .. .. .. Mexico 68.4 74.7 76.0 78.7 53 35.0 24.8 100 87 91 41 Micronesia, Fed. Sts. .. 22.3 22.3 23.6 157 .. .. 95 94 61 14 Moldova 49.4 46.1 46.7 50.0 118 .. .. 97 88 86 52 Monaco .. 100.0 100.0 100.0 16667 .. .. 100 .. 100 .. Mongolia 34.4 56.6 56.7 58.8 2 .. 59.6 87 30 75 37 Montenegro 37.4 .. .. .. 44 .. .. .. .. .. .. Morocco 53.0 55.1 58.7 65.0 68 15.9 17.7 99 56 88 52 Mozambique 24.1 30.7 34.5 42.4 26 6.4 18.6 72 26 53 19 Myanmar 33.1 28.0 30.7 37.4 73 8.6 28.0 80 77 88 72 Namibia 13.3 32.4 35.1 41.1 2 .. .. 98 81 50 13 Nepal 26.0 13.4 15.8 20.9 189 .. 19.0 96 89 62 30 Netherlands 88.2 76.8 80.2 84.9 482 13.8 8.8 100 100 100 100 Netherlands Antilles 81.5 69.3 70.4 73.4 233 .. .. .. .. .. .. New Caledonia 50.6 61.9 63.7 67.4 13 .. .. .. .. .. .. New Zealand 64.7 85.7 86.2 87.4 15 27.8 32.2 100 .. .. .. Nicaragua 48.0 57.2 59.0 63.0 45 21.3 36.2 90 63 56 34 Niger 14.3 16.2 16.8 19.3 10 .. 38.1 80 36 43 4 Nigeria 40.8 43.9 48.2 55.9 155 13.3 16.0 67 31 53 36 Northern Mariana Islands .. 93.3 94.5 95.9 175 .. .. 98 97 94 96 Norway 46.8 76.1 77.4 78.6 15 .. 22.4 100 100 .. .. Oman 68.5 71.6 71.5 72.3 8 .. .. .. .. 97 .. Pakistan 53.6 33.2 34.9 39.6 202 17.8 21.4 96 89 92 41 Palau .. 69.6 69.7 70.9 44 .. .. 79 94 96 52 Panama 52.6 65.8 70.8 77.9 43 37.6 53.1 99 79 89 51 Papua New Guinea 3.5 13.2 13.4 15.0 13 .. .. 88 32 67 41 Paraguay 45.7 55.3 58.5 64.4 15 31.5 53.8 99 68 94 61 Selected indicators 337 Urbanization Population density Rural-urban disparities Population Pop. in Pop. in % of rural Urban Urban density cities > largest % of urban % of rural % of urban population Agglomeration population Urban population (number of 1 million city population population population with index, 0 (low) (% of total population (% (% of total people (% of total (% of with water with water with sanitation sanitation to 100 (high) pop.) of total pop.) pop.) per km2) pop.) urban pop.) access access services services Economy 2000 2000 2005 2015 2006 2005 2005 2004 2004 2004 2004 Peru 52.1 71.6 72.6 74.9 21 26.3 36.3 89 65 74 32 Philippines 56.1 58.6 62.7 69.6 284 14.2 20.2 87 82 80 59 Poland 67.2 61.7 62.1 64.0 125 4.4 7.1 100 .. .. .. Portugal 62.6 54.4 57.6 63.6 115 38.6 45.4 .. .. .. .. Puerto Rico 90.3 94.7 97.6 99.3 441 66.6 68.2 .. .. .. .. Qatar 87.1 95.0 95.4 96.2 72 .. .. 100 100 100 100 Romania 65.2 54.6 53.7 56.1 94 8.9 16.6 91 16 89 .. Russian Federation 64.8 73.4 73.0 72.6 9 19.2 10.2 100 88 93 70 Rwanda 14.3 13.8 19.3 28.7 374 .. 43.7 92 69 56 38 Samoa .. 21.9 22.4 24.9 65 .. .. 90 87 100 100 San Marino 60.0 93.5 97.2 99.3 470 .. .. .. .. .. .. Sao Tome and Princip 46.2 53.4 58.0 65.8 159 .. .. 89 73 32 20 Saudi Arabia 75.7 79.9 81.0 83.2 12 36.2 22.4 97 .. 100 .. Senegal 43.0 40.6 41.6 44.7 61 18.3 44.1 92 60 79 34 Serbia 60.5 .. .. .. 84 .. .. .. .. .. .. Seychelles .. 51.1 52.9 58.2 180 .. .. 100 75 .. 100 Sierra Leone 29.3 37.0 40.7 48.2 78 .. 35.2 75 46 53 30 Singapore 96.2 100.0 100.0 100.0 6302 99.6 99.6 100 .. 100 .. Slovak Republic 59.3 56.3 56.2 58.0 112 .. .. 100 99 100 98 Slovenia 48.1 50.8 51.0 53.3 99 .. .. .. .. .. .. Solomon Islands 6.9 15.7 17.0 20.5 17 .. .. 94 65 98 18 Somalia 21.4 33.3 35.2 40.1 13 16.1 45.7 32 27 48 14 South Africa 50.2 56.9 59.3 64.1 39 30.0 11.7 99 73 79 46 Spain 76.7 76.3 76.7 78.3 87 24.0 16.8 100 100 100 100 Sri Lanka 38.2 15.7 15.1 15.7 304 .. .. 98 74 98 89 St. Kitts and Nevis .. 32.8 32.2 33.5 185 .. .. 99 99 96 96 St. Lucia 75.4 28.0 27.6 29.0 270 .. .. 98 98 89 89 St. Vincent and the Grenadines .. 44.4 45.9 50.0 305 .. .. .. 93 .. 96 Sudan 31.9 36.1 40.8 49.4 16 12.2 30.0 78 64 50 24 Suriname 70.4 72.1 73.9 77.4 3 .. .. 98 73 99 76 Swaziland 20.2 23.3 24.1 27.5 66 .. .. 87 54 59 44 Sweden 54.4 84.0 84.2 85.1 22 18.9 22.5 100 100 100 100 Switzerland 75.8 73.1 75.2 78.8 186 15.4 20.5 100 100 100 100 Syrian Arab Republic 57.2 50.1 50.6 53.4 103 25.4 26.4 98 87 99 81 Taiwan, China 84.1 .. .. .. 705 .. .. .. .. .. .. Tajikistan 36.2 25.9 24.7 24.6 47 .. .. 92 48 70 45 Tanzania 28.2 22.3 24.2 28.9 43 7.0 28.7 85 49 53 43 Thailand 35.6 31.1 32.3 36.2 123 10.5 32.4 98 100 98 99 Timor-Leste 0.0 24.5 26.5 31.2 66 .. .. 77 56 66 33 Togo 35.7 36.6 40.1 47.4 115 21.4 53.4 80 36 71 15 Tonga .. 23.2 24.0 27.4 138 .. .. 100 100 98 96 Trinidad and Tobago 81.6 10.8 12.2 15.8 258 .. .. 92 88 100 100 Tunisia 48.7 63.4 65.3 69.1 65 .. .. 99 82 96 65 Turkey 60.1 64.7 67.3 71.9 94 25.6 20.0 98 93 96 72 Turkmenistan 42.6 45.1 46.3 50.8 10 .. .. 93 54 77 50 Uganda 28.0 12.1 12.6 14.5 147 4.6 36.2 87 56 54 41 Ukraine 63.9 67.2 67.8 70.2 81 13.1 8.4 99 91 98 93 United Arab Emirates 61.0 77.4 76.7 77.4 49 32.4 42.2 100 100 98 95 United Kingdom 84.4 89.4 89.7 90.6 249 26.1 15.7 100 100 .. .. United States 71.9 79.1 80.8 83.7 32 43.3 7.8 100 100 100 100 Uruguay 64.1 91.4 92.0 93.1 19 38.2 41.6 100 100 100 99 Uzbekistan 54.2 37.3 36.7 38.0 62 8.3 22.7 95 75 78 61 Vanuatu .. 21.7 23.5 28.1 18 .. .. 86 52 78 42 Venezuela, RB 80.5 91.1 93.4 95.9 30 36.9 11.7 85 70 71 48 Vietnam 47.1 24.3 26.4 31.6 268 13.4 23.1 99 80 92 50 Virgin Islands (U.S.) .. 92.6 94.2 96.0 311 .. .. .. .. .. .. West Bank and Gaza 57.9 71.5 71.6 72.9 602 .. .. 94 88 78 61 Yemen, Rep. 23.0 25.4 27.3 31.9 40 8.5 31.3 71 65 86 28 Zambia 30.8 34.8 35.0 37.0 15 11.0 31.4 90 40 59 52 Zimbabwe 33.4 33.8 35.9 40.9 34 11.5 32.2 98 72 63 47 338 WORLD DEVELOPMENT REPORT 2009 Table A3 Territorial development Leading area: defined as the area with the highest measure of welfare Lagging area: defined as the area with the lowest measure of welfare (income or consumption or gdp) per capita (income or consumption or gdp) per capita Area's Area's welfare welfare measure measure Area's as a % of Population Area's as a % of Population number of poor country's density number of poor country's density as percentage average (number of as percentage average (number of Poverty of total country welfare Area people per Poverty of total country welfare Area people per Incidence poor measure (km2) km2) Incidence poor measure (km2) km2) Economy Area name 1995­2006a 1995­2006a 1995­2006a 2007 latest census Area name 1995­2006a 1995­2006a 1995­2006a 2007 latest census Albania Tirane 21.8 13.0 111 1,193 439 Bulqize 56.2 2.8 66 718 60 Argentina Ciudad Autonoma 18.7 5.4 256 203 13,676 Santiago del 31.4 2.6 38 136,351 6 de Buenos Aires Estero Armenia Yerevan 44.7 34.2 124 210 5,196 Armavir 52.0 9.3 68 1,241 20 Australia Australian Capital 14.8 0.9 139 2,432 129 Tasmania 31.1 2.8 89 68,127 7 Territory Azerbaijan Baku 49.0 25.1 109 2,130 944 Nakhchivan AR 45.0 4.1 90 5,500 65 Bangladesh Dhaka 46.7 30.0 157 30,772 1,257 Rajshahi 56.7 28.2 22 13,218 2,269 Belarus Minsk 13.3 14.4 107 40,800 38 Gomel 17.3 18.5 93 40,400 38 Belize Belize 24.5 21.8 123 4,204 16 Toledo 57.6 17.5 74 4,649 5 Benin Littoral 8.9 2.5 185 79 6,795 Mono 58.6 8.5 64 1,396 201 Bolivia Santa Cruz 40.6 17.8 131 370,621 5 Potosi 76.6 11.8 56 118,218 6 Brazil Sao Paulo 17.8 12.6 154 248,177 149 Piaui 57.1 3.1 31 251,312 11 Bulgaria Sofia-city 3.2 4.4 111 1,349 858 Kardzhali 19.3 3.7 81 3,209 50 Burkina Faso Centre 22.3 4.9 197 2,805 413 Centre Sud 66.1 6.1 61 11,313 43 Burundi Muramvya 37.6 3.5 141 696 363 Ruyigi 55.4 6.2 62 2,339 130 Cambodia Phnom Penh 11.9 3.6 249 375 2,547 Siemreap 53.7 11.8 63 10,299 67 Cameroon Douala (capital of 10.9 2.6 183 .. .. Extreme-Nord 56.3 24.9 67 34,246 80 Littoral) Canada Alberta 13.3 8.6 111 661,848 5 Newfoundland 18.7 1.9 77 405,212 1 Chad Ennedi 21.0 0.01 295 .. .. Mayo-Dala 79.0 4.9 68 .. .. Chile Region 13.5 28.9 130 15,782 384 Maule 23.1 7.4 70 30,518 30 Metropolitana Costa Rica Central 17.1 34.8 117 10,669 47 Huetar Atlantic 23.6 10.8 38 9,189 12 Côte d'Ivoire Lagunes 17.9 6.1 160 14,200 230 Marahoue 56.4 5.4 62 8,500 59 Croatia Grad Zagreb 2.7 4.1 133 641 1,216 Viroviticko- 19.8 3.6 68 2,024 46 Podravska zupanija Djibouti Ali Sabieh 92.4 7.0 192 2,600 6 Djibouti 36.2 58.3 92 600 528 Dominican Distrito Nacional 21.5 6.2 159 91 9,897 Elias Pina 74.0 1.5 39 1,397 46 Republic Ecuador Pichincha 5.1 7.9 144 9,110 259 Pastaza 34.7 1.6 49 29,774 2 El Salvador San Salvador 6.8 12.6 151 886 1,668 Cabanas 32.6 5.5 45 1,104 123 Estonia Harjumaa 7.9 33.7 228 4,333 120 Hiiumaa 11.4 1.0 2 1,023 10 Ethiopia Addis Ababa City 57.0 4.0 197 530 4,574 Benishangul 71.0 1.1 72 49,289 11 Gumuz Gabon Estuaire 23.0 35.8 121 20,740 29 Ogooue-Ivindo 59.9 7.4 55 46,075 1 Gambia, The Banjul 50.0 19.2 183 88 4,060 Upper River 80.0 15.7 52 2,070 88 Ghana Greater Accra 2.4 1.4 182 2,593 1,121 Upper East 79.6 14.9 35 8,842 104 Guatemala Guatemala 11.7 4.6 212 2,126 810 San Marcos 86.7 12.5 39 3,791 166 Guinea Conakry 24.4 7.6 140 308 3,523 Labe 66.3 15.1 75 24,144 33 Haiti Ouest 57.0 25.7 162 4,595 543 Nord-Est 94.0 4.2 41 1,698 147 Honduras Islas de la Bahia 57.6 0.4 154 261 120 Lempira 94.7 5.2 43 4,290 57 India Kerala 15.0 1.7 140 38,863 819 Bihar 41.4 12.2 76 99,200 837 Indonesia Jakarta 4.3 0.8 289 664 12,516 Jawa Central 28.4 18.3 16 32,549 930 Jamaica Saint Andrew and 16.4 16.8 138 431 1,282 Saint Ann 33.8 10.3 61 1,213 136 Kingston Jordan Amman 7.8 28.6 124 8,231 236 Al-Mafraq 29.2 13.5 72 26,435 9 Kenya Nairobi Province 44.0 6.5 244 684 3,133 Eastern Province 57.6 18.5 65 159,891 29 Kyrgyz Republic Chuy Oblast 33.1 9.4 136 20,200 38 Naryn 98.1 9.0 65 45,200 6 Madagascar Antananarivo 61.7 25.3 158 58,283 79 Fianarantsoa 81.1 24.4 65 102,373 33 Malawi Southern Region 68.1 48.7 119 31,754 146 Northern Region 62.5 11.9 81 26,931 46 Mali Bamako 28.2 5.2 174 267 3,952 Sikasso 76.4 21.7 78 71,741 24 Mauritania Nouakchott 29.0 15.3 130 1,000 498 Guidimakha 71.6 10.8 60 10,300 14 Mexico Distrito Federal 31.8 5.6 182 1,479 5,896 Oaxaca 68.0 4.8 49 93,952 37 Mongolia Ulaanbaatar 26.0 22.5 116 .. .. West 49.0 23.8 79 .. .. Morocco Grand Casablanca 4.0 3.0 159 1,615 1,870 Gharb-Chrarda- 23.3 9.2 47 8,805 182 Beni Hssen Mozambique Maputo (city) 47.8 4.2 192 602 1,631 Inhambane 82.6 8.6 70 68,615 17 Namibia Khomas 23.5 5.1 274 36,805 7 Ohangwena 85.8 16.9 38 10,582 22 Nepal Western 27.1 17.1 119 29,398 155 Far Western 41.0 12.4 76 19,539 112 Nicaragua Managua 3.6 4.3 161 3,465 314 Esteli 23.4 4.5 21 2,230 78 Niger Niamey 26.2 2.7 180 670 1,065 Maradi 80.4 26.2 64 38,581 58 Nigeria Bayelsa 26.2 0.6 162 9,363 182 Jigawa 89.5 5.4 36 23,415 186 Pakistan Punjab 32.4 54.1 104 205,344 359 Azad Kashmir 15.6 1.0 85 11,639 241 Panama Panamá 22.7 29.0 127 9,633 140 Ngöbe Buglé 98.7 10.3 15 6,673 16 Paraguay Asuncion 24.8 6.2 135 117 4,244 San Pedro 51.1 8.1 72 20,002 16 Peru Lima 24.5 16.8 137 32,137 254 Huancavelica 88.7 3.5 39 22,131 21 Philippines National Capital 5.7 2.6 216 630 15,766 Region V (Bicol 49.0 10.5 63 14,544 321 region (NCR) region) Poland Mazowieckie 10.8 9.1 163 35,728 142 Lubelskie 21.2 7.9 67 25,115 89 Selected indicators 339 Leading area: defined as the area with the highest measure of welfare Lagging area: defined as the area with the lowest measure of welfare (income or consumption or gdp) per capita (income or consumption or gdp) per capita Area's Area's welfare welfare measure measure Area's as a % of Population Area's as a % of Population number of poor country's density number of poor country's density as percentage average (number of as percentage average (number of Poverty of total country welfare Area people per Poverty of total country welfare Area people per Incidence poor measure (km2) km2) Incidence poor measure (km2) km2) Economy Area name 1995­2006a 1995­2006a 1995­2006a 2007 latest census Area name 1995­2006a 1995­2006a 1995­2006a 2007 latest census Romania Bucharest 3.0 2.2 216 1,821 1,186 North-East 25.0 31.4 69 36,850 100 Russian Saint Petersburg 7.8 1.1 117 23,900 197 Republic Tyva 66.5 0.6 56 37,300 8 Federation Rwanda Prefecture de la 12.3 0.6 379 313 753 Gikongoro 77.2 8.0 70 1,974 237 Ville de Kigali Sierra Leone Western Area 80.8 19.0 200 557 1,707 Eastern 80.0 23.5 65 15,553 76 Slovak Republic Bratislava 9.4 10.3 172 2,052 292 Presov 9.7 14.0 73 8,981 88 South Africa Gauteng 19.0 6.6 186 17,010 520 Northern Province 77.0 18.0 46 123,910 40 (Limpopo) Sri Lanka Colombo 6.0 3.7 179 642 3,480 Monaragala 37.0 4.0 51 7,133 56 Tajikistan Dushanbe (City) 43.5 6.9 144 300 1,873 Khatlon 73.3 44.8 79 24,600 87 (Qurghonteppa) Tanzania Dar Es Salaam 19.1 4.3 188 1,393 1,793 Rukwa 36.4 3.8 68 68,635 17 Thailand Krung Thep Maha 0.5 1.2 174 1,569 13,016 Nong Bua Lam 35.2 2.1 39 3,859 125 Nakhon (Bangkok) Phu Turkmenistan Lebap 28.5 17.4 116 93,800 11 Ahal 34.8 12.4 79 95,400 6 Uganda Central 22.3 17.1 146 .. .. Northern 63.3 29.9 54 .. .. United States New Jersey 8.7 1.9 129 20,168 429 Mississippi 21.0 1.5 73 123,515 23 Venezuela, RB Capital 15.0 13.6 126 9,880 518 Zuliana 25.2 15.7 83 63,100 56 Vietnam Ho Chi Minh city 5.3 1.0 241 2,090 2,409 Lai Chau 79.8 1.7 46 9,065 65 Yemen, Rep. Sana'a 16.6 4.2 184 380 4,827 Al Jawf 40.8 2.7 71 39,500 12 Zambia Lusaka 52.0 10.0 170 21,898 64 Western 89.0 9.4 59 126,386 6 340 WORLD DEVELOPMENT REPORT 2009 Table A4 International integration People Ideas Trade Countries International Average for which Cost of voice traffic International Telephone tariffs and Share of Countries this obtaining International (incoming Internet average cost custom trade with that need a country's a passport migration stock and outgoing, bandwith of call to the Total trade Index of duties (% neighboring visa to visit residents relative to (% of minutes per (bits per US (US$ per as share of shipping of import countries (% this country need a visa GDP per foreigners) person) person) three minutes) GDP difficulties value) of total trade) (number) (number) capita (%) (%) (minutes) (bits) (US$) (%) (Index) (%) (%) Average Economy 2004 2004 2005 2005 2000­06a 2000­06a 2000­06a 2005­06a 2008 2005 2000­2005 Afghanistan 192 168 .. 0.1 0.6 0.2 0.39 68.1 174 11.2 .. Albania 142 159 2.2 2.6 160.0 3.8 1.34 74.2 70 7.6 71.9 Algeria 183 157 .. 0.7 16.9 4.8 2.08 71.4 114 3.0 10.5 American Samoa 192 156 .. 35.0 .. .. .. .. .. .. 0.0 Andorra 132 89 .. 79.1 .. 6344.4 .. .. .. .. 0.0 Angola 191 162 4.9 0.4 6.7 11.5 3.23 111.7 164 .. .. Antigua and Barbuda 105 121 0.2 22.0 596.4 16588.0 .. 130.8 55 .. 17.4 Argentina 124 88 0.7 3.9 32.5 689.9 .. 43.9 107 15.8 35.9 Armenia 162 144 0.0 7.8 127.8 22.5 2.42 58.5 118 3.3 10.4 Aruba .. .. .. 24.4 .. 1794.6 .. .. .. .. 5.4 Australia 161 59 0.4 20.2 213.8 11593.4 0.68 42.1 34 1.8 0.0 Austria 132 57 0.3 15.0 264.7 6633.5 0.71 109.7 12 0.0 81.3 Azerbaijan 181 143 2.5 2.2 32.6 35.7 4.18 111.3 173 ,, 26.7 Bahamas, The 119 119 0.2 9.8 585.0 278.4 .. .. .. 55.0 0.0 Bahrain 140 139 0.3 40.7 587.1 564.3 1.74 127.0 .. 3.7 23.7 Bangladesh 19 155 5.2 0.7 6.4 8.0 2.02 44.2 112 32.6 8.9 Barbados 72 113 0.7 9.7 565.2 2055.3 1.95 117.8 8.0 22.6 Belarus 179 141 .. 12.2 64.0 191.7 1.90 124.1 137 7.2 68.9 Belgium 132 54 0.3 6.9 316.3 11278.5 0.75 172.8 48 73.8 Belize 120 127 0.4 13.9 178.2 604.8 2.59 125.4 116 14.1 Benin 167 144 8.6 2.1 6.4 5.4 4.80 39.6 124 24.5 20.4 Bermuda .. .. .. 29.6 8699.1 .. .. .. .. 0.0 Bhutan 191 153 .. 1.5 40.9 33.9 0.66 76.8 149 1.5 .. Bolivia 140 110 6.7 1.3 48.7 43.3 1.89 75.1 115 2.1 50.8 Bosnia and Herzegovina 154 154 7.0 1.0 207.6 .. 39.6 3.62 72.3 53 0.0 57.9 Botswana 139 138 0.1 4.5 74.1 16.1 2.88 83.8 145 .. 50.2 Brazil 140 85 1.3 0.3 11.7 149.9 0.71 26.4 93 .. 13.2 Brunei Darussalam 157 89 33.2 142.7 1453.1 .. 96.2 36 .. 7.3 Bulgaria 147 103 0.7 1.3 71.7 1756.1 0.57 147.0 89 2.4 25.5 Burkina Faso 175 147 12.7 5.8 10.8 15.0 1.14 35.8 170 12.7 40.7 Burundi 191 163 50.9 1.3 1.6 0.5 2.45 58.7 167 .. 14.8 Cambodia 191 159 .. 2.2 9.5 1.3 2.94 144.6 139 21.6 10.2 Cameroon 184 157 11.4 0.8 8.8 8.7 .. 52.7 132 .. 12.4 Canada 149 57 0.3 18.9 438.7 6731.9 .. 72.0 39 1.3 73.0 Cape Verde 174 145 .. 2.2 139.5 46.3 6.08 74.6 51 .. 1.1 Cayman Islands 121 156 .. 35.8 1630.2 .. .. .. .. .. 0.0 Central African Republic 173 153 17.7 1.9 2.3 0.4 1.99 35.5 172 19.4 16.1 Chad 181 154 60.2 4.5 2.0 0.5 9.11 97.3 157 .. Channel Islands .. .. .. 45.8 .. .. 2.90 .. .. .. 0.0 Chile 114 84 1.5 1.4 48.3 779.6 2.18 76.3 43 1.6 12.4 China 191 161 2.9 0.0 7.3 195.7 2.90 72.4 42 ­16.2 15.4 Colombia 64 150 1.5 0.3 68.2 560.2 .. 47.3 105 8.8 19.2 Comoros 192 158 .. 11.2 33.1 3.3 .. 47.3 119 .. 0.0 Congo, Dem. Rep. 190 162 125.0 0.9 5.3 0.1 .. 70.4 154 27.4 .. Congo, Rep. 169 151 8.5 7.2 .. 0.3 5.39 137.0 171 6.6 .. Costa Rica 125 100 0.4 10.2 126.8 176.2 1.93 105.5 54 5.0 7.6 Cote d'Ivoire 169 145 .. 13.1 16.9 3.0 2.25 92.3 147 43.6 20.2 Croatia 131 105 2.0 14.9 231.2 1073.7 .. 104.6 96 1.6 50.0 Cuba 169 153 .. 0.7 30.6 13.8 7.49 .. .. .. 0.4 Cyprus 137 80 0.3 15.3 693.2 593.8 0.33 .. .. 0.7 8.5 Czech Republic 136 95 0.1 4.4 94.8 2169.8 1.06 148.3 30 0.0 61.8 Denmark 132 53 0.4 7.2 318.0 34796.1 0.89 100.8 2 .. 74.9 Djibouti 192 159 .. 2.6 26.7 56.0 4.73 97.2 66 .. .. Dominica 0 130 1.0 6.3 .. 419.8 .. 107.2 80 .. 30.5 Dominican Republic 130 152 .. 1.7 218.5 6.1 0.22 73.5 35 13.9 .. Ecuador 27 139 3.6 0.9 215.6 227.2 1.75 67.5 131 15.4 Egypt, Arab Rep. 60 156 1.9 0.2 30.1 126.4 1.45 61.5 26 6.4 5.4 El Salvador 120 108 0.4 0.3 409.9 22.9 2.40 74.0 68 6.0 29.1 Equatorial Guinea 191 159 .. 1.2 .. 34.7 .. 144.7 133 .. .. Eritrea 190 163 .. 0.3 8.7 1.7 3.59 58.1 159 .. 4.5 Estonia 132 98 0.2 15.0 109.0 11174.9 0.90 169.3 7 0.0 54.6 Ethiopia 190 163 31.1 0.8 3.5 0.1 4.01 57.5 150 26.6 6.6 Faeroe Islands .. .. .. 11.1 .. 3312.6 .. .. .. .. 3.1 Fiji 94 131 1.6 2.0 112.1 87.0 2.84 127.6 111 15.9 1.3 Finland 132 55 0.2 3.0 178.3 4311.2 1.80 82.3 5 0.0 65.4 France 132 54 0.2 10.6 182.8 3285.5 0.84 55.1 25 0.0 62.4 French Polynesia .. .. .. 13.1 .. 887.2 3.67 29.1 .. .. 0.0 Gabon 174 156 .. 17.7 74.0 152.6 2.77 89.1 106 .. 2.4 Gambia, The 41 135 6.0 15.3 .. 5.6 1.81 110.2 73 .. 4.1 Georgia 131 143 .. 4.3 57.5 7.2 0.68 89.9 64 4.0 39.4 Selected indicators 341 People Ideas Trade Countries International Average for which Cost of voice traffic International Telephone tariffs and Share of Countries this obtaining International (incoming Internet average cost custom trade with that need a country's a passport migration stock and outgoing, bandwith of call to the Total trade Index of duties (% neighboring visa to visit residents relative to (% of minutes per (bits per US (US$ per as share of shipping of import countries (% this country need a visa GDP per foreigners) person) person) three minutes) GDP difficulties value) of total trade) (number) (number) capita (%) (%) (minutes) (bits) (US$) (%) (Index) (%) (%) Average Economy 2004 2004 2005 2005 2000­06a 2000­06a 2000­06a 2005­06a 2008 2005 2000­2005 Germany 132 54 0.3 12.3 190.8 6863.8 0.43 84.7 10 63.6 Ghana 171 142 1.4 7.5 20.1 9.3 0.39 103.0 61 28.5 10.7 Greece 132 56 0.4 8.8 181.8 586.5 1.09 45.6 65 0.0 56.4 Greenland .. .. 1.0 21.4 .. 106.8 2.41 .. .. .. 0.0 Grenada 11 127 .. 10.2 624.5 3976.0 .. 109.0 52 .. 25.9 Guam .. .. .. 66.9 .. .. .. .. .. .. 0.0 Guatemala 118 110 1.8 0.4 194.7 55.5 1.21 46.2 116 8.9 21.4 Guinea 171 146 .. 4.5 6.8 0.2 4.61 67.4 102 .. 10.1 Guinea-Bissau 176 150 .. 1.2 8.9 1.2 .. 95.4 109 .. .. Guyana 158 127 0.6 0.1 118.3 48.7 .. 211.8 101 .. 25.2 Haiti 4 157 .. 0.4 .. 16.7 2.15 57.3 153 .. .. Honduras 128 110 3.4 0.4 96.4 6.0 2.52 107.3 103 6.3 20.6 Hong Kong, China 42 89 0.2 43.2 1178.7 13438.6 0.77 399.4 3 .. 49.1 Hungary 132 85 0.3 3.1 105.1 993.3 1.01 155.1 45 0.0 26.5 Iceland 132 59 0.2 7.8 240.0 7289.6 0.84 83.1 11 1.1 0.7 India 189 160 3.6 0.5 3.0 24.3 1.19 48.8 79 14.6 9.9 Indonesia 162 148 1.9 0.1 5.3 6.8 2.79 56.9 41 3.0 4.1 Iran, Islamic Rep. 188 166 .. 2.9 8.8 53.2 0.55 75.2 135 5.5 4.8 Iraq .. .. .. 0.1 .. .. .. .. 175 .. .. Ireland 107 57 0.3 14.1 709.5 5911.6 0.71 149.9 20 0.1 60.6 Isle of Man .. .. .. 48.6 .. .. .. .. .. .. 0.0 Israel 123 88 0.3 38.4 364.2 2455.4 0.59 88.4 8 0.7 0.9 Italy 132 55 0.2 4.3 236.0 2044.0 0.79 56.5 62 .. 59.4 Jamaica 80 126 1.4 0.7 233.2 15822.2 0.87 108.8 92 7.7 0.8 Japan 137 56 0.2 1.6 43.4 1037.8 1.63 27.3 18 .. 0.0 Jordan 61 159 .. 41.1 138.8 57.3 1.44 146.6 59 10.4 31.8 Kazakhstan 175 141 .. 16.5 26.4 62.5 .. 91.6 178 6.0 34.4 Kenya 21 140 1.2 1.0 5.6 20.8 3.00 62.2 148 10.2 13.8 Kiribati 168 133 .. 2.6 24.2 5.4 8.82 133.3 97 .. 0.0 Korea, Dem. Rep. 192 163 .. 0.2 .. .. .. .. .. .. .. Korea, Rep. 89 78 0.3 1.1 91.6 1027.8 0.76 85.3 13 3.4 0.0 Kuwait 154 135 .. 65.8 .. 347.9 1.51 98.0 99 1.3 .. Kyrgyz Republic 144 143 .. 5.6 29.5 38.9 5.40 115.7 177 13.2 35.8 Lao PDR 192 158 11.8 0.4 6.6 3.5 1.11 78.2 158 .. .. Latvia 132 98 .. 19.5 66.7 3229.7 1.63 108.6 19 0.6 55.0 Lebanon 100 169 4.0 16.4 279.1 111.0 2.19 63.6 83 7.6 9.9 Lesotho 127 135 1.5 0.3 18.1 2.2 3.28 149.0 129 49.5 66.4 Liberia 176 151 .. 1.5 .. 0.1 .. 99.6 98 .. .. Libya 177 163 .. 10.5 65.6 20.6 .. 84.1 .. .. 8.6 Liechtenstein 117 74 .. 33.7 .. 4298.0 .. .. .. .. .. Lithuania 132 100 0.4 4.8 49.0 2714.4 1.55 129.8 23 0.5 48.0 Luxembourg 132 55 0.0 38.0 1399.1 20459.0 15.96 326.6 32 .. 89.1 Macao, China 0 122 .. 55.9 497.2 6491.7 1.12 150.5 .. .. 39.9 Macedonia, FYR 148 141 .. 6.0 63.3 16.7 .. 118.3 72 .. 45.8 Madagascar 0 156 .. 0.3 1.3 1.8 0.59 70.7 126 24.8 0.0 Malawi 123 139 .. 2.2 4.8 1.5 3.56 46.4 161 .. 16.4 Malaysia 23 63 1.7 6.5 87.9 124.5 0.71 217.0 21 5.6 22.0 Maldives 3 134 .. 1.0 91.2 179.5 5.86 178.1 110 23.6 10.7 Mali 172 148 .. 0.3 7.4 25.9 12.28 72.3 162 3.8 25.1 Malta 132 73 0.3 2.6 222.4 4729.1 0.77 179.6 .. 0.1 19.5 Marshall Islands 162 144 .. 2.6 76.5 26.2 .. .. 46 .. 0.0 Mauritania 169 141 8.7 2.1 20.3 29.6 .. 113.7 152 .. 2.1 Mauritius 88 131 0.5 1.7 149.8 153.2 1.59 127.1 17 16.8 0.0 Mayotte .. .. .. .. .. .. .. .. .. .. 0.0 Mexico 139 92 1.2 0.6 174.0 109.0 0.83 65.1 76 4.1 72.3 Micronesia, Fed. Sts. 0 146 2.5 3.2 80.8 54.5 6.00 .. 85 .. 0.0 Moldova 149 141 .. 11.4 109.9 147.4 1.46 139.2 122 4.1 39.7 Monaco 132 88 .. 75.8 .. .. .. .. .. .. .. Mongolia 174 155 .. 0.4 4.8 13.3 4.92 125.0 168 5.7 64.7 Montenegro .. .. .. 6.4 .. .. .. 128.9 113 .. .. Morocco 134 147 2.1 0.4 65.1 377.1 1.69 71.4 67 9.4 16.6 Mozambique 0 158 .. 2.1 12.7 0.9 1.17 88.9 140 .. 43.7 Myanmar 192 161 2.3 0.2 2.8 1.9 0.17 .. .. 2.3 .. Namibia 141 145 1.0 7.1 58.0 17.8 4.28 110.0 144 31.8 63.1 Nepal 0 158 26.3 3.0 5.6 4.6 2.04 45.3 151 18.2 60.5 Netherlands 132 56 0.2 10.0 310.8 20501.3 0.32 140.6 14 0.8 56.4 Netherlands Antilles .. .. .. 26.5 .. .. .. .. .. .. 0.0 New Caledonia .. .. .. 18.4 215.0 562.9 3.13 .. .. .. 0.4 New Zealand 146 59 0.2 15.7 361.1 1106.7 1.30 58.2 16 1.7 0.0 Nicaragua 38 116 3.0 0.5 61.7 1.1 3.15 92.1 87 4.6 28.4 342 WORLD DEVELOPMENT REPORT 2009 Table A4 International integration--continued People Ideas Trade Countries International Average for which Cost of voice traffic International Telephone tariffs and Share of Countries this obtaining International (incoming Internet average cost custom trade with that need a country's a passport migration stock and outgoing, bandwith of call to the Total trade Index of duties (% neighboring visa to visit residents relative to (% of minutes per (bits per US (US$ per as share of shipping of import countries (% this country need a visa GDP per foreigners) person) person) three minutes) GDP difficulties value) of total trade) (number) (number) capita (%) (%) (minutes) (bits) (US$) (%) (Index) (%) (%) Average Economy 2004 2004 2005 2005 2000­06a 2000­06a 2000­06a 2005­06a 2008 2005 2000­2005 Niger 167 144 19.8 0.9 2.0 2.3 8.77 38.9 163 .. 20.7 Nigeria 174 149 10.0 0.7 2.3 1.1 1.49 91.1 138 .. 4.4 Northern Mariana Islands 157 153 .. 6.5 .. .. .. .. .. .. 0.0 Norway 132 56 0.3 7.4 192.9 9304.8 0.31 75.0 4 0.2 18.8 Oman 128 141 0.2 24.4 189.1 173.6 1.87 99.2 104 2.8 32.1 Pakistan 184 165 4.2 2.1 10.5 4.6 1.03 38.6 94 13.0 11.1 Palau 0 148 0.7 15.1 .. .. .. 153.5 121 .. 0.0 Panama 130 107 3.2 54.9 286.6 3.64 144.5 9 8.6 9.7 Papua New Guinea 114 138 5.3 0.4 8.0 1.0 4.32 134.8 82 26.4 1.7 Paraguay 163 103 2.9 31.3 83.1 0.90 115.2 123 8.2 41.4 Peru 97 135 2.6 0.1 99.1 366.6 1.80 48.5 71 5.7 21.5 Philippines 44 144 0.8 0.5 28.3 38.0 1.20 94.0 57 20.4 0.0 Poland 132 86 0.5 1.8 60.6 560.2 1.35 82.0 40 0.4 52.7 Portugal 132 57 0.4 7.2 178.1 829.0 1.04 70.0 31 0.0 28.2 Puerto Rico .. .. .. 10.7 .. 511.2 .. 181.2 95 .. 0.0 Qatar 156 136 .. 78.3 842.9 943.6 1.95 101.7 .. 3.2 9.2 Romania 145 107 1.1 0.6 49.1 1503.2 0.82 78.5 38 3.0 18.7 Russian Federation 183 134 0.4 8.4 15.3 100.3 2.03 55.1 155 29.2 33.1 Rwanda 180 156 41.5 1.3 .. 7.4 2.43 43.2 166 .. 43.2 Samoa 0 129 2.1 5.0 149.7 49.0 1.36 78.1 108 .. 6.5 San Marino 132 82 .. 33.4 6448.4 5419.6 .. .. .. 1.7 .. Sao Tome and Princip 0 155 .. 4.8 51.3 25.8 5.11 .. 91 .. 3.1 Saudi Arabia 187 142 0.8 27.5 215.8 126.1 2.40 92.9 33 .. 