58997 MENA DEVELOPMENT REPORT Poor Places, Thriving People How the Middle East and North Africa Can Rise Above Spatial Disparities Poor Places, Thriving People M E N A D E V E LO P M E N T R E P O R T Poor Places, Thriving People How the Middle East and North Africa Can Rise Above Spatial Disparities © 2011 The International Bank for Reconstruction and Development / The World Bank 1818 H Street NW Washington DC 20433 Telephone: 202-473-1000 Internet: www.worldbank.org All rights reserved 1 2 3 4 13 12 11 10 This volume is a product of the staff of the International Bank for Reconstruction and Develop- ment / 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 gov- ernments 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 com- plete information to the Copyright Clearance Center Inc., 222 Rosewood Drive, Danvers, MA 01923, USA; telephone: 978-750-8400; fax: 978-750-4470; Internet: www.copyright.com. All other queries on rights and licenses, including subsidiary rights, should be addressed to the Office of the Publisher, The World Bank, 1818 H Street NW, Washington, DC 20433, USA; fax: 202-522-2422; e-mail: pubrights@worldbank.org. ISBN: 978-0-8213-8321-6 eISBN: 978-0-8213-8423-7 DOI: 10.1596/978-0-8213-8321-6 Library of Congress Cataloging-in-Publication Data Poor places, thriving people : how the Middle East and North Africa can rise above spatial disparities. p. cm. Includes bibliographical references and index. ISBN 978-0-8213-8321-6 -- ISBN 978-0-8213-8423-7 (electronic) 1. Industrial location--Middle East. 2. Industrial location--Africa, North. 3. Middle East-- Economic conditions--Regional disparities. 4. Africa, North--Economic conditions--Regional disparities. 5. Space in economics. I. World Bank. HC415.15.Z9D576 2011 339.4'60956--dc22 2010038030 Cover photo: Loďc Whitmore/World Bank Contents Preface xvii Acknowledgments xix Abbreviations xxi 1 Overview 1 The Political Demand: Spatial Equity with Productivity 1 Spatial Disparities in Living Standards: Uncovering the Truth 3 Policy Package 1. Level the Playing Field and the Opportunity for Human Development in Poor Places 11 Policy Package 2. Connect Poor Places to the Poles of Development 23 Policy Package 3. Underpin Private Sector Interest in Nonleading Areas 26 Proximity for All: Public Institutions for Spatial Policies 29 Endnote 32 References 32 2 The Political Demand: Spatial Equity without Compromising Productivity 35 MENA's Spatial Disparities in Living Standards: Six Facts 36 Politics Makes Geography Important 38 MENA Needs Economic Agglomeration 39 Spatial Disparities Create Political Stresses 44 Women and Men Experience Spatial Inequalities Differently 46 v vi Contents Spatial Equity with Spatial Efficiency 48 Matching Policy Packages to Lagging Areas: A Three-Way Framework 49 Endnotes 58 References 59 3 Diagnose to Prescribe: Uncovering the Truth about Spatial Disparities 63 What Kind of Spatial Inequality Does the Country Have? 64 Is the Lagging Area a Fringe, Belt, or Pocket? 112 Endnotes 124 References 124 4 Policy Package 1. Level the Playing Field and the Opportunity for Human Development in Lagging Areas 131 Apply in All Circumstances: The Only Package for "Fringe" Areas 131 How Governance Affects Urbanization--Some Numbers and an Explanation 133 How Differences in the Subnational Investment Climate Can Handicap Lagging Areas 141 How to Manage the Spatial Dimension of Public Expenditure 146 How to Correct Spatial Disparities in Public Education Systems 152 Reducing Gender Disparities in Education in Lagging Areas 163 Health: Correcting Spatial Disparities in Access to Care 169 The Better Social Safety Nets Are Targeted, the Better Spatial Disparities Are Mitigated 176 Smoothing the Path for City Growth 189 Endnotes 203 References 205 5 