3.2 Senegal 151 146 4.1 2.8 39.4 102.7 1.02 69.8 136 33.1 9.1 Serbia .. .. .. 6.4 .. 94.6 .. 73.4 58 .. .. Seychelles 0 131 1.1 5.8 .. 307.3 3.78 244.5 84 10.7 0.0 Sierra Leone 124 139 .. 2.2 .. 0.1 .. 59.4 130 27.0 2.3 Singapore 32 66 0.2 42.4 1045.4 7052.4 0.69 473.5 1 0.1 14.7 Slovak Republic 131 95 0.5 2.3 90.0 2912.6 1.06 176.0 90 0.0 66.8 Slovenia 132 92 0.3 8.4 1254.7 0.65 139.1 69 0.1 67.1 Solomon Islands 123 130 .. 0.7 23.2 17.1 .. 102.2 74 .. 0.0 Somalia 192 165 .. 3.4 .. 0.4 .. .. .. .. .. South Africa 120 118 0.7 2.4 27.6 18.8 0.79 63.1 134 4.2 2.9 Spain 132 55 0.1 11.0 173.3 2775.7 0.60 58.4 47 0.0 64.7 Sri Lanka 114 156 2.4 1.9 27.9 25.1 2.11 74.8 60 14.6 0.6 St. Kitts and Nevis 92 120 0.2 9.3 571.6 42.5 .. 113.5 22 31.3 15.5 St. Lucia 130 121 0.5 5.3 217.9 94.9 .. 117.6 88 .. 22.8 St. Vincent and the Grenadines 122 124 0.6 8.7 288.5 25.0 3.97 125.0 75 .. 33.8 Sudan 190 166 .. 1.8 12.2 5.4 39.18 42.8 143 .. 5.2 Suriname 171 137 .. 1.2 276.7 439.3 1.33 76.3 86 .. 4.4 Swaziland 137 138 0.3 4.0 47.5 0.9 2.97 167.5 146 47.7 80.0 Sweden 132 54 0.2 12.4 17468.5 0.41 94.5 6 .. 25.9 Switzerland 119 57 0.2 22.3 664.8 9609.1 0.32 89.0 37 1.1 61.4 Syrian Arab Republic 160 163 .. 5.2 44.0 8.0 4.81 75.0 127 .. 16.0 Taiwan, China .. .. .. .. 292.3 6569.5 0.51 134.0 29 .. 10.8 Tajikistan .. .. 13.4 4.7 10.4 0.3 7.84 80.7 176 11.1 .. Tanzania 35 142 13.4 2.1 1.4 0.4 3.17 55.1 100 .. 8.5 Thailand 136 146 1.0 1.6 14.1 156.2 0.67 143.5 50 6.2 7.5 Timor-Leste .. .. .. 0.6 .. .. .. .. 78 .. .. Togo 0 148 .. 3.0 21.5 15.6 3.98 83.9 81 .. 28.1 Tonga 142 135 2.4 1.1 .. 20.1 1.09 54.3 44 .. 0.0 Trinidad and Tobago 47 119 0.3 2.9 375.7 370.2 2.19 108.0 49 4.8 10.3 Tunisia 126 142 1.1 0.4 72.6 126.4 2.28 108.7 28 6.2 26.3 Turkey 118 126 8.9 1.8 27.1 630.7 2.40 64.1 56 1.1 9.3 Turkmenistan 174 148 .. 4.6 6.0 15.7 .. 126.2 .. .. .. Uganda 0 145 .. 1.8 3.1 4.4 3.21 44.4 141 20.5 26.1 Ukraine 181 139 2.7 14.5 56.6 17.3 1.65 97.3 120 4.3 40.6 United Arab Emirates 155 136 .. 70.9 .. 2371.4 1.73 170.6 24 .. .. United Kingdom 104 55 0.2 9.0 262.2 13062.0 0.77 61.6 27 .. 55.2 United States 158 52 0.2 12.9 279.5 3306.6 .. 26.8 15 1.1 30.7 Uruguay 134 98 .. 2.5 120.7 484.0 0.52 60.2 125 5.1 36.2 Uzbekistan 184 148 .. 4.8 12.4 8.7 13.95 63.4 165 .. .. Vanuatu 106 137 3.4 0.5 .. 23.2 7.45 100.3 142 .. 0.0 Venezuela, RB 131 95 0.5 3.8 23.1 50.3 0.84 57.6 156 4.9 14.6 Vietnam 185 160 2.3 0.0 8.4 84.1 1.95 150.3 63 .. 11.7 Virgin Islands (U.S.) .. .. .. 33.7 .. 414.0 .. .. .. .. 0.0 West Bank and Gaza .. .. .. 46.3 65.7 198.7 1.17 85.4 77 .. Yemen, Rep. 145 161 .. 1.3 12.0 0.3 2.39 79.4 128 .. 7.6 Zambia 18 142 3.7 2.4 6.9 10.9 1.41 67.8 160 9.0 14.1 Zimbabwe 117 141 .. 3.9 24.9 4.2 4.36 129.8 169 .. 53.9 a. Data are for the latest year available in the period shown. Selected indicators 343 Table A5 Other indicators Terrain characteristics Geography and people Population living Population living Population Population at less than 25 kms at less than living at less than living at less than plains lowlands plateaus hills mountains from an international 75 kms from an 25 kms from a 75 kms from a (% of total (% of total (% of total (% of total (% of total border international border coastline coastline land area) land area) land area) land area) land area) (%) (%) (%) (%) Economy 2007 2007 2007 2007 2007 2000 2000 2000 2000 Afghanistan 0.9 0.0 8.7 0.3 90.1 11.2 37.1 0.0 0.0 Albania 0.0 0.4 0.0 33.6 66.0 29.2 83.2 41.9 87.2 Algeria 12.4 4.2 51.7 5.1 26.6 3.3 13.0 35.8 63.4 American Samoa 0.0 100.0 0.0 0.0 0.0 0.0 0.0 100.0 100.0 Andorra 0.0 0.0 0.0 0.0 100.0 100.0 100.0 0.0 0.0 Angola 16.9 3.0 47.0 6.0 27.1 5.2 11.5 23.0 26.9 Antigua and Barbuda 0.0 100.0 0.0 0.0 0.0 0.0 0.0 100.0 100.0 Argentina 33.0 9.4 8.7 11.0 37.9 5.0 10.4 33.3 42.8 Armenia 0.0 0.0 0.0 0.0 100.0 70.5 100.0 0.0 0.0 Aruba 35.4 64.6 0.0 0.0 0.0 0.0 0.0 100.0 100.0 Australia 29.7 23.3 33.5 8.8 4.8 0.0 0.0 69.1 87.5 Austria 0.0 0.2 0.0 26.7 73.1 38.7 97.4 0.0 0.0 Azerbaijan 0.7 31.2 0.0 5.6 62.5 26.9 61.9 0.0 0.0 Bahamas, The 100.0 0.0 0.0 0.0 0.0 0.0 0.0 100.0 100.0 Bahrain 57.3 42.7 0.0 0.0 0.0 0.0 0.0 100.0 100.0 Bangladesh 67.1 20.3 0.0 10.8 1.8 28.2 78.8 25.3 48.1 Barbados 0.0 100.0 0.0 0.0 0.0 0.0 0.0 100.0 100.0 Belarus 37.8 58.7 3.5 0.0 0.0 18.3 53.5 0.0 0.0 Belgium 19.2 45.5 0.0 35.3 0.0 53.4 100.0 18.1 69.9 Belize 1.4 51.5 7.7 39.5 0.0 37.9 94.2 57.9 94.4 Benin 13.3 15.5 65.2 6.0 0.0 42.7 96.0 27.7 46.0 Bermuda 100.0 0.0 0.0 0.0 0.0 0.0 0.0 100.0 100.0 Bhutan 0.0 0.4 0.0 16.2 83.4 54.8 99.9 0.0 0.0 Bolivia 26.3 8.4 14.8 13.1 37.3 5.0 22.5 0.0 0.0 Bosnia and Herzegovina 0.0 6.0 0.0 23.0 71.0 40.5 90.1 0.8 12.4 Botswana 49.4 0.0 48.5 0.0 2.1 34.2 66.2 0.0 0.0 Brazil 21.1 20.8 37.6 8.0 12.5 1.0 2.4 25.4 46.5 Brunei Darussalam 0.3 75.7 0.0 0.0 23.9 98.9 100.0 80.8 99.3 Bulgaria 0.0 21.9 0.0 45.4 32.7 19.1 80.5 9.4 15.8 Burkina Faso 34.4 0.1 65.6 0.0 0.0 14.4 42.2 0.0 0.0 Burundi 0.0 0.0 15.9 0.0 84.1 60.9 100.0 0.0 0.0 Cambodia 6.5 74.3 0.0 17.9 1.4 18.1 57.0 4.9 12.1 Cameroon 2.2 2.0 44.7 20.7 30.4 17.5 40.2 15.8 20.8 Canada 6.7 20.8 27.9 21.3 23.2 17.6 61.8 20.6 23.1 Cape Verde 0.7 40.9 0.0 31.7 26.8 0.0 0.0 100.0 100.0 Cayman Islands 100.0 0.0 0.0 0.0 0.0 0.0 0.0 100.0 100.0 Central African Republic 17.6 0.0 78.5 0.0 3.9 31.4 58.5 0.0 0.0 Chad 35.9 0.0 42.4 3.7 18.0 23.6 51.5 0.0 0.0 Channel Islands 0.0 100.0 0.0 0.0 0.0 0.0 0.0 100.0 100.0 Chile 0.0 9.0 0.0 26.9 64.1 3.3 48.9 26.2 53.1 China 3.9 10.2 8.9 12.7 64.2 1.1 3.7 11.6 21.8 Colombia 24.0 23.1 9.7 12.5 30.7 4.7 11.1 11.5 24.9 Comoros 0.0 0.0 12.4 26.2 61.4 0.0 0.0 100.0 100.0 Congo, Dem. Rep. 17.1 0.3 61.1 1.0 20.5 24.9 46.5 0.6 1.8 Congo, Rep. 29.1 1.2 58.6 8.2 2.8 50.5 91.8 19.8 21.5 Costa Rica 0.0 1.9 0.0 62.0 36.1 8.2 29.4 22.1 97.7 Cote d'Ivoire 0.0 0.1 0.1 72.9 27.0 11.3 30.0 26.0 33.7 Croatia 6.8 25.3 60.8 6.4 0.6 68.1 98.1 31.2 36.3 Cuba 0.0 41.3 0.0 25.7 32.9 1.3 11.4 63.5 100.0 Cyprus 5.8 80.1 0.0 14.2 0.0 0.0 0.0 90.1 100.0 Czech Republic 0.0 44.1 0.0 55.9 0.0 38.3 88.0 0.0 0.0 Denmark 30.1 69.9 0.0 0.0 0.0 2.0 9.1 94.8 100.0 Djibouti 0.0 18.6 0.0 30.0 51.4 86.3 100.0 79.8 99.1 Dominica 0.0 0.0 0.0 100.0 0.0 0.0 0.0 100.0 100.0 Dominican Republic 0.1 31.2 0.0 33.2 35.5 3.7 17.3 61.2 96.7 Ecuador 7.4 6.1 9.7 28.1 48.7 5.2 15.8 33.7 49.0 Egypt, Arab Rep. 5.7 27.8 41.4 16.4 8.7 0.2 0.5 16.4 36.9 El Salvador 0.0 1.1 0.0 52.5 46.4 25.4 96.7 26.7 94.7 Equatorial Guinea 0.0 20.6 18.6 7.7 53.1 35.6 77.9 41.3 59.5 Eritrea 0.1 11.6 0.0 14.6 73.8 16.7 56.2 11.1 51.8 Estonia 39.8 60.2 0.0 0.0 0.0 13.7 48.1 64.4 78.5 Ethiopia 0.7 0.3 14.1 4.8 80.1 3.0 9.1 0.0 0.0 Faeroe Islands 0.0 100.0 0.0 0.0 0.0 0.0 0.0 100.0 100.0 Fiji 0.0 43.2 0.0 56.8 0.0 0.0 0.0 88.3 100.0 Finland 2.0 71.7 6.3 19.2 0.8 4.7 14.2 48.1 67.4 France 3.9 41.0 3.0 30.1 21.9 12.0 23.2 20.8 36.2 French Polynesia 0.0 9.2 6.2 84.5 0.0 0.0 0.0 100.0 100.0 Gabon 5.2 21.3 52.0 15.7 5.7 13.2 32.7 44.9 53.0 Gambia, The 100.0 0.0 0.0 0.0 0.0 92.2 100.0 74.2 85.7 Georgia 0.0 0.0 0.0 6.9 93.0 21.8 91.7 15.2 28.1 344 WORLD DEVELOPMENT REPORT 2009 Table A5 Other indicators--continued Terrain characteristics Geography and people Population living Population living Population Population at less than 25 kms at less than living at less than living at less than plains lowlands plateaus hills mountains from an international 75 kms from an 25 kms from a 75 kms from a (% of total (% of total (% of total (% of total (% of total border international border coastline coastline land area) land area) land area) land area) land area) (%) (%) (%) (%) Economy 2007 2007 2007 2007 2007 2000 2000 2000 2000 Germany 3.1 45.2 4.6 34.3 12.8 15.2 52.0 7.3 12.6 Ghana 15.2 41.8 24.0 19.0 0.0 14.7 33.3 25.9 39.9 Greece 0.0 7.4 0.0 50.6 41.9 7.1 27.2 81.4 96.4 Greenland 0.3 3.0 33.6 8.1 54.9 0.0 0.0 100.0 100.0 Grenada 0.2 99.8 0.0 0.0 0.0 0.0 0.0 100.0 100.0 Guam 0.0 100.0 0.0 0.0 0.0 0.0 0.0 100.0 100.0 Guatemala 0.3 19.1 7.6 22.8 50.2 18.0 57.2 5.6 28.5 Guinea 5.9 11.9 23.7 20.9 37.6 25.0 63.3 19.0 26.4 Guinea-Bissau 71.9 24.6 0.0 3.5 0.0 26.0 96.5 60.9 86.5 Guyana 9.7 49.6 16.8 18.7 5.1 10.2 23.3 56.9 81.5 Haiti 0.0 0.8 0.0 67.1 32.2 16.3 63.3 80.9 99.8 Honduras 2.1 12.4 0.0 40.0 45.4 19.8 77.6 22.0 46.1 Hong Kong, China 0.0 100.0 0.0 0.0 0.0 49.4 100.0 100.0 100.0 Hungary 15.8 66.6 0.0 16.4 1.2 30.5 91.4 0.0 0.0 Iceland 0.0 9.6 0.0 43.2 47.2 0.0 0.0 99.2 100.0 India 13.2 17.8 27.4 22.8 18.7 5.6 16.1 10.3 19.7 Indonesia 13.3 30.5 0.1 32.3 23.8 0.1 0.7 52.2 90.4 Iran, Islamic Rep. 1.1 4.3 0.3 4.9 89.4 6.0 21.3 2.2 5.5 Iraq 19.2 27.2 35.1 12.0 6.4 9.2 28.7 0.7 5.1 Ireland 0.0 100.0 0.0 0.0 0.0 7.5 31.4 73.4 98.0 Isle of Man 0.0 100.0 0.0 0.0 0.0 0.0 0.0 100.0 100.0 Israel 0.0 31.1 0.0 49.1 19.8 90.6 100.0 66.4 95.2 Italy 0.0 14.5 0.0 49.8 35.7 6.8 28.4 45.2 68.3 Jamaica 0.0 23.3 0.0 76.7 0.0 0.0 0.0 91.3 100.0 Japan 0.1 17.3 0.0 64.6 18.0 0.0 0.0 69.3 95.3 Jordan 0.0 0.0 53.3 9.8 36.9 33.6 100.0 1.5 1.8 Kazakhstan 22.7 22.4 27.4 7.9 19.6 14.2 45.8 0.0 0.0 Kenya 9.4 10.3 10.7 3.3 66.4 8.5 32.0 6.1 7.5 Kiribati 100.0 0.0 0.0 0.0 0.0 0.0 0.0 100.0 100.0 Korea, Dem. Rep. 0.0 19.4 0.0 34.5 46.1 12.8 28.9 43.9 89.2 Korea, Rep. 0.0 35.2 0.0 52.5 12.3 1.3 37.3 62.3 70.7 Kuwait 7.9 70.9 21.2 0.0 0.0 13.7 100.0 88.5 97.7 Kyrgyz Republic 0.0 0.0 0.2 0.0 99.8 55.3 94.9 0.0 0.0 Lao PDR 0.0 11.0 0.0 25.3 63.7 49.0 91.3 0.0 0.0 Latvia 8.5 91.5 0.0 0.0 0.0 19.4 94.3 50.0 73.5 Lebanon 0.0 0.0 0.0 0.0 100.0 35.0 100.0 84.3 100.0 Lesotho 0.0 0.0 0.0 0.0 100.0 60.5 100.0 0.0 0.0 Liberia 2.1 45.9 31.8 17.1 3.2 28.2 55.2 39.7 54.7 Libya 23.5 11.4 52.8 2.6 9.7 1.2 4.8 73.6 86.5 Liechtenstein 0.0 0.0 0.0 0.0 100.0 100.0 100.0 0.0 0.0 Lithuania 3.5 96.5 0.0 0.0 0.0 28.7 87.1 9.1 14.8 Luxembourg 0.0 0.0 0.0 100.0 0.0 100.0 100.0 0.0 0.0 Macao, China 0.0 100.0 0.0 0.0 0.0 100.0 100.0 100.0 100.0 Macedonia, FYR 0.0 0.0 0.0 0.0 100.0 67.5 100.0 0.0 0.0 Madagascar 0.3 20.2 3.8 31.0 44.7 0.0 0.0 23.2 45.0 Malawi 0.0 0.3 0.0 6.0 93.7 34.3 98.8 0.0 0.0 Malaysia 2.9 48.2 0.0 29.5 19.5 8.2 25.1 59.0 95.6 Maldives 99.0 0.0 0.0 0.0 0.0 0.0 0.0 100.0 100.0 Mali 59.3 2.4 36.6 1.8 0.0 12.6 38.1 0.0 0.0 Malta 0.0 100.0 0.0 0.0 0.0 0.0 0.0 100.0 100.0 Marshall Islands 96.2 0.0 0.0 0.0 0.0 0.0 0.0 100.0 100.0 Mauritania 56.0 13.4 29.6 1.1 0.0 22.4 46.8 27.0 32.6 Mauritius 0.0 5.3 0.0 94.2 0.0 0.0 0.0 100.0 100.0 Mayotte 0.0 100.0 0.0 0.0 0.0 0.0 0.0 100.0 100.0 Mexico 5.1 12.9 5.4 14.7 62.0 5.8 8.0 11.2 23.5 Micronesia, Fed. Sts. 0.0 100.0 0.0 0.0 0.0 0.0 0.0 100.0 100.0 Moldova 0.0 85.4 14.6 0.0 0.0 52.7 100.0 0.0 0.0 Monaco 0.0 0.0 0.0 0.0 100.0 100.0 100.0 100.0 100.0 Mongolia 0.0 0.0 14.3 0.0 85.7 5.8 18.1 0.0 0.0 Montenegro 0.0 0.0 0.0 1.3 98.7 Morocco 1.3 4.8 6.7 17.8 69.5 4.6 13.2 39.5 59.7 Mozambique 6.5 30.2 6.6 28.4 28.3 8.9 34.6 32.7 52.1 Myanmar 2.8 18.8 0.0 31.7 46.6 3.5 15.2 26.0 42.7 Namibia 22.4 0.9 26.4 3.0 47.4 26.8 54.1 6.0 6.7 Nepal 0.0 5.5 0.0 19.5 75.0 44.1 95.2 0.0 0.0 Netherlands 40.0 58.3 0.0 1.8 0.0 33.2 85.6 53.8 88.6 Netherlands Antilles 74.1 25.9 0.0 0.0 0.0 0.0 0.0 100.0 100.0 New Caledonia 10.9 4.7 0.0 84.5 0.0 0.0 0.0 100.0 100.0 New Zealand 0.0 19.1 0.0 28.3 52.6 0.0 0.0 87.0 98.3 Nicaragua 7.9 53.2 0.0 23.5 15.4 10.6 39.1 17.3 63.2 Selected indicators 345 Terrain characteristics Geography and people Population living Population living Population Population at less than 25 kms at less than living at less than living at less than plains lowlands plateaus hills mountains from an international 75 kms from an 25 kms from a 75 kms from a (% of total (% of total (% of total (% of total (% of total border international border coastline coastline land area) land area) land area) land area) land area) (%) (%) (%) (%) Economy 2007 2007 2007 2007 2007 2000 2000 2000 2000 Niger 50.2 0.0 41.8 0.0 7.9 21.5 55.5 0.0 0.0 Nigeria 16.1 21.8 35.4 17.5 9.2 5.3 24.4 12.3 20.9 Northern Mariana Islands 0.0 100.0 0.0 0.0 0.0 0.0 0.0 100.0 100.0 Norway 0.0 9.8 0.0 42.7 47.5 5.2 48.7 82.3 92.4 Oman 31.3 17.9 14.4 20.0 16.3 4.7 15.1 66.5 81.5 Pakistan 21.9 10.2 0.9 15.6 51.4 13.8 42.2 7.1 8.3 Palau 0.0 100.0 0.0 0.0 0.0 0.0 0.0 100.0 100.0 Panama 0.2 30.8 0.0 53.6 15.4 8.6 17.2 78.8 100.0 Papua New Guinea 13.1 17.1 0.0 25.9 43.9 1.0 3.4 35.8 53.3 Paraguay 51.3 22.3 23.6 2.8 0.0 53.0 73.1 0.0 0.0 Peru 24.5 4.2 8.5 11.0 51.9 2.5 11.8 39.0 50.4 Philippines 0.1 26.4 0.0 56.4 17.0 0.0 0.0 78.3 98.9 Poland 10.8 63.5 12.2 10.4 3.2 12.8 44.0 5.9 11.2 Portugal 0.0 37.4 0.0 36.4 26.2 6.4 32.4 65.1 89.8 Puerto Rico 0.0 38.9 0.0 61.1 0.0 0.0 0.0 94.1 100.0 Qatar 79.2 20.8 0.0 0.0 0.0 0.4 18.6 94.0 100.0 Romania 0.9 37.9 0.6 21.5 39.1 20.1 61.5 3.4 5.1 Russian Federation 10.0 36.0 6.4 22.1 25.5 3.5 13.0 8.0 10.7 Rwanda 0.0 0.0 15.5 0.0 84.5 57.8 100.0 0.0 0.0 Samoa 0.0 8.0 0.0 92.0 0.0 0.0 0.0 San Marino 0.0 0.0 0.0 100.0 0.0 100.0 100.0 0.0 0.0 Sao Tome and Princip 0.0 0.6 31.2 35.9 32.3 0.0 0.0 100.0 100.0 Saudi Arabia 0.0 16.4 0.0 83.6 0.0 2.8 9.0 27.0 38.3 Senegal 19.7 5.2 48.0 3.4 23.6 20.2 39.9 48.6 74.4 Serbia 84.2 10.0 2.2 3.6 0.0 .. .. .. .. Seychelles 1.4 18.8 0.0 39.3 40.5 0.0 0.0 100.0 100.0 Sierra Leone 35.4 64.6 0.0 0.0 0.0 15.0 68.1 35.0 55.3 Singapore 0.2 53.3 2.8 43.7 0.0 .. .. 100.0 100.0 Slovak Republic 0.0 100.0 0.0 0.0 0.0 57.0 100.0 0.0 0.0 Slovenia 0.0 5.2 0.0 67.5 27.2 68.4 100.0 4.3 14.0 Solomon Islands 0.0 0.0 0.0 49.4 50.6 0.0 0.0 100.0 100.0 Somalia 0.1 51.4 0.0 48.4 0.0 8.0 25.2 30.5 52.7 South Africa 13.1 19.2 36.1 11.6 20.1 5.0 14.8 23.4 35.9 Spain 1.8 3.7 29.1 7.4 58.0 4.7 15.3 48.1 63.5 Sri Lanka 0.0 2.8 3.7 31.6 61.8 0.0 0.0 47.3 88.0 St. Kitts and Nevis 20.3 43.6 0.0 36.0 0.0 0.0 0.0 100.0 100.0 St. Lucia 0.0 100.0 0.0 0.0 0.0 0.0 0.0 100.0 100.0 St. Vincent and the 0.0 71.5 0.0 28.5 0.0 0.0 0.0 100.0 100.0 Grenadines Sudan 0.0 0.0 0.0 100.0 0.0 5.9 14.2 1.8 2.6 Suriname 29.7 0.3 51.1 3.1 15.9 3.4 14.1 86.8 97.3 Swaziland 9.5 65.2 16.8 8.5 0.0 60.5 100.0 0.0 0.0 Sweden 0.0 5.3 0.0 8.0 86.7 1.0 7.1 63.9 82.9 Switzerland 2.3 37.8 4.4 34.4 21.1 63.8 100.0 0.0 0.0 Syrian Arab Republic 0.0 0.0 0.0 0.5 99.5 25.3 85.8 11.3 28.5 Taiwan, China 0.0 5.1 0.0 12.6 82.3 0.0 0.0 71.4 100.0 Tajikistan 0.0 0.0 0.3 0.7 99.0 49.9 96.2 0.0 0.0 Tanzania 1.5 5.1 29.2 7.2 57.1 10.9 32.3 13.6 17.3 Thailand 5.0 41.1 2.3 38.4 13.2 10.5 34.4 18.2 37.6 Timor-Leste .. .. .. .. .. .. .. .. .. Togo 0.0 58.2 14.8 27.0 0.0 72.5 100.0 26.8 38.6 Tonga 89.3 10.7 0.0 0.0 0.0 0.0 0.0 100.0 100.0 Trinidad and Tobago 0.0 100.0 0.0 0.0 0.0 0.0 0.0 100.0 100.0 Tunisia 1.2 40.5 17.4 36.4 4.4 6.6 22.4 55.8 79.4 Turkey 0.0 4.3 0.6 9.1 86.0 4.2 12.7 37.8 52.7 Turkmenistan 36.5 34.6 11.1 8.1 9.7 27.4 54.9 0.0 0.0 Uganda 10.2 0.0 40.8 0.0 49.1 21.3 53.1 0.0 0.0 Ukraine 23.0 51.1 18.1 4.3 3.5 10.1 41.1 11.2 16.4 United Arab Emirates 29.8 50.7 0.0 19.5 0.0 23.3 83.7 54.8 84.4 United Kingdom 2.3 67.4 0.0 30.3 0.0 0.7 2.1 45.5 87.6 United States 12.6 13.1 24.6 12.8 36.9 3.1 6.9 28.9 41.4 Uruguay 17.4 82.5 0.0 0.0 0.0 10.7 17.5 65.7 77.4 Uzbekistan 23.5 27.2 12.4 17.6 19.4 55.8 86.6 0.0 0.0 Vanuatu 0.3 7.9 1.4 90.4 0.0 0.0 0.0 100.0 100.0 Venezuela, RB 11.0 31.9 4.0 27.2 25.9 3.7 9.8 35.4 64.8 Vietnam 8.4 25.3 0.0 30.5 35.9 9.9 44.7 42.0 73.2 Virgin Islands (U.S.) 0.0 100.0 0.0 0.0 0.0 0.0 0.0 100.0 100.0 West Bank and Gaza 0.0 10.1 0.0 89.9 0.0 100.0 100.0 34.8 66.3 Yemen, Rep. 2.5 1.8 21.5 3.7 70.4 2.0 7.9 13.3 43.7 Zambia 20.3 0.0 46.4 0.0 33.3 22.7 52.6 0.0 0.0 Zimbabwe 2.2 0.0 35.9 0.0 61.9 11.7 28.8 0.0 0.0 346 WORLD DEVELOPMENT REPORT 2009 Sources and Definitions Table A1 Geography and access Column Source Notes 1 Location CIA Factbook This entry includes rounded latitude and longitude figures for the purpose of finding the approximate geographic center of an entity and is based on the Gazetteer of Conventional Names, Third Edition, August 1988, US Board on Geographic Names and on other sources. 2 Surface area (sq. km) World Bank Surface area is a country's total area, including areas under inland bodies of water and some coastal waterways. Food and Agriculture Organization, Production Yearbook and data files. 3 Arable land (% of land area) World Bank Arable land includes land defined by the FAO as land under temporary crops (double-cropped areas are counted once), temporary meadows for mowing or for pasture, land under market or kitchen gardens, and land temporarily fallow. 4 Forest land (% of land area) World Bank Forest area is land under natural or planted stands of trees, whether productive or not. 5 Coastline (kms) CIA Factbook Total length of the boundary between the land area (including islands) and the sea. 6 Land boundaries (kms) CIA Factbook Total length of all land boundaries and the individual lengths for each of the contiguous border countries. When available, official lengths published by national statistical agencies are used. Because surveying methods may differ, country border lengths reported by contiguous countries may differ. 7 Airports with paved runway CIA Factbook Total number of airports with paved runways (concrete or asphalt surfaces) by length. For airports (number) with more than one runway, only the longest runway is included according to the following five groups - (1) over 3,047 m, (2) 2,438 to 3,047 m, (3) 1,524 to 2,437 m, (4) 914 to 1,523 m, and (5) under 914 m. Only airports with usable runways are included in this listing. Not all airports have facilities for refueling, maintenance, or air traffic control. 8 Ports and terminals (number) CIA Factbook Number of ports 9 Rail density (rail km per 100 km2) CIA Factbook Total route length of the railway network and of its component parts by gauge (broad, standard, narrow, and dual) divided by the country size (in % terms). 10 Road density (road km per 100 km2) WB and CIA Total length of the road network and includes the length of the paved and unpaved portions divided Factbook by the country size (in % terms). 11 National average distance to capital WDR 2009 Team Population weighted average distance to the capital city city (kms) Table A2 Urbanization Column Source Notes 12 Agglomeration index WDR 2009 Team See chapter 1 13 % of urban population to total UN Urban population as a % of total population in 2000 population (2000) 14 % of urban population to total UN Urban population as a % of total population in 2005 population (2005) 15 % of urban population to total UN Urban population as a % of total population in 2015 population (2015) 16 Population density (number of World Bank Number of people per km2 people per km2) 17 Pop. in cities > 1 million (% of total World Bank Proportion of the population living in cities bigger than 1 million people pop.) 18 % of Population in largest city (% of World Bank Proportion of the population living in the largest city (as % of the urban population) urban pop.) 19 % of urban population with water World Bank Percentage of the urban population with reasonable access to an adequate amount of water from access an improved source, such as a household connection, public standpipe, borehole, protected well or spring, and rainwater collection. Reasonable access is defined as the availability of at least 20 liters a person a day from a source within one kilometer of the dwelling. 20 % of rural population with water World Bank Percentage of the urban population with reasonable access to an adequate amount of water from access an improved source, such as a household connection, public standpipe, borehole, protected well or spring, and rainwater collection. Reasonable access is defined as the availability of at least 20 liters a person a day from a source within one kilometer of the dwelling. 21 % of urban population with World Bank Percentage of the urban population with at least adequate access to excreta disposal facilities that sanitation services can effectively prevent human, animal, and insect contact with excreta. Improved facilities range from simple but protected pit latrines to flush toilets with a sewerage connection. To be effective, facilities must be correctly constructed and properly maintained. 22 % of rural population with sanitation World Bank Percentage of the rural population with at least adequate access to excreta disposal facilities that services can effectively prevent human, animal, and insect contact with excreta. Improved facilities range from simple but protected pit latrines to flush toilets with a sewerage connection. To be effective, facilities must be correctly constructed and properly maintained. Selected indicators 347 Table A3 Territorial development Column Source Notes 23 Leading area: Area name WDR 2009 Team Name of the leading area. Leading area is defined as the area with the highest measure of welfare (income or consumption or gdp) per capita 24 Leading area: Poverty Incidence WDR 2009 Team The poverty incidence is the proportion of the population living below the poverty line 25 Leading area: Poor in area as % of WDR 2009 Team Proportion of the country's poor residing in the leading area total country poor 26 Leading area: Welfare measure in WDR 2009 Team Leading measure of welfare (income, consumption or gdp) per capita, relative to the national area (as a % of country's average measure of welfare (income, consumption or gdp) per capita welfare measure ) 27 Leading area: Area (km2) Statoids Surface area of the leading area 28 Leading area: Population density Statoids People per km2 in the leading area (number of people per km2) 29 Lagging area: Area name WDR 2009 Team Name of the lagging area. Lagging area is defined as the area with the lowest measure of welfare (income or consumption or gdp) per capita 30 Lagging area: Poverty Incidence WDR 2009 Team The poverty incidence is the proportion of the population living below the poverty line 31 Lagging area: Poor in area as % of WDR 2009 Team Proportion of the country's poor residing in the lagging area total country poor 32 Lagging area: Welfare measure in WDR 2009 Team Lagging measure of welfare (income, consumption or gdp) per capita, relative to the national area (as a % of country's average measure of welfare (income, consumption or gdp) per capita welfare measure ) 33 Lagging area: Area (km2) Statoids Surface area of the lagging area 34 Lagging area: Population density Statoids People per km2 in the leading area (number of people per km2) Table A4 International integration Column Source Notes 35 Countries that need a visa to visit Neumayer (2005) Number of countries for which their nationals need a visa to visit this country. this country 36 Countries for which this country's Neumayer (2005) Number of countries for which the nationals of this country need a visa. residents need a visa 37 Cost of obtaining a passport relative McKenzie (2005) The price of a standard passport standardized using the GNI per capita to GDP per capita 38 International migration stock (% of World Bank Migration stock is the number of people born in a country other than that in which they live. It also foreigners) includes refugees 39 International voice traffic (incoming World Bank International voice traffic is the sum of international incoming and outgoing telephone traffic (in and outgoing, minutes per person) minutes). 40 International internet bandwith (bits World Bank International Internet bandwidth is the contracted capacity of international connections between per person) countries for transmitting Internet traffic. 41 Telephone average cost of call to World Bank Cost of international call to U.S. is the cost of a three-minute, peak rate, fixed line call from the the US (US$ per three minutes) country to the United States. 42 Total trade as share of GDP World Bank Trade is the sum of exports and imports of goods and services measured as a share of gross domestic product. 43 Index of shipping difficulties WB Doing Indicator of shipping difficulties (border delays, read carpet, fees, red tape, etc) Business 44 Average tariffs and custom duties World Bank Average tariffs (% of import value) 45 Share of trade with neighboring WDR 2009 Team The proportion of trade with neighboring countries as a proportion of the total trade of the country. countries (% of total trade) Calculation using COMTRADE data. 348 WORLD DEVELOPMENT REPORT 2009 Table A5 Other indicators Column Source Notes 46 % of terrain: plains Nelson (2007) % of terrain: plains 47 % of terrain: lowlands Nelson (2007) % of terrain: lowlands 48 % of terrain: plateaus Nelson (2007) % of terrain: plateaus 49 % of terrain: hills Nelson (2007) % of terrain: hills 50 % of terrain: mountains Nelson (2007) % of terrain: mountains 51 Population living less than 25 kms WDR 2009 Team Proportion of the total population living in less than 25 kms from an international border from an international border 52 Population living less than 75 kms WDR 2009 Team Proportion of the total population living in less than 75 kms from an international border from an international border 53 Population living less than 25 kms WDR 2009 Team Proportion of the total population living in less than 25 kms from a coastline from a coastline 54 Population living less than 75 kms WDR 2009 Team Proportion of the total population living in less than 75 kms from a coastline from a coastline Selected World Development Indicators I n this year's edition, development data International Monetary Fund (IMF), and are presented in five tables showing the the Organisation for Economic Co-opera- comparative socioeconomic data for tion and Development (OECD) (see the more than 130 economies for the most Data Sources following the Technical notes recent year for which data are available and, for a complete listing). Although interna- for some indicators, for an earlier year. An tional standards of coverage, definition, additional table presents basic indicators and classification apply to most statistics for 77 economies with sparse data or with reported by countries and international populations of less than 3 million. agencies, there are inevitably differences in The indicators presented here are a selec- timeliness and reliability arising from dif- tion from more than 800 included in World ferences in the capabilities and resources Development Indicators 2008. Published devoted to basic data collection and compi- annually, World Development Indicators lation. For some topics, competing sources (WDI) reflects a comprehensive view of the of data require review by World Bank staff development process. The WDI's six sec- to ensure that the most reliable data avail- tions recognize the contribution of a wide able are presented. In some instances, where range of factors: progress on the Millennium available data are deemed too weak to pro- Development Goals and human capital vide reliable measures of levels and trends development, environmental sustainability, or do not adequately adhere to international macroeconomic performance, private sec- standards, the data are not shown. tor development and the investment climate, The data presented are generally con- and the global links that influence the exter- sistent with those in World Development nal environment for development. Indicators 2008. However, data have been World Development Indicators is comple- revised and updated wherever new infor- mented by a separately published database mation has become available. Differences that gives access to more than 1,000 data may also reflect revisions to historical series tables and 800 time-series indicators for and changes in methodology. Thus data 222 economies and regions. This database of different vintages may be published in is available through an electronic subscrip- different editions of World Bank publica- tion (WDI Online) or as a CD-ROM. tions. Readers are advised not to compile data series from different publications or Data sources and methodology different editions of the same publication. Socioeconomic and environmental data Consistent time-series data are available presented here are drawn from several on World Development Indicators 2008 CD- sources: primary data collected by the ROM and through WDI Online. World Bank, member country statistical All dollar figures are in current U.S. dollars publications, research institutes, and inter- unless otherwise stated. The various methods national organizations such as the United usedtoconvertfromnationalcurrencyfigures Nations and its specialized agencies, the are described in the Technical notes. 349 350 WORLD DEVELOPMENT REPORT 2009 Because the World Bank's primary busi- third or more of the overall estimate, how- ness is providing lending and policy advice ever, the group measure is reported as not to its low- and middle-income members, the available. The section on Statistical methods issues covered in these tables focus mainly in the Technical notes provides further infor- on these economies. Where available, infor- mation on aggregation methods. Weights mation on the high-income economies is used to construct the aggregates are listed also provided for comparison. Readers may in the technical notes for each table. wish to refer to national statistical publica- tions and publications of the OECD and the Terminology and country coverage European Union for more information on The term country does not imply political the high-income economies independence but may refer to any territory for which authorities report separate social or Classification of economies economic statistics. Data are shown for econ- and summary measures omies as they were constituted in 2007, and The summary measures at the bottom of historical data are revised to reflect current most tables include economies classified by political arrangements. Throughout the income per capita and by region. Gross tables, exceptions are noted. Unless other- national income (GNI) per capita is used to wise noted, data for China do not include determine the following income classifica- data for Hong Kong, China; Macao, China; tions: low-income, $935 or less in 2007; or Taiwan, China. Data for Indonesia include middle-income, $936 to $11,455; and high- Timor-Leste through 1999 unless otherwise income, $11,456 and above. A further divi- noted. Montenegro declared independence sion at GNI per capita $3,705 is made between from Serbia and Montenegro on June 3, 2006. lower-middle-income and upper-middle- When available, data for each country are income economies. The classification of shown separately. However, some indicators economies based on per capita income occurs for Serbia continue to include data for Monte- annually, so the country composition of the negro through 2005; these data are footnoted in income groups may change annually. When the tables. Moreover, data for most indicators these changes in classification are made based from 1999 onward for Serbia exclude data for on the most recent estimates, aggregates Kosovo, a territory within Serbia that is cur- based on the new income classifications are rently under international administration recalculated for all past periods to ensure that pursuant to UN Security Council Resolution a consistent time series is maintained. See the 1244 (1999); any exceptions are noted. table on classification of economies at the end of this volume for a list of economies in each Technical notes group (including those with populations of Because data quality and intercountry com- less than 3 million). parisons are often problematic, readers are Summary measures are either totals encouraged to consult the Technical notes, (indicated by t if the aggregates include esti- the table on Classification of Economies by mates for missing data and nonreporting Region and Income, and the footnotes to the countries, or by an s for simple sums of the tables. For more extensive documentation, data available), weighted averages (w), or see World Development Indicators 2008. median values (m) calculated for groups of Readers may find more information on economies. Data for the countries excluded the WDI 2008, and orders can be made from the main tables (those presented in online, by phone, or fax as follows: table 5) have been included in the sum- For more information and to order mary measures, where data are available, online: http://www.worldbank.org/data/ or by assuming that they follow the trend of wdi2006/index.htm. reporting countries. This gives a more con- To order by phone or fax: 1-800-645- sistent aggregated measure by standardizing 7247 or 703-661-1580; Fax 703-661-1501. country coverage for each period shown. To order by mail: The World Bank, P.O. Where missing information accounts for a Box 960, Herndon, VA 20172-0960, U.S.A. Selected world development indicators 351 Classification of economies by region and income, FY2009 East Asia and the Pacific Latin America and the Caribbean South Asia High income OECD American Samoa UMC Argentina UMC Afghanistan LIC Australia Cambodia LIC Belize UMC Bangladesh LIC Austria China LMC Bolivia LMC Bhutan LMC Belgium Fiji UMC Brazil UMC India LMC Canada Indonesia LMC Chile UMC Maldives LMC Czech Republic Kiribati LMC Colombia LMC Nepal LIC Denmark Korea, Dem. People's Rep. of LIC Costa Rica UMC Pakistan LIC Finland Lao PDR LIC Cuba UMC Sri Lanka LMC France Malaysia UMC Dominica UMC Germany Marshall Islands LMC Dominican Republic LMC Greece Micronesia, Fed. States of LMC Ecuador LMC Sub-Saharan Africa Hungary Mongolia LMC El Salvador LMC Angola LMC Iceland Myanmar LIC Grenada UMC Benin LIC Ireland Palau UMC Guatemala LMC Botswana UMC Italy Papua New Guinea LIC Guyana LMC Burkina Faso LIC Japan Philippines LMC Haiti LIC Burundi LIC Korea, Rep. of Samoa LMC Honduras LMC Cameroon LMC Luxembourg Solomon Islands LIC Jamaica UMC Cape Verde LMC Netherlands Thailand LMC Mexico UMC Central African Republic LIC New Zealand Timor-Leste LMC Nicaragua LMC Chad LIC Norway Tonga LMC Panama UMC Comoros LIC Portugal Vanuatu LMC Paraguay LMC Congo, Dem. Rep. of LIC Slovak Republic Vietnam LIC Peru LMC Congo, Rep. of LMC Spain St. Kitts and Nevis UMC Côte d'Ivoire LIC Sweden St. Lucia UMC Eritrea LIC Switzerland Europe and Central Asia St. Vincent and the Grenadines UMC Ethiopia LIC United Kingdom Albania LMC Suriname UMC Gabon UMC United States Armenia LMC Uruguay UMC Gambia, The LIC Azerbaijan LMC Venezuela, R.B. UMC Ghana LIC Belarus UMC Guinea LIC Other high income Bosnia and Herzegovina LMC Guinea-Bissau LIC Andorra Bulgaria UMC Middle East and North Africa Kenya LIC Antigua and Barbuda Croatia UMC Algeria LMC Lesotho LMC Aruba Georgia LMC Djibouti LMC Liberia LIC Bahamas, The Kazakhstan UMC Egypt, Arab Rep. of LMC Madagascar LIC Bahrain Kyrgyz Republic LIC Iran, Islamic Rep. of LMC Malawi LIC Barbados Latvia UMC Iraq LMC Mali LIC Bermuda Lithuania UMC Jordan LMC Mauritania LIC Brunei Darussalam Macedonia, FYR LMC Lebanon UMC Mauritius UMC Cayman Islands Moldova LMC Libya UMC Mayotte UMC Channel Islands Montenegro UMC Morocco LMC Mozambique LIC Cyprus Poland UMC Syrian Arab Republic LMC Namibia LMC Equatorial Guinea Romania UMC Tunisia LMC Niger LIC Estonia Russian Federation UMC West Bank and Gaza LMC Nigeria LIC Faeroe Islands Serbia UMC Yemen, Rep. of LIC Rwanda LIC French Polynesia Tajikistan LIC São Tomé and Príncipe LIC Greenland Turkey UMC Senegal LIC Guam Turkmenistan LMC Seychelles UMC Hong Kong, China Ukraine LMC Sierra Leone LIC Isle of Man Uzbekistan LIC Somalia LIC Israel South Africa UMC Kuwait Sudan LMC Liechtenstein Swaziland LMC Macao, China Tanzania LIC Malta Togo LIC Monaco Uganda LIC Netherlands Antilles Zambia LIC New Caledonia Zimbabwe LIC Northern Mariana Islands Oman Puerto Rico Qatar San Marino Saudi Arabia Singapore Slovenia Taiwan, China Trinidad and Tobago United Arab Emirates Virgin Islands (U.S.) This table classifies all World Bank member economies and all other economies with populations of more than 30,000. Economies are divided among income groups according to 2007 GNI per capita, calculated using the World Bank Atlas method. The groups are low income (LIC), $935 or less; lower middle income (LMC), $936­3,705; upper middle income (UMC), $3,706­11,455; and high income, $11,456 or more. Source: World Bank data. 352 WORLD DEVELOPMENT REPORT 2009 Table 1 Key indicators of development Life expectancy at Population Population GNIa PPP GNIb Gross birth Adult Carbon age domestic literacy dioxide Average composition product rate % emissions annual % Density % $ per $ per per capita Male Female ages 15 per capita Millions growth people per ages 0­14 $ billions capita $ billions capita % growth years years and older metric tons 2007 2000­07 sq. km 2007 2007 2007 2007 2007 2007 2006­07 2006 2006 2005 2004 Afghanistan .. .. .. .. 10.1 ..c 24.9d ..d .. .. .. .. 0.0 Albania 3 0.5 116 25 10.5 3,290 20.9 6,580 5.7 73 80 99 1.2 Algeria 34 1.5 14 28 122.5 3,620 185.8d 5,490d 1.6 71 73 70 6.0 Angola 17 2.9 14 46 43.6 2,560 74.9 4,400 20.1 41 44 67 0.5 Argentina 40 1.0 14 26 238.9 6,050 513.0 12,990 7.6 71 79 97 3.7 Armenia 3 ­0.4 106 19 7.9 2,640 17.7 5,900 14.0 68 75 99 1.2 Australia 21 1.3 3 19 755.8 35,960 700.6 33,340 2.9 79 83 .. 16.2 Austria 8 0.5 101 15 355.1 42,700 316.8 38,090 3.0 77 83 .. 8.5 Azerbaijan 9 0.9 104 23 21.9 2,550 54.6 6,370 18.0 70 75 .. 3.8 Bangladesh 159 1.8 1,218 34 75.1 470 212.7 1,340 4.8 63 65 47 0.3 Belarus 10 ­0.4 47 15 40.9 4,220 104.2 10,740 8.5 63 74 .. 6.6 Belgium 11 0.5 352 17 432.5 40,710 373.1 35,110 2.1 77 82 .. 9.7 Benin 9 3.2 82 44 5.1 570 11.9 1,310 1.5 55 57 35 0.3 Bolivia 10 1.9 9 37 12.0 1,260 39.4 4,140 2.8 63 67 87 0.8 Bosnia and Herzegovina 4 0.3 74 17 14.3 3,790 28.6 7,280 6.8 72 77 97 4.0 Brazil 192 1.4 23 27 1,133.0 5,910 1,795.7 9,370 4.2 69 76 89 1.8 Bulgaria 8 ­0.8 70 13 35.1 4,590 85.4 11,180 6.9 69 76 98 5.5 Burkina Faso 15 3.1 54 46 6.4 430 16.5 1,120 1.0 50 53 24 0.1 Burundi 8 3.5 331 44 0.9 110 2.8 330 ­0.3 48 50 59 0.0 Cambodia 14 1.8 82 36 7.9 540 24.5 1,690 8.4 57 61 74 0.0 Cameroon 19 2.2 40 41 19.5 1,050 39.2 2,120 1.3 50 51 68 0.2 Canada 33 1.0 4 17 1,300.0 39,420 1,164.2 35,310 1.7 78 83 .. 20.0 Central African Republic 4 1.7 7 42 1.7 380 3.2 740 2.3 43 46 49 0.1 Chad 11 3.4 9 46 5.8 540 13.8 1,280 ­2.1 49 52 26 0.0 Chile 17 1.1 22 24 138.6 8,350 209.0 12,590 4.1 75 81 96 3.9 China 1,320 0.6 142 21 3,120.9 2,360 7,083.5 5,370 11.2 70 74 91 3.9 Hong Kong, China 7 0.5 6,647 14 218.9 31,610 305.1 44,050 5.3 79 85 .. 5.5 Colombia 46 1.4 42 29 149.9 3,250 306.2 6,640 6.2 69 76 93 1.2 Congo, Dem. Rep. of 62 3.0 28 47 8.6 140 17.9 290 3.5 45 47 67 0.0 Congo, Rep. of 4 2.3 11 42 5.8 1,540 10.4 2,750 ­3.6 54 56 85 1.0 Costa Rica 4 1.8 87 27 24.8 5,560 37.2d 8,340d 4.8 76 81 95 1.5 Côte d'Ivoire 19 1.7 61 41 17.5 910 30.7 1,590 ­0.1 47 49 49 0.3 Croatia 4 ­0.2 79 15 46.4 10,460 66.8 15,050 5.6 73 79 98 5.3 Czech Republic 10 0.1 134 14 149.4 14,450 225.5 21,820 5.0 73 80 .. 11.5 Denmark 5 0.3 129 19 299.8 54,910 200.6 36,740 1.4 76 80 .. 9.8 Dominican Republic 10 1.6 202 33 34.6 3,550 49.3d 5,050d 7.0 69 75 87 2.1 Ecuador 13 1.2 48 32 41.2 3,080 93.9 7,040 0.9 72 78 91 2.3 Egypt, Arab Rep. of 75 1.8 76 33 119.4 1,580 407.6 5,400 5.2 69 73 71 2.2 El Salvador 7 1.4 331 33 19.5 2,850 33.2d 4,840d 2.8 69 75 81 0.9 Eritrea 5 3.9 48 43 1.1 230 1.9d 400d ­2.3 55 60 .. 0.2 Ethiopia 79 2.6 79 44 17.6 220 61.7 780 8.4 51 54 36 0.1 Finland 5 0.3 17 17 234.8 44,400 186.5 35,270 4.0 76 83 .. 12.6 France 62 0.7 112 18 2,447.1 38,500e 2,065.4 33,470 1.6 77 84 .. 6.2 Georgia 4 ­1.0 63 18 9.3 2,120 21.0 4,770 13.4 67 75 .. 0.9 Germany 82 0.0f 236 14 3,197.0 38,860 2,782.7 33,820 2.6 76 82 .. 9.8 Ghana 23 2.2 103 38 13.9 590 31.2 1,330 4.3 59 60 58 0.3 Greece 11 0.4 87 14 331.7 29,630 364.1 32,520 3.6 77 82 96 8.7 Guatemala 13 2.5 123 43 32.6 2,440 55.0d 4,120d 3.2 66 74 69 1.0 Guinea 9 1.9 38 43 3.7 400 10.5 1,120 ­0.6 54 57 29 0.2 Haiti 10 1.6 349 37 5.4 560 10.1d 1,050d 1.4 59 62 .. 0.2 Honduras 7 1.9 63 39 11.3 1,600 22.4d 3,160d 4.5 66 73 80 1.1 Hungary 10 ­0.2 112 15 116.3 11,570 175.2 17,430 1.5 69 77 .. 5.7 India 1,123 1.4 378 32 1,069.4 950 3,078.7 2,740 7.7 63 66 61 1.2 Indonesia 226 1.3 125 28 373.1 1,650 807.9 3,580 5.1 66 70 90 1.7 Iran, Islamic Rep. of 71 1.5 44 27 246.5 3,470 766.9 10,800 6.2 69 72 82 6.4 Iraq .. .. .. .. .. ..g .. .. .. .. .. .. 3.0 Ireland 4 2.0 63 21 210.2 48,140 161.7 37,040 2.8 77 82 .. 10.4 Israel 7 1.9 331 28 157.1 21,900 186.0 25,930 3.4 78 82 .. 10.5 Italy 59 0.6 202 14 1,991.3 33,540 1,775.3 29,900 0.8 78 84 98 7.7 Japan 128 0.1 351 14 4,813.3 37,670 4,420.6 34,600 2.1 79 86 .. 9.8 Jordan 6 2.5 65 36 16.3 2,850 29.5 5,160 2.6 71 74 91 3.1 Kazakhstan 15 0.6 6 24 78.3 5,060 150.1 9,700 7.3 61 72 .. 13.3 Kenya 38 2.6 66 43 25.6 680 57.8 1,540 4.1 52 55 74 0.3 Korea, Rep. of 49 0.5 492 18 955.8 19,690 1,201.1 24,750 4.7 75 82 .. 9.7 Kyrgyz Republic 5 0.9 27 30 3.1 590 10.2 1,950 6.4 64 72 .. 1.1 Lao PDR 6 1.6 25 38 3.4 580 11.4 1,940 5.3 63 65 69 0.2 Lebanon 4 1.2 401 28 23.7 5,770 41.2 10,050 1.0 70 74 .. 4.1 Liberia 4 2.9 39 47 0.6 150 1.1 290 4.3 44 46 52 0.1 Libya 6 2.0 4 30 55.5 9,010 70.8d 11,500d 4.8 71 77 84 10.3 Lithuania 3 ­0.5 54 16 33.5 9,920 58.0 17,180 9.4 65 77 100 3.9 Madagascar 20 2.8 34 43 6.3 320 18.2 920 3.7 57 61 71 0.2 Malawi 14 2.6 148 47 3.5 250 10.5 750 4.7 47 48 .. 0.1 Malaysia 27 1.9 81 30 173.7 6,540 360.2 13,570 4.0 72 76 89 7.0 Mali 12 3.0 10 48 6.1 500 12.8 1,040 ­0.3 52 56 24 0.1 Mauritania 3 2.8 3 40 2.6 840 6.3 2,010 ­0.6 62 66 51 0.9 Mexico 105 1.0 54 30 878.0 8,340 1,324.6 12,580 2.3 72 77 92 4.3 Moldova 4 ­1.3 115 19 4.3 1,260h 11.1 2,930 4.1 65 72 99 2.0 Morocco 31 1.2 69 29 69.4 2,250 123.3 3,990 1.1 69 73 52 1.4 Mozambique 21 2.3 27 44 6.8 320 14.8 690 5.0 42 43 .. 0.1 Myanmar 49 0.9 74 26 .. ..c .. .. 4.1 59 65 90 0.2 Selected world development indicators 353 Life expectancy at Population Population GNIa PPP GNIb Gross birth Adult Carbon age domestic literacy dioxide Average composition product rate % emissions annual % Density % $ per $ per per capita Male Female ages 15 per capita Millions growth people per ages 0­14 $ billions capita $ billions capita % growth years years and older metric tons 2007 2000­07 sq. km 2007 2007 2007 2007 2007 2007 2006­07 2006 2006 2005 2004 Nepal 28 2.0 197 38 9.7 340 29.2 1,040 0.8 63 64 49 0.1 Netherlands 16 0.4 484 18 750.5 45,820 647.1 39,500 3.3 78 82 .. 8.7 New Zealand 4 1.3 16 21 121.7 28,780 111.4 26,340 2.3 78 82 .. 7.7 Nicaragua 6 1.3 46 37 5.5 980 11.7d 2,080d 2.9 70 76 77 0.7 Niger 14 3.5 11 48 4.0 280 9.0 630 ­0.1 57 56 29 0.1 Nigeria 148 2.4 162 44 137.1 930 262.5 1,770 4.0 46 47 69 0.8 Norway 5 0.7 15 19 360.0 76,450 252.8 53,690 2.4 78 83 .. 19.1 Pakistan 162 2.3 211 36 141.0 870 417.5 2,570 4.2 65 66 50 0.8 Panama 3 1.8 45 30 18.4 5,510 27.9d 8,340d 9.4 73 78 92 1.8 Papua New Guinea 6 2.3 14 40 5.4 850 9.5d 1,500d 4.2 55 60 57 0.4 Paraguay 6 1.9 15 35 10.2 1,670 26.8 4,380 4.6 69 74 93 0.7 Peru 28 1.2 22 31 96.2 3,450 201.9 7,240 7.8 69 74 88 1.2 Philippines 88 2.0 295 35 142.6 1,620 327.8 3,730 5.3 69 74 93 1.0 Poland 38 ­0.1 124 15 374.6 9,840 593.3 15,590 6.7 71 80 .. 8.0 Portugal 11 0.5 116 16 201.1 18,950 219.0 20,640 1.7 75 82 94 5.6 Romania 22 ­0.6 94 15 132.5 6,150 236.6 10,980 6.4 69 76 97 4.2 Russian Federation 142 ­0.5 9 15 1,071.0 7,560 2,039.1 14,400 8.8 59 73 99 10.6 Rwanda 10 2.5 395 43 3.1 320 8.4 860 3.0 44 47 65 0.1 Saudi Arabia 24 2.3 12 34 373.5 15,440 554.3 22,910 1.2 71 75 83 13.7 Senegal 12 2.6 64 42 10.2 820 20.3 1,640 1.9 61 65 39 0.4 Serbia 7i ­0.3i 95i 18 35.0 4,730i 75.5 10,220 6.9 70i 76i .. .. Sierra Leone 6 3.7 82 43 1.5 260 3.9 660 4.6 41 44 35 0.2 Singapore 5 1.9 6,660 18 149.0 32,470 222.7 48,520 3.3 78 82 93 12.5 Slovak Republic 5 0.0f 112 16 63.3 11,730 104.3 19,330 10.3 70 78 .. 6.7 Somalia 9 3.0 14 44 .. ..c .. .. .. 47 49 .. .. South Africa 48 1.1 39 32 274.0 5,760 454.8 9,560 4.4 49 53 .. 9.4 Spain 45 1.6 90 15 1,321.8 29,450 1,351.1 30,110 2.0 78 84 .. 7.7 Sri Lanka 20 0.4 309 23 30.8 1,540 84.0 4,210 6.5 72 78 91 0.6 Sudan 39 2.1 16 40 37.0 960 72.6 1,880 7.8 57 60 61 0.3 Sweden 9 0.4 22 17 421.3 46,060 327.9 35,840 1.8 79 83 .. 5.9 Switzerland 8 0.7 189 16 452.1 59,880 325.3 43,080 2.2 79 84 .. 5.5 Syrian Arab Republic 20 2.7 108 36 35.0 1,760 87.0 4,370 4.0 72 76 81 3.7 Tajikistan 7 1.3 48 38 3.1 460 11.5 1,710 6.2 64 69 99 0.8 Tanzania 40 2.5 46 44 16.3 400j 48.7 1,200 4.5 51 53 69 0.1 Thailand 64 0.7 125 21 217.4 3,400 503.1 7,880 4.1 66 75 93 4.3 Togo 7 2.8 121 43 2.4 360 5.2 800 ­0.5 56 60 53 0.4 Tunisia 10 1.0 66 25 32.8 3,200 73.0 7,130 5.1 72 76 74 2.3 Turkey 74 1.3 96 27 592.9 8,020 893.1 12,090 3.2 69 74 87 3.2 Turkmenistan 5 1.4 11 30 .. ..g 32.1d 6,640d .. 59 67 .. 8.8 Uganda 31 3.2 157 49 10.5 340 28.5 920 2.9 50 51 67 0.1 Ukraine 46 ­0.8 80 14 118.4 2,550 315.9 6,810 8.2 62 74 99 7.0 United Arab Emirates 4 4.2 52 20 .. ..k .. .. 4.4 77 82 89 37.8 United Kingdom 61 0.5 252 18 2,608.5 42,740 2,097.9 34,370 2.3 77 81 .. 9.8 United States 302 0.9 33 20 13,886.5 46,040 13,829.0 45,850 1.5 75 81 .. 20.6 Uruguay 3 0.1 19 23 21.2 6,380 36.6 11,040 7.3 72 80 .. 1.7 Uzbekistan 27 1.2 63 32 19.7 730 45.2d 1,680d 7.9 64 71 .. 5.3 Venezuela, R.B. de 27 1.7 31 31 201.2 7,320 327.5 11,920 6.6 72 77 93 6.6 Vietnam 85 1.3 275 28 67.2 790 216.9 2,550 7.2 68 73 .. 1.2 West Bank and Gaza 4 3.8 643 45 4.5 1,230 .. .. ­4.9 71 74 92 .. Yemen, Rep. 22 3.0 42 45 19.4 870 49.3 2,200 0.6 61 64 54 1.0 Zambia 12 1.9 16 46 9.5 800 14.6 1,220 4.0 41 42 .. 0.2 Zimbabwe 13 0.8 35 38 4.5 340 .. .. ­6.0 43 42 89 0.8 World 6,612s 1.2w 51w 28w 52,621.4t 7,958w 64,904.9t 9,816w 2.6w 66w 70w 82w 4.3w Low income 1,296 2.2 61 39 748.8 578 1,867.5 1,441 4.3 56 58 61 0.6 Middle income 4,260 1.0 57 27 12,234.7 2,872 25,217.1 5,920 6.9 67 71 90 3.2 Lower middle income 3,437 1.1 100 27 6,485.0 1,887 15,509.6 4,512 8.6 67 70 89 2.6 Upper middle income 823 0.7 20 24 5,749.6 6,987 9,732.4 11,827 5.1 67 74 93 5.5 Low and middle income 5,556 1.3 58 29 12,985.9 2,337 27,083.9 4,875 6.5 64 68 79 2.6 East Asia & Pacific 1,914 0.8 121 23 4,173.5 2,180 9,432.2 4,928 9.6 69 73 91 3.3 Europe & Central Asia 445 0.0f 19 19 2,693.7 6,051 4,918.3 11,049 6.7 64 74 97 7.1 Latin America & the 563 1.3 28 29 3,118.0 5,540 5,191.9 9,226 4.5 70 76 90 2.6 Caribbean Middle East & North Africa 313 1.8 36 32 875.6 2,794 2,222.4 7,091 4.0 68 72 73 3.8 South Asia 1,520 1.6 318 33 1,338.6 880 3,856.4 2,536 7.0 63 66 58 1.1 Sub-Saharan Africa 800 2.5 34 43 761.6 952 1,470.7 1,839 3.7 49 52 59 0.9 High income 1,056 0.7 32 18 39,682.1 37,566 38,080.6 36,050 2.0 76 82 99 13.1 a. Calculated using the World Bank Atlas method. b. PPP is purchasing power parity; see Technical notes. c. Estimated to be low income ($935 or less). d.The estimate is based on regression; others are extrapolated from the latest International Comparison Program benchmark estimates. e. The GNI and GNI per capita estimates include the French overseas departments of French Guiana, Guadeloupe, Martinique, and Réunion. f. Less than 0.05. g. Estimated to be lower middle income ($936 to $3,705). h. Excludes data for Transnistria. i. Excludes data for Kosovo and Metohija. j. Data refer to mainland Tanzania only. k. Estimated to be high income ($11.456 or more). 354 WORLD DEVELOPMENT REPORT 2009 Table 2 Millennium Development Goals: eradicating poverty and improving lives Eradicate extreme poverty and hunger Achieve Promote gender Improve maternal health Combat universal equality HIV/AIDS Share of poorest primary Reduce child Contraceptive and other Ratio of girls to quintile in national Prevalence of education mortality Prevalence rate diseases boys enrolllments consumption or child malnutrition Births attended Primary in primary and Under-five % of married HIV prevalence income by skilled health staff % of children completion ratea secondary schoola mortality rate women ages % of population % b under 5 % % per 1,000 % of total 15­49 ages 15­49 1992­2005 1990 2000­07b 1991 2006 1991 2006 1990 2006 1990 2000­07b 1990 2000­07b 2005 Afghanistan .. .. .. .. .. .. .. .. .. .. .. .. .. .. Albania 8.2c .. 17.0 .. 96 96 97 45 17 .. 100 .. 60 0.2 Algeria 7.0c .. 10.2 80 85 83 99 69 38 77 95 47 61 0.1 Angola .. .. 27.5 35 .. .. .. 260 260 .. 45 .. 6 3.7 Argentina 3.1d,e .. 2.3 .. 97 .. 104 29 16 96 99 .. .. 0.6 Armenia 8.5c .. 4.2 .. 91 .. 104 56 24 .. 98 .. 53 0.1 Australia 5.9e .. .. .. .. 101 97 10 6 100 100 .. .. 0.1 Austria 8.6e .. .. .. 103 95 97 10 5 .. .. .. .. 0.3 Azerbaijan 7.4c .. 14.0 .. 92 100 96 105 88 .. 100 .. 55 0.1 Bangladesh 8.8c .. 39.2 49 72 .. 103 149 69 .. 20 31 58 <0.1 Belarus 8.8c .. 1.3 94 95 .. 101 24 13 .. 100 .. 73 0.3 Belgium 8.5e .. .. 79 87 101 98 10 4 .. .. 78 .. 0.3 Benin 7.4c .. 21.5 21 64 49 73 185 148 .. 79 .. 17 1.8 Bolivia 1.5e 8.9 5.9 .. 101 .. 98 125 61 43 67 30 58 0.1 Bosnia and Herzegovina 7.0c .. 1.6 .. .. .. .. 22 15 97 100 .. 36 <0.1 Brazil 2.9e .. 3.7 93 106 .. 103 57 20 72 97 59 .. 0.5 Bulgaria 8.7c .. 1.6 84 98 99 97 19 14 .. 99 .. .. <0.1 Burkina Faso 6.9c .. 35.2 20 31 62 80 206 204 .. 54 .. 17 2.0 Burundi 5.1c .. 38.9 46 36 82 89 190 181 .. 34 .. 9 3.3 Cambodia 6.8c .. 28.4 .. 87 73 89 116 82 .. 44 .. 40 1.6 Cameroon 5.6c .. 15.1 53 52 83 83 139 149 58 63 16 29 5.5g Canada 7.2e .. .. .. .. 99 98 8 6 .. 100 .. .. 0.3 Central African Republic 2.0c .. 21.8 27 24 60 .. 173 175 .. 53 .. 19 10.7 Chad .. .. 33.9 18 31 42 61 201 209 .. 14 .. 3 3.5 Chile 3.8e .. .. .. 95 100 99 21 9 .. 100 56 .. 0.3 China 4.3e .. 6.8 105 .. 87 100 45 24 50 98 71 87 0.1h Hong Kong, China 5.3e .. .. 102 100 103 98 .. .. .. 100 86 .. .. Colombia 2.9e .. 5.1 70 105 108 104 35 21 82 96 66 78 0.6 Congo, Dem. Rep. of .. .. 33.6 46 .. .. .. 205 205 .. 74k 8 21k 3.2 Congo, Rep. .. .. 11.8 54 73 85 90 103 126 .. 86 .. 44 5.3 Costa Rica 4.1e .. .. 79 89 101 102 18 12 98 99 .. 96 0.3 Côte d'Ivoire 5.2c .. 20.2 43 43 65 .. 153 127 .. 57 .. 13 7.1 Croatia 8.8c .. .. .. 92 102 102 12 6 100 100 .. 69 <0.1 Czech Republic 10.3e .. 2.1 .. 94 98 101 13 4 .. 100 78 .. 0.1 Denmark 8.3e .. .. 98 101 101 101 9 5 .. .. 78 .. 0.2 Dominican Republic 4.1e 8.4 4.2 .. 83 .. 104 65 29 93 96 56 61 1.1 Ecuador 3.3c .. 6.2 .. 106 .. 100 57 24 .. 75 53 73 0.3 Egypt, Arab Rep. of 8.9c .. 5.4 .. 98k 81 95 91 35 37 74 38 59 <0.1 El Salvador 2.7e 11.1 6.1 41 88 102 99 60 25 52 92 47 67 0.9 Eritrea .. .. 34.5 .. 49 .. 72 147 74 .. 28 .. 8 2.4 Ethiopia 9.1c .. 34.6 26 46k 68 83k 204 123 .. 6 4 15 1.4i Finland 9.6e .. .. 97 97 109 102 7 4 .. 100 77 .. 0.1 France 7.2e .. .. 104 .. 102 100 9 4 .. .. 81 .. 0.4 Georgia 5.4c .. .. .. 85 98 103 46 32 .. 92 .. 47 0.2 Germany 8.5e .. .. 100 97 99 98 9 4 .. 100 75 .. 0.1 Ghana 5.6c 24.1 18.8 61 71 79 95k 120 120 40 50 13 17 2.3 Greece 6.7e .. .. 99 103 99 98 11 4 .. .. .. .. 0.2 Guatemala 3.9e .. 17.7 .. 77 .. 92 82 41 .. 41 .. 43 0.9 Guinea 7.0c .. 22.5 17 64 45 74 235 161 31 38 .. 9 1.5 Haiti 2.4e .. 18.9 27 .. 94 .. 152 80 23 26 10 32 2.2j Honduras 3.4e .. 8.6 64 89 106 109 58 27 45 67 47 65 1.5 Hungary 8.6c 2.3 .. 93 96 100 99 17 7 .. 100 .. .. 0.1 India 8.1c .. 43.5 64 86 70 91 115 76 .. 47 43 56 0.9 Indonesia 7.1c 31.0 24.4 91 99 93 98 91 34 32 72 50 57 0.1 Iran, Islamic Rep. of 6.5c .. .. 91 101 85 105 72 34 .. 90 49 74 0.2 Iraq .. .. .. 58 .. 78 .. 53 .. 54 .. 14 .. .. Ireland 7.4e .. .. .. 96 104 103 9 5 .. 100 60 .. 0.2 Israel 5.7e .. .. .. 101 105 101 12 5 .. .. 68 .. 0.2 Italy 6.5e .. .. 104 100 100 99 9 4 .. 99 .. .. 0.5 Japan 10.6e .. .. 101 .. 101 100 6 4 100 100 58 56 <0.1 Jordan 6.7c 4.8 3.6 72 99 101 102 40 25 87 100 40 56 0.2 Kazakhstan 7.4c .. .. .. 101k 102 99k 60 29 .. 100 .. 51 0.1 Kenya 6.0c .. 16.5 .. 93 94 96 97 121 50 42 27 39 6.1 Korea, Rep. of 7.9e .. .. 98 101k 99 96k 9 5 98 100 77 .. <0.1 Kyrgyz Republic 8.9c .. .. .. 99 .. 100 75 41 .. 98 .. 48 0.1 Lao PDR 8.1c .. 36.4 46 75 76 85 163 75 .. 19 .. 32 0.1 Lebanon .. .. 3.4 .. 80 .. 103 37 30 .. 98 .. 58 0.1 Liberia .. .. 22.8 .. 63 .. .. 235 235 .. 51 .. 10 .. Libya .. .. .. .. .. .. 105 41 18 .. .. .. .. 0.2 Lithuania 6.8c .. .. .. 93 .. 100 13 8 .. 100 .. .. 0.2 Madagascar 4.9c 35.5 36.8 33 57 98 96 168 115 57 51 17 27 0.5 Malawi 7.0c 24.4 18.4 29 55 81 100 221 120 55 54 13 42 14.1 Malaysia 4.4e .. .. 91 98 101 104 22 12 .. 98 50 .. 0.5 Mali 6.1c .. 30.1 13 49 57 74 250 217 .. 41 .. 8 1.7 Mauritania 6.2c .. 30.4 34 47 71 102 133 125 40 57 3 8 0.7 Mexico 4.3c 13.9 3.4 88 104 97 99 53 35 .. 83 .. 71 0.3 Moldova 7.8c .. 3.2 .. 98 106 103 37 19 .. 100 .. 68 1.1 Morocco 6.5c 8.1 9.9 48 84 70 87 89 37 31 63 42 63 0.1 Mozambique 5.4c .. 21.2 26 42 71 85 235 138 .. 48 .. 17 16.1 Myanmar .. .. 29.6 .. 95 97 101 130 104 .. 68 17 34 1.3 Selected world development indicators 355 Eradicate extreme poverty and hunger Achieve Promote gender Improve maternal health Combat universal equality HIV/AIDS Share of poorest primary Reduce child Contraceptive and other Ratio of girls to quintile in national Prevalence of education mortality Prevalence rate diseases boys enrolllments consumption or child malnutrition Births attended Primary in primary and Under-five % of married HIV prevalence income by skilled health staff % of children completion ratea secondary schoola mortality rate women ages % of population % b under 5 % % per 1,000 % of total 15­49 ages 15­49 1992­2005 1990 2000­07b 1991 2006 1991 2006 1990 2006 1990 2000­07b 1990 2000­07b 2005 Nepal 6.0c .. 38.8 51 76 59 93 142 59 7 19 23 48 0.5 Netherlands 7.6e .. .. .. .. 97 98 9 5 .. 100 76 .. 0.2 New Zealand 6.4e .. .. 100 .. 100 103 11 6 .. 97 .. .. 0.1 Nicaragua 5.6c .. 7.8 42 73 109 102 68 36 .. 67 .. 69 0.2 Niger 2.6c 41.0 39.9 18 33 53 70 320 253 15 18 4 11 1.1 Nigeria 5.0c 35.1 27.2 .. 76 77 83 230 191 33 36 6 13 3.9 Norway 9.6e .. .. 100 96 102 100 9 4 100 .. 74 .. 0.1 Pakistan 9.1c 39.0 31.3 .. 62 .. 78 130 97 19 31 15 28 0.1 Panama 2.5e .. .. 86 94 .. 101 34 23 .. 91 .. .. 0.9 Papua New Guinea 4.5c .. .. 46 .. 80 .. 94 73 .. 42 .. .. 1.8 Paraguay 2.4e 2.8 .. 68 95 98 99 41 22 66 77 48 73 0.4 Peru 3.7e 8.8 5.2 .. 101 96 101 78 25 80 87 59 46 0.6 Philippines 5.4c .. 20.7 86 94 100 102 62 32 .. 60 36 49 <0.1 Poland 7.4c .. .. 98 97 101 99 18 7 .. 100 49 .. 0.1 Portugal 5.8e .. .. 95 104 103 101 14 5 98 100 .. .. 0.4 Romania 8.2c .. 3.5 96 101 99 100 31 18 .. 98 .. 70 <0.1 Russian Federation 6.1c .. .. .. 94 104 99 27 16 .. 99 34 .. 1.1 Rwanda 5.3c 24.3 18.0 35 35 92 102 176 160 26 39 21 17 3.0i Saudi Arabia .. .. .. 55 .. 84 .. 44 25 .. 96 .. .. 0.2 Senegal 6.6c .. 14.5 42 49 69 92 149 116 .. 52 .. 12 0.7i Serbia 8.3c,f .. 1.8 .. .. .. .. .. 8 .. 99 .. 41 0.2f Sierra Leone 6.5c .. 24.7 .. 81k 67 86k 290 270 .. 43 .. 5 1.6 Singapore 5.0e .. 3.3 .. .. 95 .. 8 3 .. 100 65 .. 0.3 Slovak Republic 8.8e .. .. 96 93 .. 100 14 8 .. 100 74 .. <0.1 Somalia .. .. 32.8 .. .. .. .. 203 145 .. 33 1 15 0.9 South Africa 3.5c .. .. 76 100 104 100 60 69 .. 92 57 60 18.8 Spain 7.0e .. .. .. 103 104 103 9 4 .. .. .. .. 0.6 Sri Lanka 7.0c .. 22.8 102 108 102 .. 32 13 .. 96 .. 70 <0.1 Sudan .. .. 38.4 42 47 77 89 120 89 69 49 9 8 1.6 Sweden 9.1e .. .. 96 .. 102 100 7 3 .. .. .. .. 0.2 Switzerland 7.6e .. .. 53 88 97 97 9 5 .. 100 .. .. 0.4 Syrian Arab Republic .. .. 8.5 89 115 85 95 38 14 .. 93 .. 58 0.2 Tajikistan 7.8c .. .. .. 106 .. 88 115 68 .. 83 .. 38 0.1 Tanzania 7.3c 25.1 16.7 62 85k 97 .. 161 118 53 43 10 26 6.5 Thailand 6.3c .. 7.0 .. .. 97 104 31 8 .. 97 .. 77 1.4 Togo .. 21.2 .. 35 67 59 73 149 108 31 62 34 17 3.2 Tunisia 6.0c 8.5 .. 74 120 86 104 52 23 69 90 50 63 0.1 Turkey 5.3c .. 3.5 90 96 81 90 82 26 .. 83 63 71 0.2 Turkmenistan 6.1c .. .. .. .. .. .. 99 51 .. 100 .. 48 <0.1 Uganda 5.7c 19.7 19.0 .. 54 82 98 160 134 38 42 5 24 6.4l Ukraine 9.0c .. 4.1 94 105 .. 99 25 24 .. 100 .. 66 1.4 United Arab Emirates .. .. .. 103 100 104 101 15 8 .. 100 .. .. 0.2 United Kingdom 6.1e .. .. .. .. 102 102 10 6 .. .. .. 84 0.2 United States 5.4e .. 1.1 .. 95 100 100 11 8 99 99 71 .. 0.6 Uruguay 4.5d,e .. 6.0 94 99 .. 106 23 12 .. 99 .. .. 0.5 Uzbekistan 7.2c .. 4.4 .. 100 94 98k 74 43 .. 100 .. 65 0.2 Venezuela, R.B. de 3.3e .. .. 43 96 105 103 33 21 .. 95 .. .. 0.7 Vietnam 7.1c .. 20.2 .. .. .. .. 53 17 .. 88 53 76 0.5i West Bank and Gaza .. .. .. .. 95 .. 104 40 22 .. 99 .. 50 .. Yemen, Rep. 7.2c .. 41.3 .. 60 .. 66 139 100 16 27 10 23 0.2 Zambia 3.6c 21.2 23.3 .. 84 .. 93 180 182 51 43 15 34 17.0 Zimbabwe 4.6c 8.0 14.0 97 .. 92 97 76 105 70 80 43 60 18.1j World ..w 23.9w 79w 86w ..w 95w 92w 72w 49w 65w 57w 60w 1.0w Low income .. 28.9 49 65 76 88 164 135 33 41 22 33 2.5 Middle income .. 22.9 82 93 86 97 75 49 48 73 61 69 0.7 Lower middle income .. 25.2 83 91 83 96 81 54 44 69 63 69 0.5 Upper middle income .. .. 88 101 99 100 46 26 79 94 50 67 1.6 Low and middle income .. 24.9 77 85 84 94 101 79 45 62 54 60 1.1 East Asia & Pacific .. 13.3 101 98 90 100 56 29 47 87 75 79 0.2 Europe & Central Asia .. .. 90 98 98 97 49 26 81 95 44 63 0.6 Latin America & the .. 5.1 82 100 99 101 55 26 75 88 57 67 0.6 Caribbean Middle East & North .. .. 77 90 79 93 78 42 48 77 42 60 0.1 Africa South Asia .. 41.3 62 80 70 90 123 83 30 41 40 53 0.7 Sub-Saharan Africa .. 26.8 51 60 82 87 184 157 44 45 15 22 5.8 High income .. .. .. 97 100 99 12 7 .. 99 71 .. 0.4 a. Because of the change from International Standard Classification of Education 1976 (ISCED76) to ISCED97, data before 1998 are not fully comparable with data from 1999 onward. b. Data are for the most recent year available. c. Refers to expenditure shares by percentiles of population, ranked by per capita expenditure. d. Urban data. e. Refers to income shares by percentiles of population, ranked by per capita income. f. Includes Montenegro. g. Survey data, 2004. h. Includes Hong Kong, China. i. Survey data 2005. j. Survey data 2005­2006. k. Data are for 2007. l. Survey data, 2004­2005. 356 WORLD DEVELOPMENT REPORT 2009 Table 3 Economic activity Gross domestic product Agricultural tchr Value added as % of GDP External productivity Household General gov't. Gross balance of GDP implicit final cons. final cons. capital goods and deflator Agricultural value expenditure expenditure formation services Millions of Avg. annual added per worker Avg. annual dollars % growth 2000 $ Agriculture Industry Services % of GDP % of GDP % of GDP % of GDP % growth 2007 2000­07 1990­92 2003­05 2007 2007 2007 2007 2007 2007 2007 2000­07 Afghanistan 11,627 11.5 .. .. 36 24 39 111 10 28 ­49 5.9 Albania 10,569 5.3 778 1,449 23 22 56 90 9 27 ­26 3.5 Algeria 135,285 4.5 1,911 2,225 8 61 30 34 12 30 24 8.8 Angola 58,547 13.1 165 174 10 68 22 59 ..a 14 26 55.3 Argentina 262,331 4.7 6,767 10,072 8 36 56 59 12 24 6 12.3 Armenia 9,177 12.7 1,476b 3,692 18 44 38 74 9 32 ­15 4.2 Australia 821,716 3.3 20,838 29,924 3 28 69 57 18 27 ­1 3.7 Austria 377,028 1.9 12,048 22,203 2 31 67 56 18 21 6 1.8 Azerbaijan 31,248 17.2 1,084b 1,143 6 62 32 32 13 27 27 9.3 Bangladesh 67,694 5.8 254 338 19 29 53 77 6 24 ­7 4.3 Belarus 44,771 8.3 1,977b 3,153 9 40 51 62 20 28 ­9 27.6 Belgium 448,560 1.9 21,479 41,631 1 24 75 53 23 22 3 2.0 Benin 5,428 3.8 326 519 32 13 54 78 15 20 ­13 2.9 Bolivia 13,120 3.6 670 773 14 32 54 68 13 13 5 6.8 Bosnia and Herzegovina 14,661 5.3 .. 8,270 10 26 64 76 24 18 ­18 3.8 Brazil 1,314,170 3.3 1,506 3,126 5 31 64 48 28 22 2 8.5 Bulgaria 39,549 5.7 2,500 7,159 8 33 59 75 12 35 ­22 5.0 Burkina Faso 6,767 5.8 110 173 33 22 44 75 22 18 ­15 2.2 Burundi 974 2.7 108 70 35 20 45 91 29 17 ­37 8.4 Cambodia 8,628 9.8 .. 306 30 26 44 82 3 21 ­7 3.8 Cameroon 20,644 3.5 389 646 19 29 52 73 9 17 1 2.2 Canada 1,326,376 2.7 28,243 43,055 .. .. .. 55 19 22 4 2.0 Central African Republic 1,712 0.1 287 381 56 16 28 88 10 9 ­7 1.9 Chad 7,085 12.2 173 215 23 44 32 60 6 19 15 8.2 Chile 163,915 4.5 3,600 5,308 4 48 48 55 10 21 14 7.0 China 3,280,053 10.2 254 401 12 48 40 34 14 44 8 3.7 Hong Kong, China 206,706 5.2 .. .. 0 9 91 60 8 21 11 ­2.3 Colombia 171,979 4.5 3,405 2,847 11 29 60 63 13 24 ­1 6.3 Congo, Dem. Rep. of 8,955 5.0 184 149 42 28 29 82 9 18 ­9 31.0 Congo, Rep. of 7,646 4.1 .. .. 5 60 35 29 14 27 30 5.7 Costa Rica 25,225 5.2 3,143 4,499 9 31 60 75 5 27 ­7 9.8 Côte d'Ivoire 19,570 0.2 598 795 23 26 51 73 8 10 8 3.1 Croatia 51,277 4.8 4,921b 9,987 7 32 61 56 20 33 ­8 3.7 Czech Republic 168,142 4.5 .. 5,423 3 39 58 48 21 27 3 2.1 Denmark 308,093 1.8 15,190 40,780 2 26 72 49 26 23 3 2.2 Dominican Republic 36,686 4.8 2,268 4,586 12 28 60 80 7 20 ­6 17.4 Ecuador 44,184 5.0 1,686 1,676 7 35 58 67 13 21 ­1 9.6 Egypt, Arab Rep. of 128,095 4.5 1,528 2,072 13 36 51 75 11 22 ­8 6.9 El Salvador 20,215 2.7 1,633 1,638 11 29 60 93 11 16 ­21 3.4 Eritrea 1,201 2.3 .. 61 18 24 58 81 42 19 ­42 14.8 Ethiopia 19,395 7.5 146 158 46 13 40 84 11 25 ­19 6.6 Finland 246,020 3.1 18,822 31,214 3 32 65 51 21 21 7 0.9 France 2,562,288 1.7 22,234 44,017 2 21 77 57 24 21 ­1 2.0 Georgia 10,176 8.3 2,443b 1,790 11 24 65 75 22 29 ­26 6.9 Germany 3,297,233 1.1 13,724 26,549 1 30 69 58 18 18 5 1.0 Ghana 15,246 5.5 293 320 36 25 38 78 13 33 ­23 19.5 Greece 360,031 4.3 7,668 9,011 3 21 76 68 14 26 ­8 3.4 Guatemala 33,432 3.6 2,119 2,350 12 27 60 87 9 20 ­16 4.7 Guinea 4,564 2.8 142 190 17 45 38 84 6 13 ­2 18.1 Haiti 6,137 0.2 .. .. .. .. .. 91 9 29 ­29 16.5 Honduras 12,279 5.3 1,193 1,483 13 28 58 79 15 28 ­23 6.2 Hungary 138,182 4.0 4,105 6,987 4 29 66 66 10 22 2 5.1 India 1,170,968 7.8 324 392 18 29 53 55 10 38 ­3 4.4 Indonesia 432,817 5.1 484 583 14 47 39 63 8 25 4 10.1 Iran, Islamic Rep. of 270,937 5.9 1,954 2,542 9 42 49 47 14 37 3 16.9 Iraq .. ­11.4 .. 1,756 .. .. .. .. .. .. .. .. Ireland 254,970 5.2 .. 17,879 2 36 62 44 16 27 13 3.1 Israel 161,822 3.3 .. .. .. .. .. 56 26 20 ­1 1.2 Italy 2,107,481 0.8 11,542 23,967 2 27 71 59 20 21 ­1 2.7 Japan 4,376,705 1.7 20,445 35,517 2 30 69 57 18 23 1 ­1.2 Jordan 15,832 6.3 1,892 1,360 3 32 65 89 20 26 ­35 3.0 Kazakhstan 103,840 10.0 1,795b 1,557 7 44 49 48 11 31 9 14.3 Kenya 29,509 4.4 333 332 23 19 58 75 16 20 ­10 6.0 Korea, Rep. of 969,795 4.7 5,679 11,286 3 39 58 55 15 29 1 1.7 Kyrgyz Republic 3,505 4.0 675b 979 33 20 47 101 19 17 ­37 5.8 Lao PDR 4,008 6.6 360 458 42 32 26 65 9 33 ­6 9.4 Lebanon 24,001 3.3 .. 30,099 6 23 71 89 15 12 ­16 2.0 Liberia 725 ­2.7 .. .. 66 16 18 86 11 16 ­14 10.0 Libya 58,333 3.7 .. .. .. .. .. .. .. .. .. 21.0 Lithuania 38,328 8.0 .. 4,703 5 33 61 66 17 30 ­12 2.9 Madagascar 7,326 3.3 186 174 27 15 58 81 10 29 ­19 11.6 Malawi 3,552 3.2 72 116 34 20 45 71 12 28 ­11 21.3 Malaysia 180,714 5.1 3,803 5,126 9 51 41 50 13 23 14 4.8 Mali 6,863 5.4 208 241 37 24 39 76 11 23 ­10 3.6 Mauritania 2,644 5.1 574 356 13 47 41 61 20 26 ­7 11.3 Mexico 893,364 2.6 2,256 2,792 4 25 71 71 9 23 ­2 6.3 Moldova 4,396 6.5 1,286b 816 17 15 67 97 18 30 ­44 11.5 Morocco 73,275 4.9 1,430 1,775 12 29 59 58 16 32 ­6 1.1 Mozambique 7,752 8.0 109 153 28 27 45 72 12 24 ­8 7.8 Myanmar .. 9.2 .. .. .. .. .. .. .. .. .. 21.2 Selected world development indicators 357 Gross domestic product Agricultural tchr Value added as % of GDP External productivity Household General gov't. Gross balance of GDP implicit final cons. final cons. capital goods and deflator Agricultural value expenditure expenditure formation services Millions of Avg. annual added per worker Avg. annual dollars % growth 2000 $ Agriculture Industry Services % of GDP % of GDP % of GDP % of GDP % growth 2007 2000­07 1990­92 2003­05 2007 2007 2007 2007 2007 2007 2007 2000­07 Nepal 10,207 3.2 191 207 35 16 49 82 9 25 ­16 5.6 Netherlands 754,203 1.6 24,914 42,198 2 25 73 47 25 20 8 2.1 New Zealand 129,372 3.2 19,204 25,109 .. .. .. 60 18 25 ­3 2.4 Nicaragua 5,676 3.5 .. 2,071 20 30 51 92 9 29 ­30 7.6 Niger 4,170 3.9 152 157b .. .. .. 75 12 23 ­9 2.1 Nigeria 165,690 6.7 .. .. 33 39 28 .. .. .. 9 17.8 Norway 381,951 2.5 19,500 37,776 2 45 54 41 19 22 18 4.0 Pakistan 143,597 5.8 593 695 20 27 54 75 10 23 ­8 6.5 Panama 19,740 6.0 2,363 3,914 7 16 77 71 7 23 ­1 1.8 Papua New Guinea 6,261 2.4 500 595 36 45 19 47 12 20 21 7.2 Paraguay 12,004 3.3 1,596 2,052 26 20 54 74 9 21 ­4 10.6 Peru 109,088 5.4 930 1,498 6 35 59 63 9 20 8 3.8 Philippines 144,129 5.1 905 1,075 14 31 55 80 10 15 ­5 5.1 Poland 420,321 4.1 1,502b 2,182 4 30 66 63 17 22 ­1 2.4 Portugal 220,241 0.8 4,612 5,980 3 25 72 65 21 22 ­8 3.0 Romania 165,980 6.1 2,196 4,646 8 26 65 77 14 22 ­13 18.0 Russian Federation 1,291,011 6.6 1,825b 2,519 5 39 57 50 17 25 8 16.7 Rwanda 3,320 5.8 168 182 36 14 50 85 11 22 ­18 9.7 Saudi Arabia 381,683 4.1 7,875 15,780 3 65 32 28 23 22 27 8.1 Senegal 11,151 4.5 225 215 15 22 63 76 10 32 ­18 2.2 Serbia 41,581 5.5 .. .. 13 26 62 73 22 25 ­20 19.5 Sierra Leone 1,672 11.2 .. .. 44 24 32 84 13 17 ­14 8.9 Singapore 161,347 5.8 22,695 40,419 0 31 69 38 10 23 29 1.0 Slovak Republic 74,932 6.0 .. 5,026 3 37 60 55 18 27 0 3.9 Somalia .. .. .. .. .. .. .. .. .. .. .. .. South Africa 277,581 4.3 1,786 2,484 3 31 66 64 20 20 ­4 6.6 Spain 1,429,226 3.4 9,511 19,030 3 30 67 58 18 31 ­6 4.0 Sri Lanka 32,354 5.3 679 702 12 30 58 68 15 27 ­10 10.0 Sudan 47,632 7.1 418 666 32 28 41 68 14 24 ­6 9.6 Sweden 444,443 2.8 21,463 33,023 1 29 70 47 27 18 8 1.5 Switzerland 415,516 1.6 22,344 23,418 1 28 70 60 11 22 7 0.9 Syrian Arab Republic 38,081 4.5 2,344 3,261 20 32 48 71 12 16 2 6.2 Tajikistan 3,712 8.8 397b 465 21 28 51 113 9 23 ­45 20.5 Tanzaniac 16,181 6.7 238 295 45 17 37 73 16 17 ­6 9.0 Thailand 245,818 5.4 497 621 11 44 45 57 10 30 4 2.9 Togo 2,493 2.6 312 347 43 23 34 85 10 18 ­13 0.8 Tunisia 35,020 4.8 2,422 2,719 11 27 62 64 13 23 0 2.7 Turkey 657,091 5.9 .. 1,846 9 28 63 71 12 22 ­5 18.8 Turkmenistan 12,933 .. 1,222b .. .. .. .. 46 13 23 17 .. Uganda 11,214 5.7 184 229 29 18 53 80 14 24 ­19 6.2 Ukraine 140,484 7.6 1,195b 1,702 7 32 61 66 15 22 ­4 14.0 United Arab Emirates 129,702 8.2 10,454 25,841 2 56 42 46 11 24 18 4.9 United Kingdom 2,727,806 2.6 22,659 26,933 1 24 75 64 22 18 ­4 2.6 United States 13,811,200 2.7 20,793 41,797 1 23 76 71 16 19 ­6 2.6 Uruguay 23,087 3.3 5,714 7,973 9 32 59 73 11 18 ­2 9.4 Uzbekistan 22,308 6.2 1,272b 1,800 24 27 49 54 16 20 10 26.5 Venezuela, R.B. de 228,071 4.7 4,483 6,292 .. .. .. 48 11 24 17 26.8 Vietnam 71,216 7.8 214 305 20 42 38 67 6 35 ­8 6.7 West Bank and Gaza 4,007 0.4 .. .. .. .. .. 96 33 23 ­52 3.0 Yemen, Rep. of 22,523 4.0 271 328b .. .. .. .. .. .. .. 13.5 Zambia 11,363 ­5.7 159 204 22 38 40 59 10 24 6 18.3 Zimbabwe 3,418 ­4.4 240 222 19 24 57 72 27 17 ­16 232.0 World 54,347,038t 3.2w 730w 911w 3w 28w 69w 61w 17w 22w 0w Low income 810,300 5.6 259 321 25 28 48 74 9 24 ­6 Middle income 13,342,194 6.2 454 654 8 32 59 60 15 25 0 Lower middle income 6,888,343 8.0 370 509 13 41 46 49 13 35 3 Upper middle income 6,450,429 4.3 2,134 2,954 5 31 64 59 17 23 1 Low and middle income 14,155,882 6.2 417 583 10 32 59 61 15 25 ­1 East Asia & Pacific 4,438,135 8.9 303 446 12 47 41 41 13 38 7 Europe & Central Asia 3,155,221 6.1 1,588 2,109 7 33 60 61 16 24 ­1 Latin America & the 3,444,374 3.6 2,155 3,053 5 29 66 60 17 22 1 Caribbean Middle East & North 828,691 4.5 1,583 2,205 11 35 53 62 14 26 ­1 Africa South Asia 1,438,594 7.3 335 406 18 29 53 59 10 35 ­4 Sub-Saharan Africa 842,914 5.0 246 281 15 32 54 67 16 21 ­3 High income 40,197,253 2.4 14,586 25,456 2 26 72 62 18 21 ­1 a. Data on general government final consumption expenditure are not available separately; they are included in household final consumption expenditure. b. Data for all three years are not available. c. Data refer to mainland Tanzania only. 358 WORLD DEVELOPMENT REPORT 2009 Table 4 Trade, aid, and finance Merchandise trade Manufactured High technology Domestic Exports Imports External debt exports exports Current Foreign direct Official credit provided by account investment development banking % of total % of Present balance net inflows assistancea sector merchandise manufactured Total $ value % of Net migration $ millions $ millions exports exports $ millions $ millions $ per capita millions GNI % of GDP thousands 2007 2007 2006 2006 2007 2006 2006 2006 2006 2007 2000­05b Afghanistan 480 2,950 .. .. .. .. .. 1,771 18d 0 1,112 Albania 1,072 4,196 27 13 ­671 325 101 2,340 21 64 ­110 Algeria 59,518 27,439 1 2 .. 1,795 6 5,583 5 ­3 ­140 Angola 38,100 11,400 .. .. 10,690 ­38 10 9,563 33 2 175 Argentina 55,933 44,780 32 7 7,210 4,840 3 122,190 68 29 ­100 Armenia 1,219 3,282 56 1 ­571 343 71 2,073 29 12 ­100 Australia 141,079 165,331 23 12 ­56,783 26,599 .. .. .. 142 593 Austria 162,204 161,800 80 13 12,031 157 .. .. .. 124 180 Azerbaijan 9,300 6,050 8 2 9,019 ­584 24 1,900 12 18 ­100 Bangladesh 12,360 18,470 92 0 1,196 697 8 20,521 22 59 ­500 Belarus 24,339 28,674 50 3 ­2,944 354 7 6,124 17 27 .. Belgium 432,327 415,752 77 8c 8,254 61,990 .. .. .. 114 180 Benin 590 1,110 9 0 ­226 63 43 824 12d 9 99 Bolivia 4,485 3,446 7 4 1,319 240 62 5,292 20d 54 ­100 Bosnia and Herzegovina 4,155 9,726 62 3 ­1,939 423 126 5,669 43 56 115 Brazil 160,649 126,581 51 12 1,460 18,782 0 194,150 26 96 ­229 Bulgaria 18,450 30,034 53 6 ­8,592 5,172 .. 20,925 74 59 ­43 Burkina Faso 660 1,700 8 10 .. 26 61 1,142 11d 12 100 Burundi 55 350 6 4 ­135 0 51 1,411 105 38 192 Cambodia 4,400 5,300 97 0 ­506 483 37 3,527 48 13 10 Cameroon 3,750 3,760 3 3 .. 309 93 3,171 4d 6 6 Canada 418,493 389,670 56 15 12,815 69,068 .. .. .. 166 1,041 Central African Republic 195 230 36 0 .. 24 31 1,020 53d 18 ­45 Chad 3,450 1,500 .. .. .. 700 27 1,772 23d 0 219 Chile 68,296 46,108 11 7 7,200 7,952 5 47,977 42 90 30 China 1,217,939e 955,845 92e 30 249,866 78,095 1 322,845 14 136 ­1,900 Hong Kong, China 349,663 370,733 91 11 27,405 42,891 .. .. .. 126 300 Colombia 29,360 32,897 37 4 ­5,851 6,463 22 39,698 32 50 ­120 Congo, Dem. Rep. of 2,600 2,950 .. .. .. 180 34 11,201 119d 5 ­237 Congo, Rep. of 6,100 2,900 .. .. 903 344 69 6,130 108d ­10 ­10 Costa Rica 9,367 12,955 65 45 ­1,499 1,469 5 6,832 35 48 84 Côte d'Ivoire 8,400 6,100 15 42 ­146 315 13 13,840 72d 21 ­339 Croatia 12,360 25,830 66 10 ­4,412 3,376 45 37,480 93 83 100 Czech Republic 122,414 117,980 89 14 ­4,586 6,021 .. .. .. 55 67 Denmark 103,307 99,375 65 20 4,279 3,343 .. .. .. 207 46 Dominican Republic 6,700 13,100 .. .. ­2,231 1,183 6 8,905 35 54 ­148 Ecuador 13,751 13,565 10 8 1,503 271 14 16,536 52 19 ­400 Egypt, Arab Rep. of 16,201 27,064 21 1 2,635 10,043 12 29,339 28 91 ­525 El Salvador 3,980 8,677 55 3 ­855 204 23 9,136 55 46 ­143 Eritrea 15 515 .. .. .. 4 28 800 49d 139 229 Ethiopia 1,290 5,320 .. .. ­1,786 364 25 2,326 8d 47 ­140 Finland 89,656 81,145 81 22 11,402 5,311 .. .. .. 85 33 France 552,193 613,224 79 21 ­30,567 81,045 .. .. .. 123 722 Georgia 1,240 5,217 48 16 ­1,931 1,060 81 1,964 22 32 ­248 Germany 1,326,521 1,059,439 83 17 150,746 43,410 .. .. .. 126 1,000 Ghana 4,320 7,980 31 0 ­1,040 435 51 3,192 21d 33 12 Greece 23,574 75,553 52 11 ­29,565 5,401 .. .. .. 95 154 Guatemala 6,926 13,578 35 3 ­1,592 354 37 5,496 18 42 ­300 Guinea 1,100 1,190 .. .. .. 108 18 3,281 58d 16 ­425 Haiti 550 1,550 .. .. 1 160 62 1,189 22d 25 ­140 Honduras 2,160 6,760 21 1 ­195 385 84 4,076 25d 51 ­150 Hungary 94,160 94,792 84 24 ­7,421 6,098 .. 107,677 100 75 65 India 145,228 216,682 70 5 ­9,415 17,453 1 153,075 15 63 ­1,350 Indonesia 118,163 91,715 45 13 11,009 5,580 6 130,956 45 41 ­1,000 Iran, Islamic Rep. of 83,000 45,000 10 6 .. 901 2 20,113 10 48 ­1,250 Iraq 36,400 29,020 .. .. .. .. .. .. .. .. ­375 Ireland 121,068 81,678 85 34 ­12,695 ­882 .. .. .. 199 188 Israel 54,065 58,950 82 14 4,994 14,302 .. .. .. 76 115 Italy 491,532 504,591 85 7 ­51,032 38,884 .. .. .. 129 1,125 Japan 712,839 620,967 91 22 210,490 ­6,784 .. .. .. 294 270 Jordan 5,760 13,310 71 1 ­1,909 3,219 105 8,000 58 124 130 Kazakhstan 46,540 32,940 13 21 ­7,184 6,143 11 74,148 132 41 ­200 Kenya 4,140 9,210 26 3 ­526 51 26 6,534 26 34 25 Korea, Rep. of 371,554 356,648 89 32 5,954 3,645 .. .. .. 110 ­80 Kyrgyz Republic 1,105 2,475 46 3 ­234 182 60 2,382 52d 15 ­75 Lao PDR 980 1,400 .. .. .. 187 63 2,985 87 8 ­115 Lebanon 3,574 12,251 70 2 ­2,046 2,794 174 23,963 116 190 .. Liberia 157 490 .. .. ­138 ­82 75 2,674 1,128d 92 ­119 Libya 45,000 8,600 .. .. 22,170 .. 6 .. .. ­70 10 Lithuania 17,173 24,116 58 8 ­3,218 1,812 .. 18,955 79 61 ­30 Madagascar 1,190 2,590 41 1 ­554 230 39 1,453 13d 9 ­5 Malawi 670 1,380 13 11 .. 30 49 850 6d 16 ­30 Malaysia 176,211 146,982 74 54 28,931 6,064 9 52,526 39 117 150 Mali 1,620 2,000 10 4 ­231 185 69 1,436 15d 15 ­134 Mauritania 1,360 1,510 0 .. .. ­3 62 1,630 93d .. 30 Mexico 272,044 296,578 76 19 ­1,993 19,222 2 160,700 21 43 ­3,983 Moldova 1,370 3,720 31 5 ­695 242 60 2,416 65 40 ­250 Morocco 14,646 31,468 68 10 1,851 2,699 34 18,493 30 92 ­550 Mozambique 2,650 3,210 5 2 ­634 154 77 3,265 12d 10 ­20 Myanmar 5,350 3,250 .. .. 802 279 3 6,828 47 28 ­99 Selected world development indicators 359 Merchandise trade Manufactured High technology Domestic Exports Imports External debt exports exports Current Foreign direct Official credit provided by account investment development banking % of total % of Present balance net inflows assistancea sector merchandise manufactured Total $ value % of Net migration $ millions $ millions exports exports $ millions $ millions $ per capita millions GNI % of GDP thousands 2007 2007 2006 2006 2007 2006 2006 2006 2006 2007 2000­05b Nepal 888 2,904 .. .. 6 ­7 19 3,409 27d 49 ­100 Netherlands 550,636 490,582 66 28 50,706 7,197 .. .. .. 208 110 New Zealand 26,950 30,890 27 11 ­10,233 7,941 .. .. .. 152 102 Nicaragua 1,210 3,510 9 7 ­855 282 132 4,391 30d 74 ­210 Niger 650 970 15 11 ­312 20 29 805 8d 7 ­29 Nigeria 66,500 27,500 .. .. 24,202 5,445 79 7,693 9 4 ­170 Norway 139,424 80,347 16 19 64,070 4,653 .. .. .. .. 84 Pakistan 17,457 32,598 81 1 ­8,253 4,273 14 35,909 26 46 ­1,239 Panama 1,200 7,010 10 0 ­1,577 2,574 9 9,989 77 88 8 Papua New Guinea 4,610 2,950 .. .. 640 32 45 1,675 35 23 .. Paraguay 3,374 7,280 16 8 ­217 189 9 3,426 43 20 ­45 Peru 27,956 20,185 14 2 2,589 3,467 17 28,174 42 16 ­510 Philippines 50,276 57,160 87 68 5,897 2,345 7 60,324 57 41 ­900 Poland 137,609 160,804 79 4 ­15,794 19,198 .. 125,831 41 47 ­200 Portugal 50,994 77,050 74 9 ­18,281 7,366 .. .. .. 174 276 Romania 40,257 69,712 79 4 ­23,136 11,394 .. 55,114 58 36 ­270 Russian Federation 355,177 223,059 17 9 78,310 30,827 .. 251,067 34 25 917 Rwanda 165 600 .. .. ­147 11 62 419 8d 9 43 Saudi Arabia 228,550 94,235 8 1 99,066 660 1 .. .. 18 285 Senegal 1,650 4,250 44 6 .. 58 68 1,984 14d 25 ­100 Serbia 8,780 18,295 .. 4 .. 5,128 214 13,831 52 30 ­339 Sierra Leone 260 420 .. .. ­101 59 63 1,428 10d 10 472 Singapore 299,271e 263,150 80e 58 36,326 24,191 .. .. .. 81 200 Slovak Republic 58,082 60,103 85 6 .. 4,165 .. 27,085 58 52 3 Somalia .. .. .. .. .. 96 46 2,836 .. .. 100 South Africa 69,788 90,990 53f 6 ­20,631 ­120 15 35,549 15 89 75 Spain 241,962 373,585 76 6 ­145,275 20,167 .. .. .. 194 2,846 Sri Lanka 7,750 10,840 70 2 ­1,334 480 40 11,446 40 47 ­442 Sudan 8,160 8,450 0 1 ­4,722 3,534 55 19,158 77d 0 ­532 Sweden 168,223 150,039 78 16 28,413 27,299 .. .. .. 135 152 Switzerland 171,621 160,798 91 22 72,354 27,185 .. .. .. 194 100 Syrian Arab Republic 11,330 14,820 32 1 920 600 1 6,502 23 33 200 Tajikistan 1,468 2,455 .. .. ­21 339 36 1,154 36 15 ­345 Tanzania 2,005 5,337 18 0 ­1,442 474 46 4,240 16d,g 13 ­345 Thailand 152,469 141,347 76 27 14,921 9,010 ­3 55,233 30 96 231 Togo 690 1,450 58 0 ­461 57 12 1,806 68d 22 ­4 Tunisia 15,029 18,980 75 4 ­634 3,270 43 18,480 66 72 ­29 Turkey 107,154 169,987 42 .. ­32,774 20,070 8 207,854 61 49 ­30 Turkmenistan 8,920 4,460 .. .. .. 731 5 881 11 .. ­10 Uganda 1,530 3,350 21 34 ­745 392 52 1,264 6d 9 ­5 Ukraine 49,100 60,440 73 3 ­5,927 5,604 10 49,887 58 62 ­173 United Arab Emirates 154,000 121,100 .. .. .. .. .. .. .. 59 577 United Kingdom 435,615 617,178 77 34 ­115,243 139,745 .. .. .. 194 948 United States 1,163,183 2,016,978 79 30 ­738,641 180,580 .. .. .. 240 6,493 Uruguay 4,480 5,480 32 3 ­186 1,346 6 9,804 66 25 ­104 Uzbekistan 8,040 4,470 .. .. .. 164 6 3,892 26 .. ­300 Venezuela, R.B. de 69,165 48,591 5 2 20,001 ­543 2 44,635 34 23 40 Vietnam 48,387 60,830 50 5 ­6,992 2,315 22 20,202 33 96 ­200 West Bank and Gaza .. .. .. .. .. .. 384 .. .. 9 11 Yemen, Rep. of 7,160 5,890 1 5 206 1,121 13 5,563 25 10 ­100 Zambia 4,876 4,014 6 2 ­505 575 122 2,325 9d 17 ­82 Zimbabwe 2,050 2,420 38 2 .. 40 21 4,677 110 93 ­75 World 13,899,267t 14,107,100t 74w 20w 1,352,442s 16w ..s 167w ..wh Low income 230,215 251,819 .. .. 20,380 35 201,382 30 ­2,858 Middle income 3,919,104 3,641,914 60 20 334,242 9 2,642,418 77 ­15,770 Lower middle income 2,179,289 1,947,080 69 25 162,047 9 1,080,416 102 ­11,295 Upper middle income 1,738,728 1,690,142 52 16 172,195 7 1,562,002 55 ­4,475 Low and middle income 4,149,329 3,893,700 60 20 354,621 19 2,843,800 75 ­18,629 East Asia & Pacific 1,783,695 1,475,731 80 33 104,972 4 659,985 119 ­3,847 Europe & Central Asia 874,122 935,854 39 8 114,318 14 912,265 38 ­1,798 Latin America & the 750,092 732,907 53 12 70,457 12 734,499 63 ­6,811 Caribbean Middle East & North 297,678 234,252 19 5 26,551 54 136,499 49 ­2,618 Africa South Asia 184,991 286,021 72 4 22,916 6 227,303 60 ­2,484 Sub-Saharan Africa 261,373 237,971 .. .. 15,408 52 173,248 48 ­1,070 High income 9,752,088 10,219,990 77 21 997,821 0.1 196 18,522 a. The distinction between official aid, for countries on the Part II list of the OECD Development Assistance Committee (DAC), and official development assistance was dropped in 2005. Regional aggregates include data for economies not listed in the table. World and income group totals include aid not allocated by country or region. b. Total for the 5-year period. c. Includes Luxembourg. d. Data are from debt sustainability analysis undertaken as part of the Heavily Indebted Poor Countries (HIPC) initiative. e. Includes re-exports. f. Data on total exports and imports refer to South Africa only. Data on export commodity shares refer to the South African Customs Union (Botswana, Lesotho, Namibia, and South Africa). g. GNI refers to mainland Tanzania only. h. World total computed by the UN sums to zero, but because the aggregates shown here refer to World Bank definitions, regional and income group totals do not equal zero. Table 5 Key indicators for other economies Gross national income PPP gross national Life expectancy Population (GNI)a income (GNI)b Gross at birth Carbon Population domestic Adult dioxide Density age product per literacy rate emissions Avg. annual people per composition Millions of per capita Millions of per capita capita % Male Female % ages 15 per capita Thousands % growth sq. km % ages 0­14 dollars dollars dollars dollars growth years years and older metric tons 2007 2000­2007 2007 2007 2007 2007 2007 2007 2006­2007 2006 2006 2005 2004 American Samoa 60 1.4c 301 .. .. ..d .. .. .. .. .. .. 5.1 Andorra 67 0.5c 143 .. .. ..e .. .. .. .. .. .. .. Antigua and Barbuda 85 1.4 193 .. 977 11,520 1,070f 12,610f 2.9 .. .. .. 5.1 Aruba 101 0.5c 561 22 .. ..e .. .. .. .. .. 97 21.8 Bahamas, The 331 1.3 33 27 .. ..e .. .. .. 70 76 .. 6.3 Bahrain 753 2.1 1,060 25 14,022 19,350 24,869 34,310 5.6 74 77 87 23.8 Barbados 294 0.4 684 18 .. ..e 3,177f 10,880f .. 74 80 .. 4.4 Belize 304 2.8 13 37 1,157 3,800 1,551f 5,100f 0.1 70 74 .. 2.8 Bermuda 64 0.4 1,280 .. .. ..e .. .. .. 76 81 .. 8.7 Bhutan 657 2.3 14 31 1,166 1,770 3,276 4,980 17.5 64 67 60 0.7 Botswana 1,881 1.2 3 35 10,991 5,840 23,369 12,420 2.5 50 50 81 2.4 Brunei Darussalam 389 2.2 74 29 10,287 26,930 19,059 49,900 2.9 75 80 93 24.1 Cape Verde 530 2.3 132 38 1,287 2,430 1,558 2,940 4.6 68 74 81 0.6 Cayman Islands 47 2.1c 180 .. .. ..e .. .. .. .. .. .. 7.1 Channel Islands 149 0.2 785 16 .. ..e .. .. .. 76 81 .. .. Comoros 626 2.1 336 42g 425 680 721 1,150 ­2.9 62g 64g .. 0.2 Cuba 11,257 0.1 103 18 .. ..d .. .. .. 76 80 100 2.3 Cyprus 787 1.8 85 19 19,617 24,940 20,741 26,370 2.5 77 82 97 9.1 Djibouti 833 1.9 36 37 908 1,090 1,886 2,260 2.2 53 56 .. 0.5 Dominica 73 0.3 97 .. 310 4,250 411f 5,650f 0.4 .. .. .. 1.5 Equatorial Guinea 508 2.3 18 42 6,527 12,860 10,773 21,230 9.9 50 52 87 11.5 Estonia 1,342 ­0.3 32 15 17,706 13,200 26,399 19,680 7.3 67 78 100 14.0 Faeroe Islands 48 0.2c 35 .. .. ..e .. .. .. 77 81 .. 13.7 Fiji 838 0.6 46 32 3,189 3,800 3,666 4,370 ­5.0 66 71 .. 1.3 French Polynesia 263 1.5 72 27 .. ..e .. .. .. 71 77 .. 2.7 Gabon 1,330 1.7 5 35 8,876 6,670 17,395 13,080 4.0 56 57 84 1.1 Gambia, The 1,707 3.0 171 41 544 320 1,951 1,140 4.3 58 60 .. 0.2 Greenland 57 0.1 0h .. .. ..e .. .. .. .. .. .. 10.0 Grenada 108 0.9 318 33 505 4,670 649f 6,010f 3.0 .. .. .. 2.0 Guam 173 1.6 321 29 .. ..e .. .. .. 73 78 .. 25.0 Guinea-Bissau 1,695 3.0 60 48 331 200 790 470 ­0.3 45 48 .. 0.2 Guyana 739 0.1 4 31 959 1,300 1,917f 2,600f 5.5 63 69 .. 2.0 Iceland 311 1.4 3 22 16,826 54,100 10,592 34,060 1.4 79 83 .. 7.6 Isle of Man 77 0.9 136 .. 3,088 40,600 2,568f 33,750f 4.9 .. .. .. .. Jamaica 2,677 0.5 247 31 9,923 3,710 13,510 5,050 1.7 70 73 .. 4.0 Kiribati 102 1.7 126 .. 120 1,170 223f 2,190f 0.8 .. .. .. 0.3 Korea, Dem. People's 23,783 0.5 198 23 .. ..i .. .. .. 65 69 .. 3.4 Rep. of Kuwait 2,663 2.8 149 23 80,221 31,640 126,703 49,970 6.7 76 80 93 40.4 Latvia 2,276 ­0.6 37 14 22,595 9,930 38,452 16,890 10.9 65 77 100 3.1 Lesotho 2,006 0.9 66 40 2,007 1,000 3,783 1,890 4.3 43 43 82 .. Liechtenstein 35 0.8c 220 .. .. ..e .. .. .. .. .. .. .. Luxembourg 480 1.3 185 18 36,420 75,880 30,909 64,400 1.9 76 82 .. 24.9 Macao, China 480 1.2 17,026 14 .. ..e .. .. 26.6 78 83 91 4.7 Macedonia, FYR 2,037 0.2 80 19 7,052 3,460 17,344 8,510 5.1 72 76 96 5.1 Maldives 305 1.6 1,018 32 977 3,200 1,540 5,040 3.8 67 69 96 2.5 Malta 409 0.7 1,279 17 6,216 15,310 8,523 20,990 2.7 77 81 .. 6.1 Marshall Islands 67 3.3 369 .. 204 3,070 .. .. 1.8 .. .. .. .. Mauritius 1,263 0.9 622 24 6,878 5,450 14,381 11,390 3.9 70 77 84 2.6 Mayotte 194 3.8c 518 .. .. ..d .. .. .. .. .. .. .. Micronesia, Fed. States 111 0.5 159 38 274 2,470 412f 3,710f 1.5 68 69 .. .. Monaco 33 0.3c 16,769 .. .. ..e .. .. .. .. .. .. .. Mongolia 2,612 1.2 2 27 3,362 1,290 8,246 3,160 8.7 66 69 98 3.4 Montenegro 600 ­1.6 43 19 3,109 5,180 6,175 10,290 7.6 72 77 .. .. Namibia 2,074 1.4 3 37 6,970 3,360 10,608 5,120 4.6 52 53 85 1.2 Netherlands Antilles 191 0.8 239 21 .. ..e .. .. .. 71 79 96 22.2 New Caledonia 242 1.8 13 26 .. ..e .. .. .. 73 78 .. 11.2 Northern Mariana 84 2.4c 182 .. .. ..e .. .. .. .. .. .. .. Islands Oman 2,600 1.1 8 32 27,887 11,120 49,487 19,740 4.6 74 77 81 12.5 Palau 20 0.8c 44 .. 167 8,210 .. .. 2.0 .. .. .. 11.9 Puerto Rico 3,943 0.5 445 21 .. ..e .. .. .. 74 83 90 0.5 Qatar 836 4.3 76 21 .. ..e .. .. 1.8 75 76 89 69.2 Samoa 187 0.7 66 40 454 2,430 667f 3,570f 2.2 68 75 99 0.8 San Marino 29 1.1j 482 .. 1,291 45,130 1,046f 37,080f 3.5 79 85 .. .. São Tomé and Príncipe 158 1.7 165 41 138 870 258 1,630 4.1 63 67 85 0.6 Seychelles 85 0.7 185 .. 762 8,960 737f 8,670f 5.8 69 76 92 6.6 Slovenia 2,018 0.2 100 14 42,306 20,960 53,756 26,640 5.5 74 81 100 8.1 Solomon Islands 495 2.5 18 40 363 730 693f 1,400f 3.2 63 64 .. 0.4 St. Kitts and Nevis 49 1.4 188 .. 470 9,630 509f 10,430f 2.5 .. .. .. 2.7 St. Lucia 168 1.1 275 27 929 5,530 1,191f 7,090f 2.0 73 76 .. 2.3 St. Vincent and the 120 0.5 309 28 507 4,210 689f 5,720f 6.2 69 74 .. 1.7 Grenadines Suriname 458 0.7 3 29 2,166 4,730 2,745f 6,000f 4.7 67 73 90 5.1 Swaziland 1,145 1.3 67 39 2,951 2,580 5,649 4,930 1.7 42 40 80 0.9 Timor-Leste 1,066 4.4 72 45 1,604 1,510 3,402f 3,190f 4.1 56 58 .. 0.2 Tonga 101 0.4 140 37 233 2,320 345f 3,430f ­4.3 72 74 .. 1.2 Trinidad and Tobago 1,333 0.4 260 21 18,795 14,100 19,441 14,580 5.6 68 72 98 24.7 Vanuatu 226 2.5 19 39 417 1,840 654f 2,890f 2.6 68 72 .. 0.4 Virgin Islands (U.S.) 108 0.0k 310 23 .. ..e .. .. .. 77 80 .. 124.3 a. Calculated using the World Bank Atlas method. b. PPP is purchasing power parity; see Definitions. c. Data are for 2003­2007. d. Estimated to be upper middle ($3,706 to $11,455). e. Estimated to be high income ($11,456 or more). f. The estimate is based on regression; others are extrapolated from the latest International Comparison Program benchmark estimates. g. Includes the island of Mayotte. h. Less than 0.5. i. Estimated to be low income ($935 or less). j. Data are for 2004­2007. k. More than ­0.05. 360 Selected world development indicators 361 Technical notes nitions differ widely; and cross-country These technical notes discuss the sources and intertemporal comparisons involve and methods used to compile the indicators complex technical and conceptual prob- included in this edition of Selected World lems that cannot be unequivocally resolved. Development Indicators. The notes follow Data coverage may not be complete because the order in which the indicators appear in of special circumstances or for economies the tables. experiencing problems (such as those stem- ming from conflicts) affecting the collec- Sources tion and reporting of data. For these The data published in the Selected World reasons, although the data are drawn from Development Indicators are taken from the sources thought to be most authorita- World Development Indicators 2008. Where tive, they should be construed only as indi- possible, however, revisions reported since cating trends and characterizing major the closing date of that edition have been differences among economies rather than incorporated. In addition, newly released offering precise quantitative measures of estimates of population and gross national those differences. Discrepancies in data income (GNI) per capita for 2007 are presented in different editions reflect included in table 1 and table 5. updates by countries as well as revisions to The World Bank draws on a variety of historical series and changes in methodol- sources for the statistics published in the ogy. Thus readers are advised not to com- World Development Indicators. Data on pare data series between editions or between external debt for developing countries are different editions of World Bank publica- reported directly to the World Bank by tions. Consistent time series are available developing member countries through the from the World Development Indicators Debtor Reporting System. Other data are 2008 CD-ROM and in WDI Online. drawn mainly from the United Nations and its specialized agencies, from the Interna- Ratios and growth rates tional Monetary Fund (IMF), and from For ease of reference, the tables usually show country reports to the World Bank. Bank ratios and rates of growth rather than the staff estimates are also used to improve cur- simple underlying values. Values in their rentness or consistency. For most countries, original form are available from the World national accounts estimates are obtained Development Indicators 2008 CD-ROM. from member governments through World Unless otherwise noted, growth rates are Bank economic missions. In some instances computed using the least-squares regression these are adjusted by staff to ensure con- method (see Statistical methods below). formity with international definitions and Because this method takes into account all concepts. Most social data from national available observations during a period, the sources are drawn from regular admin- resulting growth rates reflect general trends istrative files, special surveys, or periodic that are not unduly influenced by excep- censuses. tional values. To exclude the effects of infla- For more detailed notes about the data, tion, constant price economic indicators are please refer to the World Bank's World used in calculating growth rates. Data in Development Indicators 2008. italics are for a year or period other than that specified in the column heading--up to Data consistency and reliability two years before or after for economic indi- Considerable effort has been made to stan- cators and up to three years for social indi- dardize the data, but full comparability cators, because the latter tend to be collected cannot be assured, and care must be taken less regularly and change less dramatically in interpreting the indicators. Many factors over short periods. affect data availability, comparability, and reliability: statistical systems in many Constant price series developing economies are still weak; statis- An economy's growth is measured by the tical methods, coverage, practices, and defi- increase in value added produced by the 362 WORLD DEVELOPMENT REPORT 2009 individuals and enterprises operating in of legal status or citizenship, except for ref- that economy. Thus, measuring real growth ugees not permanently settled in the coun- requires estimates of GDP and its compo- try of asylum, who are generally considered nents valued in constant prices. The World part of the population of the country of Bank collects constant price national origin. accounts series in national currencies and Average annual population growth rate recorded in the country's original base year. is the exponential rate of change for the To obtain comparable series of constant period (see the section on Statistical meth- price data, it rescales GDP and value added ods below). by industrial origin to a common reference Population density is midyear popula- year, 2000 in the current version of the tion divided by land area. Land area is a World Development Indicators. This process country's total area excluding areas under gives rise to a discrepancy between the res- inland bodies of water and coastal water- caled GDP and the sum of the rescaled ways. Density is calculated using the most components. Because allocating the dis- recently available data on land area. crepancy would give rise to distortions in Population age composition, ages 0­14 the growth rate, it is left unallocated. refers to the percentage of the total popula- tion that is ages 0­14. Summary measures Gross national income (GNI) is the The summary measures for regions and broadest measure of national income; it income groups, presented at the end of most measures total value added from domestic tables, are calculated by simple addition and foreign sources claimed by residents. when they are expressed in levels. Aggregate GNI comprises gross domestic product growth rates and ratios are usually computed (GDP) plus net receipts of primary income as weighted averages. The summary mea- from foreign sources. Data are converted sures for social indicators are weighted by from national currency to current U.S. dol- population or subgroups of population, lars using the World Bank Atlas method. except for infant mortality, which is weighted This involves using a three-year average by the number of births. See the notes on of exchange rates to smooth the effects of specific indicators for more information. transitory exchange rate fluctuations. (See For summary measures that cover many the section on Statistical methods below for years, calculations are based on a uniform further discussion of the Atlas method.) group of economies so that the composition GNI per capita is GNI divided by mid- of the aggregate does not change over time. year population. It is converted into current Group measures are compiled only if the U.S. dollars by the Atlas method. The World data available for a given year account for at Bank uses GNI per capita in U.S. dollars to least two-thirds of the full group, as defined classify economies for analytical purposes for the 2000 benchmark year. As long as and to determine borrowing eligibility. this criterion is met, economies for which PPP gross national income, which is data are missing are assumed to behave GNI converted into international dollars like those that provide estimates. Readers using purchasing power parity (PPP) con- should keep in mind that the summary version factors, is included because nominal measures are estimates of representative exchange rates do not always reflect interna- aggregates for each topic and that nothing tional differences in relative prices. At the meaningful can be deduced about behavior PPP rate, one international dollar has the at the country level by working back from same purchasing power over domestic GNI group indicators. In addition, the estima- that the U.S. dollar has over U.S. GNI. PPP tion process may result in discrepancies rates allow a standard comparison of real between subgroup and overall totals. price levels between countries, just as con- ventional price indexes allow comparison of Table 1. Key indicators of development real values over time. The PPP conversion Population is based on the de facto defini- factors used here are derived from price tion, which counts all residents, regardless surveys covering 146 countries conducted Selected world development indicators 363 in 2005 by the International Comparison The Carbon Dioxide Information Anal- Program. For OECD countries, data come ysis Center (CDIAC), sponsored by the U.S. from the most recent round of surveys, Department of Energy, calculates annual completed in 1999; the rest are either from anthropogenic emissions of CO2. These the 1996 survey, or data from the 1993 or calculations are derived from data on fos- earlier round and extrapolated to the 1996 sil fuel consumption, based on the World benchmark. Estimates for countries not Energy Data Set maintained by the UNSD, included in the surveys are derived from and from data on world cement manufac- statistical models using available data. turing, based on the Cement Manufactur- PPP GNI per capita is PPP GNI divided ing Data Set maintained by the U.S. Bureau by midyear population. of Mines. Each year the CDIAC recalculates Gross domestic product (GDP) per the entire time series from 1950 to the pres- capita growth is based on GDP measured ent, incorporating its most recent findings in constant prices. Growth in GDP is con- and the latest corrections to its database. sidered a broad measure of the growth of Estimates exclude fuels supplied to ships an economy. GDP in constant prices can be and aircraft engaged in international estimated by measuring the total quantity transportation because of the difficulty of of goods and services produced in a period, apportioning these fuels among the coun- valuing them at an agreed set of base year tries benefiting from that transport. prices, and subtracting the cost of interme- diate inputs, also in constant prices. See the Table 2. Millennium Development Goals: section on Statistical methods for details of eradicating poverty and improving lives the least-squares growth rate. Share of poorest quintile in national con- Life expectancy at birth is the number sumption or income is the share of the of years a newborn infant would live if pat- poorest 20 percent of the population in terns of mortality prevailing at its birth consumption or, in some cases, income. It were to stay the same throughout its life. is a distributional measure. Countries with Data are presented for males and females more unequal distributions of consump- separately. tion (or income) have a higher rate of pov- Adult literacy rate is the percentage of erty for a given average income. Data are persons aged 15 and above who can, with from nationally representative household understanding, read and write a short, surveys. Because the underlying household simple statement about their everyday life. surveys differ in method and type of data In practice, literacy is difficult to measure. collected, the distribution data are not To estimate literacy using such a defini- strictly comparable across countries. The tion requires census or survey measure- World Bank staff have made an effort to ments under controlled conditions. Many ensure that the data are as comparable as countries estimate the number of literate possible. Wherever possible, consumption people from self-reported data. Some use has been used rather than income. educational attainment data as a proxy but Prevalence of child malnutrition is the apply different lengths of school attendance percentage of children under five whose or level of completion. Because definition weight for age is less than minus two stan- and methodologies of data collection differ dard deviations from the median for the across countries, data need to be used with international reference population ages caution. 0­59 months. The table presents data for Carbon dioxide emissions (CO2) mea- the new child growth standards released by sures those emissions stemming from the the World Health Organization (WHO) in burning of fossil fuels and the manufacture 2006. Estimates of child malnutrition are of cement. These include carbon dioxide from national survey data. The proportion produced during consumption of solid, of children who are underweight is the most liquid, and gas fuels and from gas flaring. common indicator of malnutrition. Being Carbon dioxide per capita is CO2 divided underweight, even mildly, increases the risk by the mid-year population. of death and inhibits cognitive development 364 WORLD DEVELOPMENT REPORT 2009 in children. Moreover, it perpetuates the specific mortality rates. The probability problem from one generation to the next, is expressed as a rate per 1,000. The main as malnourished women are more likely to sources of mortality date are vital registra- have low-birth-weight babies. tion systems and direct or indirect esti- Primary completion rate is the per- mates based on sample surveys or censuses. centage of students completing the last To produce harmonized estimates of under- year of primary school. It is calculated by five mortality rates that make use of all taking the total number of students in the available information in a transparent way, last grade of primary school, minus the a methodology that fits a regression line to number of repeaters in that grade, divided the relationship between mortality rates by the total number of children of official and their reference dates using weighted graduation age. The primary completion least squares was developed and adopted by rate reflects the primary cycle as defined by both UNICEF and the World Bank. the International Standard Classification of Births attended by skilled health staff Education (ISCED), ranging from three or are the percentage of deliveries attended four years of primary education (in a very by personnel trained to give the necessary small number of countries) to five or six supervision, care, and advice to women years (in most countries) and seven (in a during pregnancy, labor, and the postpar- small number of countries). Because cur- tum period; to conduct deliveries on their ricula and standards for school completion own; and to care for newborns. The share vary across countries, a high rate of primary of births attended by skilled health staff completion does not necessarily mean high is an indicator of a health system's abil- levels of student learning. ity to provide adequate care for pregnant Ratio of girls to boys enrollments in women. Data are from UNICEF and house- primary and secondary school is the ratio hold surveys. Good prenatal and postnatal of the female gross enrollment rate in pri- care improves maternal health and reduces mary and secondary school to the male maternal and infant mortality. But data gross enrollment rate. may not reflect such improvements because Eliminating gender disparities in educa- health information systems are often weak, tion would help to increase the status and maternal deaths are underreported, and capabilities of women. This indicator is an rates of maternal mortality are difficult to imperfect measure of the relative accessibil- measure. ity of schooling for girls. With a target date Contraceptive prevalence rate is the of 2005, this is the first of the targets to fall percentage of women married or in-union due. School enrollment data are reported ages 15­49 who are practicing, or whose to the UN Educational, Scientific, and Cul- sexual partners are practicing, any form of tural Organization (UNESCO) Institute for contraception. Safe and effective contra- Statistics by national education authorities. ception is one of the indispensable means Primary education provides children with to achieve reproductive health, helping basic reading, writing, and mathematics women avoid unintended pregnancies skills along with an elementary under- while preventing sexually transmitted dis- standing of such subjects as history, geogra- eases. Contraceptive prevalence reflects all phy, natural science, social science, art, and methods--ineffective traditional methods music. Secondary education completes the as well as highly effective modern methods. provision of basic education that began at Contraceptive prevalence rates are obtained the primary level and aims at laying foun- mainly from household surveys. dations for lifelong learning and human Prevalence of HIV is the percentage of development by offering more subject-or people ages 15­49 who are infected with skill-oriented instruction using more spe- HIV. Adult HIV prevalence rates reflect the cialized teachers. rate of HIV infection in each country's pop- Under-five mortality rate is the prob- ulation. Low national prevalence rates can ability that a newborn baby will die before be very misleading, however. They often reaching age five, if subject to current age- disguise serious epidemics that are initially Selected world development indicators 365 concentrated in certain localities or among Services value added correspond to ISIC specific population groups and threaten divisions 50­99. to spill over into the wider population. In Household final consumption expen- many parts of the developing world, most diture is the market value of all goods new infections occur in young adults, with and services, including durable products young women especially vulnerable. The (such as cars, washing machines, and home estimates of HIV prevalence are based on computers), purchased by households. It extrapolations from data collected through excludes purchases of dwellings but includes surveys and from surveillance of small, imputed rent for owner-occupied dwellings. nonrepresentative groups. It also includes payments and fees to gov- ernments to obtain permits and licenses. Table 3. Economic activity Here, household consumption expendi- Gross domestic product (GDP) is gross ture includes the expenditures of nonprofit value added, at purchasers' prices, by all institutions serving households, even when resident producers in the economy plus any reported separately by the country. In prac- taxes and minus any subsidies not included tice, household consumption expenditure in the value of the products. It is calculated may include any statistical discrepancy in without deducting for depreciation of fab- the use of resources relative to the supply ricated assets or for depletion or degrada- of resources. tion of natural resources. Value added is the Generalgovernmentfinalconsumption net output of an industry after adding up expenditure includes all government cur- all outputs and subtracting intermediate rent expenditures for purchases of goods inputs. The industrial origin of value added and services (including compensation of is determined by the International Stan- employees). It also includes most expen- dard Industrial Classification (ISIC) revi- ditures on national defense and security, sion 3. The World Bank conventionally uses but excludes government military expen- the U.S. dollar and applies the average offi- ditures that are part of government capital cial exchange rate reported by the IMF for formation. the year shown. An alternative conversion Gross capital formation consists of factor is applied if the official exchange rate outlays on additions to the fixed assets of is judged to diverge by an exceptionally the economy plus net changes in the level large margin from the rate effectively of inventories and valuables. Fixed assets applied to transactions in foreign curren- include land improvements (fences, ditches, cies and traded products. drains, and so on); plant, machinery, and Grossdomesticproductaverageannual equipment purchases; and the construction growth rate is calculated from constant of buildings, roads, railways, and the like, price GDP data in local currency. including commercial and industrial build- Agricultural productivity refers to the ings, offices, schools, hospitals, and private ratio of agricultural value added, measured dwellings. Inventories are stocks of goods in constant 1995 U.S. dollars, to the num- held by firms to meet temporary or unex- ber of workers in agriculture. pected fluctuations in production or sales, Value added is the net output of an and "work in progress." According to the industry after adding up all outputs and 1993 System of National Accounts (SNA), subtracting intermediate inputs. The indus- net acquisitions of valuables are also con- trial origin of value added is determined by sidered capital formation. the International Standard Industrial Clas- Externalbalanceofgoodsandservicesis sification (ISIC) revision 3. exports of goods and services less imports of Agriculture value added corresponds goods and services. Trade in goods and ser- to ISIC divisions 1­5 and includes forestry vices comprise all transactions between resi- and fishing. dents of a country and the rest of the world Industry value added comprises min- involving a change in ownership of general ing, manufacturing, construction, electric- merchandise, goods sent for processing and ity, water, and gas (ISIC divisions 10­45). repairs, nonmonetary gold, and services. 366 WORLD DEVELOPMENT REPORT 2009 The GDP implicit deflator reflects of earnings, other long-term capital, and changes in prices for all final demand short-term capital, as shown in the balance categories, such as government consump- of payments. Data on FDI are based on tion, capital formation, and international balance-of-payments data reported by the trade, as well as the main component, pri- IMF, supplemented by World Bank staff vate final consumption. It is derived as the estimates using data reported by the United ratio of current to constant price GDP. The Nations Conference on Trade and Develop- GDP deflator may also be calculated explic- ment, and official national sources. itly as a Paasche price index in which the Official development assistance or offi- weights are the current period quantities of cial aid from the high-income members of output. the OECD are the main source of official National accounts indicators for most external finance for developing countries, developing countries are collected from but official development assistance (ODA) national statistical organizations and cen- is also disbursed by some important donor tral banks by visiting and resident World countries that are not members of OECD's Bank missions. Data for high-income econ- Development Assistance Committee omies come from the OECD. (DAC). DAC has three criteria for ODA: it is undertaken by the official sector; it pro- Table 4. Trade, aid, and finance motes economic development or welfare Merchandise exports show the free on as a main objective; and it is provided on board (f.o.b.) value of goods provided to the concessional terms, with a grant element rest of the world valued in U.S. dollars. of at least 25 percent on loans (calculated Merchandise imports show the c.i.f. at a 10-percent discount rate). Official value of goods (the cost of the goods includ- development assistance comprises grants ing insurance and freight) purchased from and loans, net of repayments, that meet the rest of the world valued in U.S. dollars. the DAC definition of ODA and are made Data on merchandise trade come from the to countries and territories on of the DAC World Trade Organization (WTO) in its list of aid recipients. The new DAC list of annual report. recipients is organized on more objective Manufactured exports comprise the needs-based criteria than its predecessors, commodities in Standard Industrial Trade and includes all low- and middle-income Classification (SITC) sections 5 (chemi- countries, except those that are members of cals), 6 (basic manufactures), 7 (machinery the G8 or the EU (including countries with and transport equipment), and 8 (miscel- a firm date for EU admission). laneous manufactured goods), excluding Total external debt is debt owed to division 68. nonresidents repayable in foreign cur- High-technology exports are products rency, goods, or services. It is the sum of with high research and development (R&D) public, publicly guaranteed, and private intensity. They include high-technology non-guaranteed long-term debt, use of IMF products such as in aerospace, computers, credit, and short-term debt. Short-term pharmaceuticals, scientific instruments, debt includes all debt having an original and electrical machinery. maturity of one year or less and interest in Current account balance is the sum arrears on long-term debt. of net exports of goods and services, net Present value of debt is the sum of income, and net current transfers. Data are short-term external debt plus the dis- drawn from the IMF's Balance of Payments counted sum of total debt service pay- Statistics Yearbook. ments due on public, publicly guaranteed, Foreign direct investment is net inflows and private nonguaranteed long-term of investment to acquire a lasting manage- external debt over the life of existing ment interest (10 percent or more of vot- loans. Data on external debt come mainly ing stock) in an enterprise operating in an from reports to the World Bank through economy other than that of the investor. It its Debtor Reporting System via member is the sum of equity capital, re-investment countries that have received IBRD loans Selected world development indicators 367 or IDA credits, with additional informa- The least-squares growth rate, r, is esti- tion from the files of the World Bank, the mated by fitting a linear regression trend- IMF, the African Development Bank and line to the logarithmic annual values of the African Development Fund, the Asian variable in the relevant period. The regres- Development Bank and Asian Devel- sion equation takes the form opment Fund, and the Inter American Development Bank. Summary tables of the ln Xt = a + bt external debt of developing countries are which is equivalent to the logarithmic published annually in the World Bank's transformation of the compound growth Global Development Finance. equation, Net migration is the total net number of migrants during the period, that is, the Xt = Xo (1 + r)t number of immigrants less the number of emigrants, including both citizens and non- In this equation, X is the variable, t citizens. Data shown in the table are five- is time, and a = log Xo and b = ln (1 + r) year estimates. Data are from the United are the parameters to be estimated. If b* Nations Population Division's World Popu- is the least-squares estimate of b, the aver- lation Prospects: The 2006 Revision. age annual growth rate, r, is obtained as Domestic credit provided by banking [exp(b*) ­ 1] and is multiplied by 100 to sector includes all credit to various sec- express it as a percentage. tors on a gross basis, with the exception of The calculated growth rate is an average credit to the central government, which is rate that is representative of the available net. The banking sector includes monetary observations over the entire period. It does authorities, deposit money banks, and not necessarily match the actual growth other banking institutions for which data rate between any two periods. are available (including institutions that do not accept transferable deposits but do Exponential growth rate incur such liabilities as time and savings The growth rate between two points in time deposits). Examples of other banking insti- for certain demographic data, notably labor tutions include savings and mortgage loan force and population, is calculated from the institutions and building and loan associa- equation tions. Data are from the IMF's International r = ln (pn /p1)/n Finance Statistics. where pn and p1 are the last and first obser- Table 5. Key indicators for other vations in the period, n is the number of economies years in the period, and ln is the natural See Technical notes for table 1, key logarithm operator. This growth rate is indicators. based on a model of continuous, exponen- tial growth between two points in time. It Statistical methods does not take into account the intermedi- This section describes the calculation of the ate values of the series. Note also that the least-squares growth rate, the exponential exponential growth rate does not corre- (endpoint) growth rate, and the World spond to the annual rate of change mea- Bank's Atlas methodology for calculating sured at a one-year interval which is given the conversion factor used to estimate GNI by and GNI per capita in U.S. dollars. (pn ­ pn )/pn ­ 1 ­ 1 Least-squares growth rate Least-squares growth rates are used wher- World Bank Atlas method ever there is a sufficiently long time series to In calculating GNI and GNI per capita in permit a reliable calculation. No growth U.S. dollars for certain operational pur- rate is calculated if more than half the poses, the World Bank uses the Atlas con- observations in a period are missing. version factor. The purpose of the Atlas 368 WORLD DEVELOPMENT REPORT 2009 conversion factor is to reduce the impact The following formulas describe the cal- of exchange rate fluctuations in the cross- culation of the Atlas conversion factor for country comparison of national incomes. year t : The Atlas conversion factor for any year is the average of a country's exchange rate (or alternative conversion factor) for that year and its exchange rates for the two preceding years, adjusted for the difference between and the calculation of GNI per capita in the rate of inflation in the country and that U.S. dollars for year t: in Japan, the United Kingdom, the United States, and the Euro Zone. A country's Yt = (Yt/Nt)/et* $ inflation rate is measured by the change in its GDP deflator. The inflation rate for where et* is the Atlas conversion factor Japan, the United Kingdom, the United (national currency to the U.S. dollar) for States, and the Euro Zone, representing year t, et is the average annual exchange rate international inflation, is measured by the (national currency to the U.S. dollar) for S$ change in the SDR deflator. (Special draw- year t, ptis the GDP deflator for year t, pt is ing rights, or SDRs, are the IMF's unit of the SDR deflator in U.S. dollar terms for $ account.) The SDR deflator is calculated year t, Yt is the Atlas GNI per capita in U.S. as a weighted average of these countries' dollars in year t, Yt is current GNI (local GDP deflators in SDR terms, the weights currency) for year t, and Nt is the midyear being the amount of each country's cur- population for year t. rency in one SDR unit. Weights vary over Alternative conversion factors time because both the composition of the The World Bank systematically assesses the SDR and the relative exchange rates for each appropriateness of official exchange rates as currency change. The SDR deflator is calcu- conversion factors. An alternative conver- lated in SDR terms first and then converted sion factor is used when the official exchange to U.S. dollars using the SDR to dollar Atlas rate is judged to diverge by an exceptionally conversion factor. The Atlas conversion fac- large margin from the rate effectively applied tor is then applied to a country's GNI. The to domestic transactions of foreign curren- resulting GNI in U.S. dollars is divided by cies and traded products. This applies to the midyear population to derive GNI per only a small number of countries, as shown capita. in the primary data documentation table in When official exchange rates are World Development Indicators 2007. Alterna- deemed to be unreliable or unrepresenta- tive conversion factors are used in the Atlas tive of the effective exchange rate during methodology and elsewhere in the Selected a period, an alternative estimate of the World Development Indicators as single-year exchange rate is used in the Atlas formula conversion factors. (see below). Index accessibility index, Mexico, 244b8.6 economic activity link to development stages in, 83­84, 85t2.1 accidents, 191­92 as geographic scale, 3b1 ACP. See African, Caribbean, and Pacific (ACP) countries local 1-square kilometer areas, 60 administrative areas, economic activity link to development stages as units for government action, 201­2 in, 83­84, 85t2.1 Association of Southeast Asian Nations (ASEAN), 195, 196 ad valorem transport charges, 179 autarky, and labor mobility, 149 advanced urbanization, 25map6c, 26­27, 201, 202, 223­29 Africa, 265b9.3 disparities in lagging areas of, 80b2.3 backyard capitalism, 135 language diversity in, 104­5 balanced industrial growth, 90­91b2.5 life expectancy in, 111f3.12 Balkan region, 274 tariffs in, 99, 99f3.3 Bamako, Mali, 48­49 transport sector, 186, 187t6.1 Bangkok, transport infrastructure, 223b7.12 African, Caribbean, and Pacific (ACP) countries, EPAs with EU, barriers to trade, 99 265b9.3 Belgium, 50 agglomeration economies, 13­15, 26, 29, 34b0.1, 126 Bogotá, Colombia, 21, 22, 223, 225map7.5 agriculture disperses and manufacturing clusters in, 132­33 Bolivia, border divisions in, 102b3.2 amplified by density and attenuated by distance, 134­35 Bombay, India, slums in, 204, 205b7.3 changes in to adapt to market forces, 142­44 borders China, 127b4.1 Africa, 187 cities facilitation of, 134 artificial borders, 105b3.4 clustering of services vs. clustering of manufacturing in, Bolivia, 102b3.2 133­34 calculation of, 55b1.3 computing index for, 55b1.3 contrasted with divisions, 97­98 essentials of, 129 costs of border-crossing procedures, 189­90 examples of, 128t4.1 cross-border migration, 152­53 Indonesia, 131­32 cross-border production networks, 266 and knowledge production, 184 and economic impact of conflicts on, 105 and mobility factors, 85 fragmentation of, 12, 13f3c multicountry, 279­80 and growth spillovers, 102b3.1 principles for managing, 200­202 increase in, 96, 97f3.1 Sub-Saharan Africa, 284­86 and institutional cooperation, 266­67 theoretical advances recognized, 136­37 integration of pairs of large cities separated by national borders, and theories of labor mobility, 158­59 279­80 variations in by city size and profile, 137­38 link of border restrictions to economic development, 97­98, agglomeration indexes, 54­56, 57f1.3, 60 99f3.2 agriculture sector, 253 negative influence on trade, 46 Belgium, 50 openness of, 119­20 dispersing of as producers seek scale economies, 132­33 and patterns of labor mobility, 149­53 link to towns, 53­54 thickness of as obstacle to development, 265 trade of agriculture products as entry point for regional and transport costs in Africa, 187 integration, 280 brain drains, 18, 147, 168, 168b5.9 air pollution, 185b6.5, 190­91 brain gains, 168 air transport costs, 176 Brazil antidecentralization, 231b8.1 constitution and national unity in, 236b8.3 Arab Maghreb Union, 275 economic integration in, 241 areas internal migration in, 167b5.8 concept of, 230 land use in, 204­5 definition of subnational areas, 78b2.1 local and national incentives in, 257 description of, 35 migration in, 15, 17 development policies of to integrate nations, 27­29 poor people in, 232­33 369 370 WORLD DEVELOPMENT REPORT 2009 Brazil (continued) increase in population of, 12, 13b31 poverty rates in, 28map7b, 232­33 integration of pairs of large cities separated by national borders, road networks, 251 279­80 Superintendency for the Development of the Northeast, integration of slums into, 227­29 38 new city initiatives, 144­45 Breaking New Ground, 214 patterns of specialization and diversity among, 137­40 Brussels, Belgium, 50 population of, 49­50 business services, imports in, 180­81 size of related to plant size, 21b4 busways, 210, 223, 225map7.5 and telecommunication costs, 140b4.5 buyer-supplier networks, 177b6.2 urban density of, 45­46 world cities, 137 Cities Alliance, 229b7.15 Cambodia, rural-urban transformation in, 218, 219b79 civil wars, economic costs of, 105 Cameroon, rural-urban integration in, 220b7.11 climate change, and urbanization, 211b7.6 Canada, 46, 79b2.2, 86 coastal areas, 93­94, 94map2.6 capital, 99­100 cohesion policies, 236­37 divisions as barrier to movement of, 171 Colombia, freight transport, 251­52 and economic development, 84­85 commodities, export of, 118­19 capitalism, backyard capitalism, 135 communal systems capital mobility, 146­49, 171 land, 219 capital restriction index, 100f3.4 property rights, 203­4 the Caribbean, 274­75 communication costs, 13, 18­20, 180, 184 education in, 156f5.4 commuter rail systems, 208b7.4 transport costs, 192b6.7 competition Caribbean Community (CARICOM), 274­75 and the invisible hand, 135­36 Caribbean Knowledge and Learning Network, 275 and network economies, 186 Caribbean Regional Negotiating Machinery, 274­75 within same industry, 130­31 CARICOM. See Caribbean Community (CARICOM) compulsory purchases, 207 Cassa per il Mezzogiorno, 184b6.4 computer and electronics sector, 127b4.1 Central American Common Market, 270 concentration Central American Electricity Connection System, 270 conflicts and, 106 Central Asia, integration in, 277, 278b9.4 and economic geography, 41b0.4 Central European Free Trade Agreement, 274 of economic mass, 97, 238­40 checkpoints, Africa, 187 of economic production, 8­10 Chile, border divisions in, 102b3.2 geographic, 129b4.2 China measures of, 60­62 and exporting industries, 94, 94map2.6 pace and pattern of, 57­60 inequalities in per capita provincial gross product of, 89, principles for managing, 200­202 89f2.10 promotion of in Japan, 208, 209b7.5 intermediate urbanization in, 221­22 and transport costs, 183­84, 185b6.5 investments in infrastructure, 227b7.14 See also economic concentration migration of workers in, 13, 16map3, 17, 18 conflicts, 105, 106 mobility with density in Hong Kong, China, 185b6.5 congestion, 190 patterns of labor mobility in, 154­55b5.3 in intermediate urbanization, 220­23 poverty rates, 28map7a reduction of, 207­8, 209b7.5 spatial integration in, 213 See also concentration special economic zones in, 224b7.13, 254­55b8.10 constitutions, emphasis on national unity in, 236b8.3 trade policies, 76 consumption, 166 and urbanization, 26b5, 217­18, 226 convergence of, 10­12 and urban renewal and redevelopment in, 142b4.6 in developing countries, 141­42 circular causation, 171­72 and economic development, 2 cities increase in for urban residents, 8­9 changes in to adapt to market forces, 142­44 in leading areas, 9 China, 141­42 rural-urban disparities in, 64­65 city-based growth, 137 containerization, 177­79 empirical regularities of city-size distributions, 52b1.1 corruption, 186, 274 facilitation of scale economies in, 134 Costa Rica, urban housing shortage, 214 and importance of diversity and specialization in, 139b4.4 cost discovery, 263b9.2 Index 371 costs developed countries distance-related, 75, 77 impact of migration on, 161­68 of doing business, 6b2 and regional and global integration, 264­65 economic costs of conflict and territorial disputes, 105 developing economies, 128 shipping, 176­79 clustering of services vs. clustering of manufacturing in, 133­34 See also communication costs; transport costs consumption in, 141-42 Côte d'Ivoire, 236b8.3 differences in, 93­95 cotton, 280, 280map9.6 dynamics of economic geography on, 120­21 countries economic growth in, 159­60 with bigger and denser places, 61 impact of migration on, 161­68 description of, 35 labor mobility patterns in, 155, 156t5.3 and distance challenges for integration, 275­77 new cities in, 224b7.13 as geographic scale, 3b.1 patterns of urbanization in, 71­72 impact of divisions on, 98­101 regional and global integration in, 262­65 impact of transport costs on concentration within, 183­84, scale economies in, 132­33 185b6.5 slums in, 4map1, 5 integration options for market access, 273­75 spatial disparities in living standards in, 87­89 resource-poor, 102b3.1 and transport and trade costs, 173 sea-locked, 103­4 transport policies in, 184­92 size of, 101­3, 103b3.3 development taxonomy to assist integration policies in, 241 assessing performance of policies for, 259t8.4 See also landlocked countries attributes of, 2­3 cropland per farmer, 132 challenges of, 3­5 cross-border migration, 152­53 and concentration of economic wealth, 115­20 cross-border production networks, 266 and density concentration, 60­62 cultural divisions, 104­5 dimensions of, 6­8 cultural facilities, 51, 52 and distance, 6b2, 17, 36-37 cumulative causation, 2, 284 and divergence and convergence of slums within cities, 66­70 customs regimes, 188 and economic concentration, 57­60 findings concerning, 49 framework for policies that promote, 246t8.1 Daegu, Korea, 220­23 and international comparisons of economic concentration, decentralization, 221­22, 231, 257 83­84 vs. antidecentralization, 231b8.1 rural-urban disparities in, 62­66 demand. See supply and demand spatially balanced, 73 demographic development, 238 stages of related to spatial concentration of economic activity, Deng Xiaoping, 73 82­83 Denmark, land reform to facilitate urbanization in, 203b7.2 strategies for Africa, 282­83 density diseconomies, 144 agglomeration economies amplified by, 134­35 distance concentration link to development, 60­62 advantage of locations close to markets, 76­77 definition, 49­56 agglomeration economies attenuated by, 134­35 as dimension of development, 6b2, 7 challenges of for integration options for countries, 275­77 hierarchy of, 49­50 countries distant from world markets with small neighbors, institutions and infrastructure to overcome, 242­45, 246t8.1 277­83 in leading areas compared to distance in lagging areas, 78­80 as dimension of development, 6b2, 7, 36­37 link of demand management and public transportation to, 208, in East Asia, 194­96 210­12 as an economic concept, 75, 76map22 measures of, 54­56 findings concerning, 74 mobility with density in Hong Kong, China, 185b6.5 impact on trade, 108­9 movement toward, 15 institutions and infrastructure to overcome, 241­45, 246t8.1 overview, 48­49 migration as way to reduce, 77­78 of population in United States, 44 openness matters for, 94 reduction of during intermediate urbanization, 220­23 as spatial dimension, 36­37 as spatial dimension, 36­37 spatially connective infrastructure to reduce distance to density, spatially connective infrastructure to reduce distance to, 206­12 206­12 in Sub-Saharan Africa, 284­86 in Sub-Saharan Africa, 284­86 See also economic density as trade deterrent, 182 372 WORLD DEVELOPMENT REPORT 2009 diversity spatially blind institutions to facilitate, 202­6 industrial diversity and urbanization, 131­32 in three geographic scales, 8­10 and specialization in cities, 137­40 See also density divisions economic development, 97 and aspects of regional concentration on wealth, 115­20 impact of geography, globalization, and development on, as barriers to movements of capital and labor, 171 115­20 beyond control of countries, 101­5 importance of location in, 1­2 and capital flows, 99­100 link to malaria, 116, 117b3.7 contrasted with borders, 97­98 spatial disparities in, 39b0.3 definition of, 97­105 See also development as dimension of development, 6b2, 7 Economic Development Administration, U.S. Commerce in East Asia, 194­96 Department, 256 and economic concentration, 105­9 economic forecasting, 198­99 and economic costs of conflict and territorial disputes, economic geography, 115, 173, 200b7.1 105 and assistance for Africa, 282­83 and economic geography, 120­21 and concentration, 41b0.4 ethnic and cultural, 104­5 and convergence, 41b0.4 and flow of goods and services, 98­99 impact on economic development, 120­21 impacts on countries, 98­101 incentives to exploit, 252­53 institutions to overcome, 244­45 integration, 41b0.4 main findings concerning, 96­97 and trade, 263­64 as spatial dimension, 36­37 and transport costs, 181­82 in Sub-Saharan Africa, 284­86 United States, 44­45 as trade deterrent, 182 See also geography Doha Round, 187­88 economic growth, 128 domestic disparities, 5 impact of migration on, 159­61, 161b5.5 domestic transport, 173­74 and labor mobility, 162, 163b5.6, 164f5.8 Dongguan, China, scale economies in, 126, 127b4.1 at national geographic scale, 3b1, 5 reformulation of models of, 159, 160 spatial balancing of, 5­6 EAC. See East African Community (EAC) economic integration, 1 earnings. See income; wages and area development policies, 27­29 East African Community (EAC), 152b5.1 Bogota, 21, 22 East Asia impact of openness on, 119­20 average annual growth rate of GDP, 110f3.11, 112­13 political benefits of, 258­59 disparities in lagging areas of, 80b2.3 promotion of, 2, 22­24 distance and division in, 194­96 strategies and priorities for, 272, 273t9.4 economic growth in, 89, 89f2.8 Turkey, 21, 22 and income divergence then convergence, 112­13 West Africa, 21­22 Eastern Caribbean Telecommunications Authority, 275 See also integration Eastern Europe economic mass, 97, 238-40 average annual GDP growth, 113­14 Economic Partnership Agreements (EPAs), 265, 265b9.3 economic growth in, 89, 89f2.8 economic production, concentration of, 8­10 impact of labor and social policies on migration, 160b5.4 economies of scale. See scale economies income convergence in, 113­14 ECOWAS. See Economic Community of West African States income disparities, 91f2.11 (ECOWAS) economic analysis, insights from, 21b4 education Economic Community of West African States (ECOWAS), 21­22, convergence of as incomes diverge, 92­93 32, 152b5.1 impact on welfare, 247 economic concentration improvements in, 110­11 and divisions, 105­9 inequalities in access to, 111­12 historical perspective, 106­8 and internal migration, 156f5.4, 157f5.5 international comparisons of, 83­84 and mobility-enhancing regional infrastructure, 268 pace and pattern of, 57­60 and patterns of labor mobility, 155­57, 158t5.6 See also concentration Uganda, 247, 248b8.7 economic density, 195­96 Egypt, Arab Republic of, 224b7.13, 243, 245map8.6 and population density, 29 eminent domain, 44, 207 problems of in large cities, 142­44 emissions, 190, 191f6.6, 211b7.6 Index 373 employment geography density of, 54 and access to global markets, 29­32 rates of, 164­65 and concentration of economic wealth, 115­20 enclosure movement, 203 and disparities in income and welfare in developed countries, endogenous growth model, 159 85­87 EPAs. See Economic Partnership Agreements (EPAs) first-nature geography, 116, 117b3.7 ethnic divisions, 104­5 natural geography, 84 ethnic heterogeneity, 242 physical geography, 115, 173, 240, 263­64 ethnic minorities, disparities among, 80, 80b2.3 second-nature geography, 116, 118 EU. See European Union (EU) unevenness of, 2 Euclidean distances, 75 See also economic geography Euro-Mediterranean Forum, 273­74 Germany, 78b2.1, 235b8.2 Europe, compared to Asia, 194 Ghana, 253 European Union (EU), 236­38, 2324 Gibrat's law, 51, 52b1.1 criteria used to define lagging areas in, 79b2.2 globalization, 96, 97f3.1, 128t4.1 EPAs with African, Caribbean, and Pacific countries, 265b9.3 and concentration of economic wealth, 115­20 GDP per capita, 62 financial, 100 inequalities of market incomes in, 238 increase in capital mobility during, 147­49 Ireland's use of EU funds for international convergence, 237­38 and patterns of labor mobility, 153 and regional integration, 274 and transport costs, 170, 175 territorial policy, 39b0.3 global markets. See world markets Everything But Arms initiative, 38 Global Rural-Urban Mapping Project (GRUMP), 55b1.3 exports, 99 goods and services delivery, link to border permeability, 98­99 of commodities, 118­19 goods-producing industries, 130 connecting to markets, 251­52 governance, 274 and distance to markets, 6b2 Greater Mekong Subregion (GMS), cross-border migration in, perishable goods, 177b6.2 153b5.2 regional and global integration of, 262­65 gross domestic product (GDP), 49, 62, 194 expropriation risks, 204 and density of economic activity, 78b2.1, 83, 83f2.2 external scale economies, 127b4.1, 128t4.1 and economic concentration, 57­60, 105­6 generated by commuters, 56 and geographic disparities, 95 factor mobility, 241, 258, 285­86 Germany, 78b2.1 final goods, 20b3 global trade as share of, 12, 13f3b financial sector, 52, 133, 251, 266 growth of, 30f7, 96, 97map3.1, 110f3.11, 112­13, 236­37 First Law of Geography, 76 increase in from 1500-1998, 109t3.2, 110 first-nature geography, 116, 117b3.7 and landlocked countries, 102b3.1 Fiscal Investment and Loan Program, Japan, 87b2.4 and relationship between economic activity and stages of Floor Space Index (FSI), 205b7.3 development, 83­84 foreign aid, 5 and residential mortgage markets, 206 foreign born population, 150 shares of world GDP, 107f3.8 foreign direct investments (FDIs), 148f5.2, 149, 161, 266, 274 Southeast Asia, 89, 89f2.9 foreign spaces, unequal access to, 100 spatial distribution of, 39b0.3 France Sub-Saharan countries, 265b9.3 divergence then convergence of income and welfare in, 86 trade as share of, 93f2.13, 94 increasing spatial concentration and decreasing spatial gross national income (GNI), link to migration, 100 disparities in, 238, 238f8.1 gross regional product (GRP), 90­91b2.5 size of cities in, 51 GRUMP. See Global Rural-Urban Mapping Project (GRUMP) free trade, 260 freight costs, 175­76, 192b6.7, 251­52 FSI. See Floor Space Index (FSI) health convergence of as incomes diverge, 92­93 improvements in, 110­11 GDP. See gross domestic product (GDP) as indicator of well-being of urbanites, 72 geographic concentration, reasons firms locate closer to each and road fatalities, 191­92 other, 129b4.2 in slums, 67 geographic information systems, 54­56, 57f1.3 heavy industries, and internal scale economies, 129­30 geographic scales, 3­6 high-income countries, and density concentration, 61 374 WORLD DEVELOPMENT REPORT 2009 high-technology industries, 133 disparities in Eastern Europe, 91f2.11 historical accidents, 233 disposable income, 62 home places, 159 divergence of, 10­12, 112f3.14 Honduras, poverty rate in leading and lagging areas, 82map2.5 as health and education converge, 92­93 Hong Kong, China between richest and poorest countries, 110­11 mobility with density in, 185b6.5 then convergence of in developed countries, 85­87 spatial integration in, 213 East Asia, 195 and urban renewal and redevelopment in, 142b4.6 European Union, 238 hosiery industry, 131 Germany, 235b8.2 household consumption, 2, 10­12, 166 improvements in, 110 household registration, 76, 154­55b5.3 as indicators of well-being of urbanites, 72 household surveys, 88­89 Latin America, 112, 115b3.8 housing predictors of, 1 costs of, 205b7.3 spatial inequalities of in fast-growing countries, 89­92 and land use, 204 and Todaro class of migration models, 159­60 occupant-owned, 206 United States, 165b5.7 regulations for finance of, 206 upper-middle class, 8 shortage of, 214 income tax systems, United States, 238­39 Sweden, 215b7.8 India See also slums barriers to labor mobility in, 162, 163b5.6 hukou household registration system, 154­55b5.3 criteria used to define lagging areas in, 79b2.2 human capital, 146, 157 economic integration in, 241 external effects from clustering of, 160 federal transfers in, 248 investments in, 245­50 Indian Railways, 175b6.1 movement of skilled workers, 21b4 labor mobility in, 29 See also labor mobility lagging areas in, 233­34 human development index, 92, 246, 247f8.2 licensing systems, 256 hydropower, 267 migration in, 15, 17 poverty rates, 28map7c slums in, 204, 205b7.3, 227 ICT. See information and communication technology (ICT) spatially balanced growth in, 233 ideas special economic zones in, 254­55b8.10 idea-driven economies, 137 Indonesia link to division changes in countries, 100­101 agglomeration economies in, 131­32 immigration policies, 157 and distance to economic centers, 76 imports, business services, 180­81 rural-urban integration in, 220b7.11 Improved Road Transport Governance Initiative, West Africa, 187 industrialization incentives, 22­24, 25, 231b8.1, 256 and acceleration of urbanization, 61 for advanced urbanization, 223­26 in China, 154­55b5.3 area incentives, 256­57 link to labor mobility, 149 coordination of, 257, 266, 269­73 policies for, 24, 27 design of, 257­58 Soviet Union, 90­91b2.5 to facilitate regional integration, 281­83 industries financial, 253­55 industrial diversity, 138 for integrating slums into cities, 227­29 industrial zones, 224b7.13 for migration, 77 location patterns of, 138­40 to overcome distance, density, and divisions, 244­45 market-oriented, 94map2.6 and policy framework for economic integration, 240 spatial concentration of, 133 spatially focused, 37­39, 253-57 vertically integrated industries, 256 that exploit geographic advantages, 252­57 information, freedom of access to, 101 incipient urbanization, 25­26, 201, 202, 204, 216­20 information and communication technology (ICT), 77, 138, 251 inclusive urbanization, 229 export of, 101 income, 2, 6, 8, 45, 46 New York City, 143b4.7 convergence of, 10­12, 235b8.2 information technology (IT), 119, 188, 189 divergence then convergence of in developed countries, infrastructure, 22­24, 25, 173, 231b8.1 85­87 for advanced urbanization, 223­26 in faster-growing regions, 112­15 China, 26b5 in middle-income countries, 92 to connect lagging and leading areas, 250­52 Index 375 cross-border, 267­68 of neighboring markets, 105 evolvement of, 207 options for countries distant from world markets with small for increasing density and reducing congestion, 220­23 neighbors, 277­83 as instrument for integration, 37­39 of Pacific islands with world markets, 277, 278b9.5 for integrating slums into cities, 227­29 policy framework for integrating lagging and leading areas, intercity, 208 238­45, 246t8.1 interregional, 252 political benefits of economic integration, 258­59 investments in, 176­77 principles for managing, 200­202 Italy, 184b6.5 regional and global integration, 262­65 natural infrastructure, 126 strategies to integrate poor countries with world markets, to overcome distance and density, 242­45, 246t8.1 260­62 policies in Bombay, 205b7.3 urbanization framework in action, 216­29 and policy framework for economic integration, 240 See also economic integration; regional integration; rural-urban as prerequisite for spatially-targeted interventions, 212 transformations regional, 266, 267­69 intergovernmental transfers, 248, 249b8.8 and spatial integration in Hong Kong, China, 213 intermediate goods, 20, 22, 182 and targeted interventions, 214­15 intermediate inputs, 19­20 and transport providers, 186­87 intermediate urbanization, 25map6b, 26, 201, 202, 220­23 See also spatially connective infrastructure; transport internal economies, 128t4.1, 129 infrastructure internal labor mobility, 153­55, 156t5.3 innovations, 131­32, 135 comparison of internal migrants and native workers, 162, input-sharing, 130­31 164f5.8 institutions, 22­23, 25, 231b8.1 and economic growth, 162 for advanced urbanization, 223­26 India, 162, 163b5.6 China, 26b5 relationship of skills and education to, 155­57, 158f5.6 for connective infrastructure, 206­7 See also labor mobility to connect lagging and leading areas, 250­52 internal migration, Brazil, 167b5.8 evolvement of, 207 internal scale economies, 127b4.1, 128t4.1 for increasing density and reducing congestion, 220­23 in heavier industries, 129­30 institutional mechanisms used to overcome physical distances in light industries, 130, 132b4.3 in United States, 44­45 International Development Association, 281 institutional reforms, 274­75 international geographic scale, 3­5, 6b2, 7, 8, 10 as instrument for integration, 37­39 international labor mobility, relationship of skills and education for integrating slums into cities, 227­29 to, 155­57, 158f5.6 for integration in Central Asia, 277, 278b9.4 Internet, 138 land market institutions for rural-urban integration, 220b7.11 interregional infrastructure, 252 for more efficient rural-urban transformations, 216­20 interregional trade, East Asia, 194­95 to overcome density, distance, and divisions, 241­45, 246t8.1 interstate highway system, 46 and policy framework for economic integration, 240 interventions, 25, 26 as prerequisite for spatially targeted interventions, 212 as instrument for integration, 37­39 to promote portable investments, 245­50 to integrate the urban poor, 228t7.2 reforms to improve regional integration, 275­76 link to institutions and infrastructure, 214­15 for slum upgrading and prevention, 229b7.15 soft interventions, 255­56 and spatial integration in Hong Kong, China, 213 spatially targeted, 23­24, 37­39, 42, 238, 242 and targeted interventions, 214­15 intervento straordinario, 184b6.4 See also spatially blind institutions intraindustry trade, 19­20, 22, 136­37 instruments, for integration, 37­39, 271, 273b9.4 and communication costs, 180­81 integrated firms, 138 historical perspective, 171, 171f6.1, 182, 183f6.5 integration and increase in transport intensity, 175­79 antiglobal integration, 261b9.1 investments, 25 in Central Asia, 277, 278b9.4 FDIs, 148f5.2, 149, 161 challenges of distance and division in, 195­96 in industrial sector, 255 in China, 227b7.14 in infrastructure, 184b6.4 and economic geography, 41b0.4 in China, 227b7.14 framework for, 202­16, 265­73 cross-country, 276­77 in Germany, 235b8.2 institutions that promote, 245­50 institutional mechanisms for, 271 in research and development, 101 instruments for, 37­39 in social services, 87b2.4 376 WORLD DEVELOPMENT REPORT 2009 Iran, Islamic Republic of, 63b1.5, 250­51 income disparities in, 89, 89f2.9 Ireland, international convergence using EU funds, 237­38 India, 233­34 isolationism, 194 institutions and infrastructure to connect to leading areas, Istanbul, Turkey, 225­26 250­52 Italy, and transport costs, 184b6.4 integration of, 27­29 living standards in, 247f8.2 Mexico, 244b8.6 Jacobs, Jane, 49 and mobile phones, 251 Japan policy framework for integrating lagging and leading areas, divergence then convergence of income and welfare in, 86­87 238­45, 246t8.1 Fiscal Investment and Loan Program, 87b2.4 poverty rates in, 81, 82map2.5 spatially connective policies to promote concentration, 208, seeking unity with leading areas, 234­38 209b7.5 and welfare indicators, 73 jet engines, 176, 177b6.2 Lagos, Nigeria, 48 juice tankers, 176­77 land and land use, 15, 26 and building regulations, 204-6, 226­27 costs of in Russia, 241­42 Kazakhstan, 52b1.1 definition and enforcement of land rights, 218 Kenya, disparities in lagging areas of, 80b2.3 and diseconomies, 144 knowledge and economic activity link to development stages of, 83­84, access to, 95 85t2.1 production of, 184 and exclusion of poor, 228 sharing of, 183­84 Hong Kong, China, 213 spillovers of, 134, 158 as immobile factor to production, 142 Korea, Republic of, 53b1.2, 246 improving administration of, 218­20 infrastructure development project in, 220­22 institutions for land markets, 202, 203­4, 220b7.11 ranking of industries in, 133 management of, 202­3 rural-urban integration in, 220b7.11 markets for, 22 spatially targeted incentives in, 253 policies for, 25 urbanization in, 216­17, 224­25 Land Consolidation Program, Indonesia, 220b7.11 Kuznets, Simon, 33 landlocked countries, 101, 102b3.1, 102b3.2, 103map3.3 and beyond the border trade, 267 development strategies for, 282­83 labor markets, 22 impact of openness and integration on, 119­20 labor mobility, 18, 147, 171 integration options for, 277­83 barriers to in India, 162, 163b5.6 and time costs of border crossings, 190 conclusions from analysis of, 158­68 LandScan, 55b1.3 and economic growth, 84­85, 162, 163b5.6, 164f5.8 language diversity, 104­5 impact of policies on migration, 169b5.4 Lao People's Democratic Republic, 242, 243map8.5, 246, 247f8.2 India, 29 large cities, patterns of specialization and diversity among, 138­40 international, 155­57, 158f5.6 Latin America and migration theory, 159­61, 161b5.5 education in, 156f5.4 and mobility-enhancing regional infrastructure, 268 per capita incomes, 112, 115b3.8 overview, 146­47 and transport costs, 192b6.7 patterns of, 149­53, 155, 156t5.3 leading areas, 22, 74 and RTAs, 266 consumption in, 9 in Sub-Saharan Africa, 151­52, 281 density in compared to distance for lagging areas, 78­80 voluntary migration, 169 and differences in modern developing countries, 93­95 See also capital mobility; internal labor mobility divergence then convergence of economic activity in, 84­85 lagging areas, 22, 24, 29 economic concentration in, 81­84, 85t2.1 criteria used to define, 79b2.2 impact of internal labor migration on, 162­65 development policies for, 230­31 income disparities in, 89, 89f2.9 and differences in modern developing countries, 93­95 institutions and infrastructure to connect to lagging areas, distance in compared to density in leading areas, 78­80 250­52 and distribution of poor people, 232­33 integration of, 27­29 divergence then convergence of economic activity in, 84­85 living standards in, 247f8.2 Egypt, 245map8.6 policy framework for integrating lagging and leading areas, impact of internal labor migration on, 162­65 238­45, 246t8.1 Index 377 poor people in, 81, 82map2.5 advantage of distance to, 76­77 production in, 9 changes in cities to adapt to forces of, 142­44 seeking unity with lagging areas, 234­38 connecting producers to markets by mail, 251­52 Lewis, Arthur, 161b5.5 country scores for access to, 270­73 licensing systems, 256 as drivers of density and migration, 45 life expectancy index, 92 forces of, 13­20 life expectancy, 110, 111, 111f3.12 global nature of, 12, 13f3b light industries, 130, 132b4.3, 133 impact of policy makers on, 128 liner shipping, 176 importance of market access, 108­9 literacy rates, 92, 93 institution cooperation for, 265, 266­67 living standards, 1 integration options for countries close to world markets, convergence of, 10­12, 45­46, 73 273­75 disparities within cities, 66­70 market forces, 128, 130­41 divergence in, 3­5, 10­12 market-oriented industries, 94map2.6 findings linking to development, 49 Mexico, 244b8.6 impact of migration on, 165 preferential access to, 282-83 Malaysia, 27 size of, 41b0.4 spatial disparities in, 39b0.3, 87­89 and spatial dimensions, 36­37 spatial equities in, 238 and spatially targeted incentives, 253­57 United States, 73 See also world markets variations in leading and lagging areas, 247f8.2 Marshall, Alfred, 134 living standard surveys, and household consumption, 2 mass rapid transit, 210 local geographic scale, 3­5 McLean, Malcom, 178b6.3 and concentration of economic production, 8­9 MDGs. See Millennium Development Goals (MDGs) and density, 6b2, 7 mega-cities, 140­41 localization economies, 14, 21b4, 128t4.1, 129, 221­22 mercantile period Indonesia, 132b4.3 capital mobility during, 147­49 and industry patterns, 137­40 labor mobility during, 149­53 from input-sharing and competition within the industry, mercantile trade industry, 142 130­31 metropolitan economy, 49 locations Metropolitan Transportation Authority, New York City, 208b7.4 advantages of distance to markets, 76­77 Mexico, 46 and prosperity, 1­8 criteria used to define lagging areas in, 79b2.2 See also distance industrial development in, 256 logistics, and time factors in international trade, 179­80 market access in lagging areas of, 244b8.6 low-income countries, and density concentration, 61 Mezzogiorno, 184b6.4 Lucas, Robert E, Jr., 161b5.5 middle-income countries, 61, 92 mid-size cities, 137­40 migration, 34b0.1, 100 Maddison, Angus, 194 and benefits of economic growth, 159­61, 161b5.5 Le Mâitre, Alexandre, 51 China, 13, 16map3, 17, 18 malaria, 116, 117b3.7 and division changes in countries, 100 Malaysia economic perspective on, 161b5.5 agriculture sector, 253 to exploit profit from proximity, 13, 15­18 geographic convergence in welfare, 92, 92f2.12 Germany, 235b8.2 improvement of living standards in, 27 impact of labor and social policies on, 160b5.4 Mali, 48­49, 220b7.11 international, 17, 150­52 manmade barriers, to distance, 75 link to physical distance, 75 manufacturing sector policies for managing, 168­69 China, 127b4.1 rates of, 247 clustering of as producers seek scale economies, 132­33 reasons for migrating, 166­67, 166t5.4 importance of, 119 restrictions on, 100 and internal scale economies, 129­30 rural-to-rural, 153 link to economic distance, 76­77 as way to reduce distance, 77­78 spatial concentration of, 209b7.5 military confrontations, costs of, 105 marginalization index, 79b2.2 Millennium Development Goals (MDGs), 27 maritime transport, 176­79, 188, 189, 192b6.7 Millennium Development Indicators, 92­93 markets and marketing minimum wages, 160b5.4 378 WORLD DEVELOPMENT REPORT 2009 Mints, Alexei, 90­91b2.5 opportunities mobile phones, 251, 268 in Russian Federation, 241­42 mobility sought by people, 231­34 with density, 185b6.5 for women in Germany, 235b8.2 and economic development, 84­85 Organisation for Economic Co-operation and Development factor mobility, 241, 258, 285­86 (OECD), 225 free movement of people, 260 density threshold used by, 55b1.3, 56 and opportunity seeking, 231­34 and federal transfers, 248 See also capital mobility; labor mobility GDP per capita in countries of, 63 Morocco, 52b1.1 and incentives to stimulate lagging areas, 256t8.2 mortality rates patterns of migration to countries of, 150­52 as indicators of well-being of urbanites, 72 preferential access for Africa to OECD markets, 282­83 in slums, 68b1.6 Osaka-Kobe, Japan, 109b7.5 Thailand, 92­93 outcome-oriented policy framework, 247 multilocation firms, 138 outcomes Mumbai, India, slums in, 204, 205b7.3 in education, 110, 110f3.10 identification of, 259 output density, 116 NAFTA. See North American Free Trade Agreement (NAFTA) outsourcing, 134­35, 181 national geographic scale, 3­5, 7 and concentration of economic production, 8, 9 and distance, 6b2, 7 Pacific Islands National Slum Development Program, India, 227 EPAs with EU, 265b9.3 national unity, 234­38 integration into world markets, 277, 278b9.5 native workers, compared to internal migrants, 162, 164f5.8 Pan-Arab Free Trade Area, 275 natural gas, 102b3.2 passport costs, 100 natural resources, 101­3, 103b3.3 pension benefits, 250 navigation, 174 per capita gross national income, 2 neighborhoods per capita incomes, 8 close to world markets, 275 growth in, 46 effects of, 2 link to income growth, 45 identification of for institutional reform, 278­81 and market access, 108b3.5 importance of, 2 See also gross domestic product (GDP) natural neighborhoods, 35, 278­81 performance monitoring mechanisms, 229b7.15 regional, 150­53 perishable goods, 177b6.2, 179, 273 New Partnership for Africa's Development, 281 Peru, postal service in, 251, 251b8.9 New York City physical geography, 115, 240 adapting to market forces, 143b4.7 and trade, 263­64 transport system and urbanization, 207, 208b7.4 and transport costs, 173 Nigeria, 48, 236b8.3 physical infrastructure, 185 and dimensions of development in, 6b2 places intergovernmental transfers, 249b8.8 evolving portfolio of, 51­54 Nomenclature of Territorial Units for Statistics (NUTS), 50, link to income, 46 78b2.1, 135, 237 plants, and size of cities, 21b4 nontariff barriers, 99 policies and policy makers North, Douglas, 203 for area development to integrate nations, 27­29 North American Free Trade Agreement (NAFTA), 46, 113, 264 and attributes of development, 2­3 NUTS1. See Nomenclature of Territorial Units for Statistics (NUTS) cohesion policies, 236­37 concerning infrastructure to reduce distance to density, 206­12 consideration of area as unit of government action, 201­2 OECD. See Organisation for Economic Co-operation and Devel- framework for economic growth, 40­43 opment (OECD) immigration policies, 157 offshoring tasks, 180 integration framework in action, 216­29, 245­58 Ohmae, Kenichi, 96 and lagging and leading areas oil revenues, 275 for countries seeking unity between lagging and leading open registry shipping, 176 areas, 234­38 Index 379 criteria used to define lagging areas, 79b2.2 producers, connecting to markets, 251­52 development policies for lagging areas, 230­31 product cycles, 179­80 framework for integrating lagging and leading areas, 238­45, production 246t8.1 and communication costs, 180­81 for managing migration, 168­69 concentration of, 8­10 and market forces, 128 diversification through regional cooperation, 263b9.2 national industrial policies, 253 facilities for, 257 and political benefits of economic integration, 258­59 fragmentation of, 120 preferential trade agreements, 282 importance of location in, 2 pricing policies, 257 intermediate inputs of, 19­20 protectionist trade policies, 175 land as immobile factor in, 142 to reduce economic disparities, 22­24 in leading areas, 9 for regional integration, 261b9.1, 265­73, 273­83 and location of firms and facilities, 41b0.3, 179­80 spatially blind policies, 200b7.1 spatial disparities in, 6 spatially targeted interventions for slums, 212­15 spatially efficient distribution of, 46 transport polices in developing world, 184­92 specialization of, 34b0.1 and urbanization production networks, 180, 194­95, 266 framework for, 199­200 productivity, 84 policies for stages of, 24­27 gains from scale economies, 135 policy instrument for dimensions of, 215­16 increase in, 183 spatially blind institutions to facilitate, 202­3 rural-urban disparities in, 62­66 strategies for inclusive urbanization, 229 profit, migration to profit from proximity, 13, 15­18 political centers, 51, 52, 53b1.2 progressive tax systems, 239­40 poor people property registry systems, 214 in Brazil, 232­33 property rights, 203 China, 232 prosperity, and location, 1­8 distribution of, 232­33 provincial growth product, inequalities in, 89, 89f2.10 India, 233­34 proximity, migrating to for profit, 13, 15­18 interventions to integrate urban poor, 118t7.2 proximity trap, 284 in leading areas, 81, 82map2.5 psychic costs, 75 population public goods and services city comparisons, 49­50 access to, 10­12 density in United States, 44 delivery of link to border permeability, 98­99 density of link to economic mass, 54 as indicators of well-being of urbanites, 72 foreign born population, 150 provision of, 275­76, 281 geographic centers of in United States, 44mapG1.1 as reason for migration, 166t5.4, 167­68 increase in cities, 12, 13b31 taxonomy of properties of, 269­70 in lagging areas, 24, 27­29, 29 transfer mechanisms for, 247­50 prediction of, 198­99 public health, in slums, 68b1.6 spatial distribution of, 39b0.3 public housing, Sweden, 215b7.8 Sub-Saharan Africa, 284­86 public transportation in urban areas, 3, 5 New York City, 207, 208b7.4 port efficiency, 188, 189, 192b6.7 as promoter of higher density, 208, 210­12 postal service, 251, 251b8.9 poverty and landlocked countries, 101, 102b3.1, 102b3.2, rail freight costs, 175­76 103map3.3 rail systems, 174, 208b7.4 link to malaria, 116, 117b3.7 Indian Railways, 175b6.1 poverty rates in lagging areas, 81, 82map2.5 rank-size rule of cities, 51, 52b1.1 poverty maps, 27 rapid transit systems, 223, 225map7.5 poverty rates redistributive transfers, 247­50 Brazil, 232­33 reforms, institutional, 266­67, 274­76, 278­81 India, 233­34 Regional African Communications Organization, 268 pricing, policies for, 257 regional infrastructure, 266, 267­69 primary cities, 49, 51­52 investment in, 281 private sector, and slum upgrading and prevention, 229b7.15 trade enhancing, 268­69 380 WORLD DEVELOPMENT REPORT 2009 regional integration in China, 126, 127b4.1 diversifying production through, 263b9.2 cities facilitation of, 134 findings concerning, 113­14 clustering of services vs. clustering of manufacturing in, 133­34 framework in action, 273­83 and enhancement of productivity-enhancing regional geography of, 270­73 infrastructure, 268 and growth spillovers, 102b3.1 guide to, 128t4.1, 129­32 and income convergence, 112 list, 128t4.1 to increase access to world markets, 29­32 overview, 126 institutional reform to improve, 275­76 theoretical advances recognized, 136­37 instrument per dimension of, 271, 273t9.4 and transport costs, 171­72, 181­84 investments in cross-country infrastructure for, 276­77 transport sector, 19, 186­87 policy framework for, 265­73 in United States, 46 Sub-Saharan Africa, 286 sea-locked countries, 103­4 trade of agricultural products as entry point to, 280 search costs, 180 See also integration secondary cities, 49­50, 52 regional trade agreements (RTAs), 102b3.1, 266, 280 second-nature geography, 116, 118 regions security, in slums, 67 cooperation in, 19 Seoul, Korea, 220­23, 224­25 description of, 35­36, 36b0.2 services development of markets in, 261 clustering of in scale economies, 133­34 as geographic scale, 3b1 and developing economies, 132­33 income convergence within faster growing, 112­15 fragmentation of, 20b3 patterns of migration in regional neighborhoods, 150­53 and internal scale economies, 129­30 policies to integrate nations, 27­29 rural-urban disparities in, 65 regional cities, 52­54 SEZs. See special economic zones, (SEZs) regulations shipping for housing finance, 206 costs for, 176­79 land market regulations, 250 rates of, 174 for land use and building, 204­6, 226­27 routes, 172 of transport for benefits of scale economies, 186­87 sigma-convergence, 46 remittance flows, 161­62, 165­66, 168 Singapore, slums in, 214, 215b7.8 rent-seeking, 186 size research and development, 101 and economic prosperity, 101­3, 103b3.3 resource allocations, 236 influence of, 118 response rule, 40­42 as volume of demand, 183 Ricardo, David, 106 skills roads and patterns of internal and international migration, 155­57, road networks, 251 158f5.6 road space, 144 skilled emigrants, 168 road transport costs, 175 slums RTAs. See regional trade agreements (RTAs) as challenge for economic integration, 24 rural areas in developing countries, 4map1, 5 link to towns, 53­54 divergence and convergence within cities, 66­70 rural households and macroeconomic policies, 200b7.1 India, 204, 205b7.3, 227 rural-to-rural migration, 153 integration into cities, 227­29 rural-urban disparities in, 62­66 Singapore, 214, 215b7.8 rural-urban transformations, 216­20 slum improvement programs, 26, 229b7.15 in Cambodia and Vietnam, 218, 219b7.9 spatially targeted interventions for, 212­15 China, 226, 227b7.14 Sweden, 213 and land market institutions, 220b7.11 in world class cities, 69b1.7 Russian Federation, population and economic opportunities in, smaller cities, 52­53, 137­38 241­42 Smith, Adam, 14­15, 93, 126, 135, 194 social inequalities, 39b0.3, 46 social policies, impact on migration, 150b5.4 sanitation, in slums, 67 social services, 26, 216­28 scale economies China, 227b7.14 agriculture disperses and manufacturing clusters in, 132­33 and policies to manage migration, 168­69 amplified by density and attenuated by distance, 134­35 rising investments in, 87b2.4 Index 381 South Africa, housing shortage, 214 subnational governments, and decentralization, 231 South America, 115b3.6 Sub-Saharan Africa South Asia brain drain in, 168b5.9 average annual growth rate of GDP, 110f3.11 density, distance, and division in, 284­86 disparities in lagging areas of, 80b2.3 GDP for, 265b9.3 tariffs in, 99, 99f3.3 and labor mobility, 151­52, 281 Southeast Asia, per capita GDP in, 89, 89f2.9 urbanization in, 59b1.4 Southern African Development Community, 270 subway systems, 208b7.4, 210 Southern Cone of South America, 115b3.8 Superintendency for the Development of the Northeast Southern Europe, 115b3.8 (SUDENE), Brazil, 38 South Pacific Regional Trade and Economic Cooperation supply and demand Agreement (SPARTECA), 278b9.5 demand management relationship to public transportation, Soviet Union, spatial inefficiency and fall of, 90­91b2.5 208, 210­12 SPARTECA. See South Pacific Regional Trade and Economic regional global integration for, 262­65 Cooperation Agreement (SPARTECA) Sweden, slums in, 213 spatial dimensions, elements of, 36­37 symbiosis, and large cities, 52, 53 spatial disparities, 80b2.3 in developed countries, 85­87 France, 238, 238f8.1 tanneries, 6b2 in incomes in fast-growing countries, 89­92 tariffs, 99, 192b6.7 in living standards in developing countries, 87­89 taxonomy, to assist countries integration policies, 241 and U.S. tax system, 238­39 tax systems, 238­40 spatial integration, 22­24 technical information, 101 spatially balanced development, 73 technologies, for transport sector, 173, 177b6.2 spatially blind institutions, 22­24, 29, 38, 202­3 telecommunications, 138, 267, 275 and economic integration policies, 231b8.1 costs of, 140b4.5 to overcome density and distance, 242­44 impact of costs on, 94­95 See also institutions and labor mobility, 157 spatially connective infrastructure, 23­24, 29, 38 territorial development, 27­29, 286 China, 221­22 territorial disputes, economic costs of, 105 Egypt, 243­44 territorial policies, of European Union, 39b0.3 Iran, 250­51 territorial units, classification of, 78b2.1 Japan, 208, 209b7.5 textile producers, 131 Korea, 221 TFP. See total factor productivity (TFP) to reduce distance to density, 206­12 Thailand, 153b5.2 See also infrastructure time spatially targeted incentives, 37-39, 253­57 costs of, 189 spatially targeted interventions, 23­24, 37­39, 42, 238, 242 as factor in international trade, 179­80 spatial scales, 35­36 Tinbergen principle, 229 spatial transformations, 33, 198­99 Tobler, Waldo, 76 special economic zones, (SEZs), 224b7.13, 254­55b8.10 Todaro class of migration models, 159­60 specialization Tokyo-Yokohama, Japan, 109b7.5 and diversity in cities, 137­40 total factor productivity (TFP), 77, 116, 135 of production, 34b0.1 tourism, 177b6.2 and transport costs, 13, 18­20 town planning, Hong Kong, China, 213 spillovers towns, 53­54 of growth from across borders, 102b3.1 trade, 182 of knowledge, 134, 158 in agricultural products for regional integration, 280 from proximity to density, 76­77 and institutional cooperation, 266­67 regional, 116 in intermediate goods, 20, 22 Stabilization and Association Agreements, 274 international, 179­80, 182 standard of living. See living standards interregional, 193­95, 194­95 statistical areas, and economic activity link to development stages, link to border permeability, 98­99 83­84, 85t2.1 logistics and time factors in, 75, 179­80 subnational areas national efforts and regional coordination to facilitate, 187­90 definition of, 78b2.1 between neighbors, 182­83 rural-urban convergence in, 66 preferential trade agreements, 282­83 spatial disparities of living standards in, 87­89 protectionist policies for, 175 382 WORLD DEVELOPMENT REPORT 2009 trade (continued) Mexico, 244b8.6 regional and global integration of, 262­65 negative externalities of, 190­92 restraints, 45 policies in developing world, 184­92 as share of global GDP, 93f2.13, 94 scale economies in, 19 shipping routes, 172 technology inventions for, 177b6.2 trade-enhancing regional infrastructure, 268­69 urban transport systems, 183, 185b6.5 trade friction, 179­81 weight-value ratio in, 179 trade volume, 171­72 travel speed, link to market access, 77f2.1 See also intraindustry trade; markets and marketing; transport trust, as component of regional cooperation, 270 costs; world markets Turkey, 21, 22, 225­26 trade agreements, 46, 271, 273t9.4 training, and mobility-enhancing regional infrastructure, 268 tramp shipping, 176­79 Uganda, 236b8.3, 247, 248b8.7 transfers, 247­50 unemployment, 184b6.4, 215b7.8 transfrontier regionalism, 279 unemployment insurance, 160b5.4 transmigration programs, 238 UN-HABITAT, 229b7.15 transport costs, 94­95, 134, 170 United Kingdom across United States, 45 criteria used to define lagging areas in, 79b2.2 air transport costs, 176 divergence then convergence of income and welfare in, 85­86 China, 251 economic development in, 255 concentration within countries, 183­84, 185b6.5 infrastructure in, 206­7 domestic transport, 173­74 United Nations, region classification, 36b0.2 and economic concentration, 252 United States and economic geography, 181­82 borders of, 46 and growth strategies, 21b4 convergence of living standards, 73 historical perspective, 170­71, 173­81 divergence then convergence of income and welfare in, 85­86 impact of distance on, 285­86 economic development programs in, 256 and increase in transport intensity, 175­79 economic geography of, 44­45 and infrastructure, 186­87, 212 evolvement of institutions and infrastructure in, 207 international transport, 174 federal income tax system, 238­39 and intraindustry trade, 20b3 historical perspective of wages in United States South, 165b5.7 Italy, 184b6.4 investing in human capital, 246­47 maritime transport costs, 176­79 rank-size rule of, 52b1.1 rail freight costs, 175­76 slums in, 213 reduction in, 45 urban areas, 55b1.3 road transport costs, 175 and density, 49 and scale economies, 171­72, 181­84 and external economies, 141 in sea-locked countries, 103 rural-urban disparities in, 62­66 specialization and trading as costs fall, 13, 18­20 transport sector in, 222 Sub-Saharan Africa, 286 See also slums and trade between neighbors, 182­83 urban hierarchy, 51, 53b1.2 and trade-enhancing regional infrastructure, 268­69 urbanization, 3, 9, 128t4.1, 129 and trade friction, 179­81 and access to public services, 10­12 trade volume impact on, 171­72 Africa, 59b1.4, 285 transport infrastructure, 46, 75, 175 China, 26b5, 154­55b5.3 in advanced urbanization, 223­29 and climate change, 211b7.6 to connect lagging and leading areas, 250­52 and divergence and convergence of slums within cities, 66­70 transport sector economic activities in, 141­42 Africa, 186, 187t6.1 framework in action, 216­29 and air pollution, 190­91 and housing finance regulations, 206 in developing countries, 267 indicators of well-being in, 72 domestic transport, 173­74 from industrial diversity that fosters innovation, 131­32 and economic concentration, 252 infrastructure policies to reduce distance to density, 206­12 and expansion of cities, 144 Korea, 216­27 importance of sector, 192­93 land reform to facilitate in Denmark, 203b7.2 and institutional cooperation, 266­67 and land use and building regulations, 204­6 international transport, 174­75 measures of, 54­56, 57f1.3 investments in, 222 pace and pattern of, 57­59, 70­71 Index 383 policy considerations for, 24­27, 199­200 welfare policy instruments for dimensions of, 215­16 divergence then convergence of in developed countries, 85­87 principles for managing, 200­202 and factor mobility, 258 promotion of, 23­24 impact of education on, 247 rate of, 8 improvements in, 110 and reduction of congestion, 207­8, 209b7.5 indicators of, 73 rural-urban disparities in, 62­66 Malaysia, 92, 92f2.12 spatially targeted interventions for, 212­15 rural-urban disparities in, 62­66 strategy for inclusive urbanization, 229 spatial disparities in developing countries, 87­89 Sub-Saharan Africa, 285, 286 West Africa, 187 Turkey, 225­26 cotton-led industrial development in, 280, 280map9.6 United Kingdom, 206­7 and economic integration, 21­22 volume of in developing countries, 71­72 West African Economic and Monetary Union (WAEMU), 263b9.2, See also slums 264­65 urbanization economies, 15, 215­16 Western Asia urban residents, definition of, 8 average annual GDP growth, 113­14 income convergence in, 113­14 tariffs in, 99, 99f3.3 Vajpayee, Behari, 120 World Bank, region classification, 36b0.2 value added, 49, 181 World Development Report 2009 value-to-weight ratios, 179, 182 elements not covered in, 34b0.1 variable trade costs, 180 scope of, 34 vehicles structure of, 39­43 and congestion, 210 terms used in, 34­39 and enlargement of cities, 45­46 world markets growth in numbers of, 144 countries distant from world markets with small neighbors, vertically integrated industries, 256 277­83 Vietnam and geography of regional integration, 270­73 poverty in lagging and leading areas, 81map2.4 importance of, 93­94 rural-urban transformation of, 218, 219b7.9 and institutional reforms, 274­75 visa requirements, 100, 101f3.5 integration options for countries close to world markets, 273­75 for distance obstacles, 275­77 WAEMU. See West African Economic and Monetary Union integrating Pacific Islands into, 277, 278b9.5 (WAEMU) regional integration to increase access to, 29­32 wages strategies to integrate poor countries into, 260-62 Asia vs. Europe, 194 and telecommunications, 275 converging of, 10­12 World Trade Organization (WTO), 187­88, 265b9.3, 266, 278b9.4 historical perspective in United States South, 165b5.7 World Urbanization Prospects, 55b1.3, 56, 58, 285 and labor mobility, 164­65 WTO. See World Trade Organization (WTO) rural-urban disparities in, 62­66 water management, 275 wealth Zheng He, 194 concentration of, 9f1c, 10, 115­20 Zipf's law, 51, 139 rural-urban disparities in, 64 zoning policies, 207, 228 The Wealth of Nations, 14­15, 93, 126, 135 E C O - A U D I T Environmental Benefits Statement The World Bank is committed to Saved: preserving endangered forests and · 92 trees natural resources. The Office of the · 64 million BTUs of Publisher has chosen to print the total energy World Development Report 2009 · 211,008 pounds of on recycled paper with 10 percent postconsumer fiber in accordance CO2 equivalent greenhouse gases with the recommended standards for paper usage set by the Green · 700,880 gallons of Press Initiative, a nonprofit program waste water supporting publishers in using fiber · 83,411 pounds of that is not sourced from endangered solid waste forests. For more information, visit www.greenpressinitiative.org. reshaping economic geography Places do well when they promote transformations along the dimensions of economic geography: higher densities as cities grow; shorter distances as workers and businesses migrate closer to density; and fewer divisions as nations lower their economic borders and enter world markets to take advantage of scale and trade in specialized products. World Development Report 2009 concludes that the transformations along these three dimensions--density, distance, and division--are essential for development and should be encouraged. The conclusion is controversial. Slum-dwellers now number a billion, but the rush to cities continues. A billion people live in lagging areas of developing nations, remote from globalization's many benefits. And poverty and high mortality persist among the world's "bottom billion," trapped without access to global markets, even as others grow more prosperous and live ever longer lives. Concern for these three intersecting billions often comes with the prescription that growth must be spatially balanced. This report has a different message: economic growth will be unbalanced. To try to spread it out is to discourage it--to fight prosperity, not poverty. But development can still be inclusive, even for people who start their lives distant from dense economic activity. For growth to be rapid and shared, governments must promote economic integration, the pivotal concept, as this report argues, in the policy debates on urbanization, territorial development, and regional integration. Instead, all three debates overemphasize place-based interventions. Reshaping Economic Geography reframes these debates to include all the instruments of integration--spatially blind institutions, spatially connective infrastructure, and spatially targeted interventions. By calibrating the blend of these instruments, today's developers can reshape their economic geography. If they do this well, their growth will still be unbalanced, but their development will be inclusive. ISBN 978-0-8213-7607-2 SKU 17607