Policy Package 2. Connecting Poor Places to the Poles of Development 213 The Power of Spillovers 213 Road Transport: An Extensive Long-Distance Network but an Underdeveloped Short-Distance Connectivity 217 Contents vii Regional Trade Facilitation Matters Crucially for Lagging Areas 228 Priorities for Improving Spillover Connectivity 230 Electronic Proximity: New Tool for the Integration of Lagging Areas 232 Endnotes 241 References 241 6 Policy Package 3. Underpin Private Sector Interest in Nonleading Areas 247 Difficulty and Expense of Using Public Money to Lure Development to Lagging Areas 250 MENA's Focus on Using Public Resources to Override Economies of Agglomeration 251 Financial Incentives to Steer the Spatial Pattern of Development Have Not Been Successful 254 New Paradigm of Regional Economic Development Policy 258 Realizing Agricultural Potential through Clusters 262 Conclusion 268 Endnotes 269 References 269 7 Proximity for All: Public Institutions for Spatial Policies 275 Spatial Disparities: A Big Task for the State 275 How States Organize and Mobilize to Address the Spatial Dimension of Policy 278 Conclusion: Keeping Spatial Development Simple 290 Endnotes 293 References 293 Appendix: MENA Region 297 Index 341 List of Boxes Box 1.1 Convergence Trends for the Islamic Republic of Iran, Tunisia, and the Arab Republic of Egypt 9 Box 2.1 Ibn Khaldun and the Political Economy of Spatial Disparities 45 viii Contents Box 2.2 A Migrant's Wife Has Difficulties with the Local Authorities 47 Box 2.3 Mind Your Language! 50 Box 3.1 The Nile Delta: Urban, Rural . . . or Rurban? 64 Box 3.2 Decomposing MENA's Inequality into Its Spatial and Nonspatial Components: A Worthwhile Exercise, Despite the MAUP 79 Box 3.3 Caffeine-Based Networking among Cairo Construction Craftsmen 110 Box 3.4 The Arab Republic of Egypt's Human Development Reports: A Model for MENA? 120 Box 4.1 Lagging Areas: Is Culture to Blame? 132 Box 4.2 An Entrepreneur's Experience of Spatial Disparities in the Business Environment 142 Box 4.3 Making the Subnational Investment Climate Work for Lagging Areas: An Idea for Policy Makers 145 Box 4.4 Algeria Introduces Spatial Salary Incentives, with Mixed Results 160 Box 4.5 Leveling Spatial Disparities in Education: Nine Ideas for Policy Makers 168 Box 4.6 Leveling Spatial Disparities in Health: Four Ideas for Policy Makers 177 Box 4.7 Harnessing Social Safety Nets to Lagging-Areas Policy: Three Ideas for Policy Makers 190 Box 4.8 Stories of Low-Income Housing Provision: Tunisia and the Islamic Republic of Iran 194 Box 4.9 Facilitating Smoother Urbanization: Six Ideas for Policy Makers 202 Box 5.1 Seeking the Balance between Underregulation and Overregulation of Freight and Passenger Transportation 227 Box 5.2 Enhancing Spillover Connectivity in MENA: Four Ideas for Policy Makers 233 Box 5.3 Bringing ICT Connectivity to Lagging Areas: Two Ideas for Policy Makers 240 Box 6.1 Tax Breaks for Lagging Areas in MENA 253 Box 7.1 Examples of Appointed Provincial and Elected Municipal Local Administration in MENA 279 Contents ix Box 7.2 How Tunisia Integrates the Spatial Dimension into National Development Planning 281 Box 7.3 Morocco's Experimentation with Deconcentration Illustrates the Challenges 286 Box 7.4 Mobilizing Institutions for Spatial Policy Making: Seven Ideas for Policy Makers 291 List of Figures Figure 1.1 MENA's Agricultural Labor Force Is Growing Fast Relative to Its Level of Development 5 Figure 1.2 MENA's Working-Age Population Has Been Booming Compared with Other Countries and Economies at the Same Level of Development 5 Figure 1.3 A Composite Index of Objective Measures of the Subnational Investment Climate in Algeria, the Arab Republic of Egypt, Jordan, Morocco, and the West Bank and Gaza 15 Figure 1.4 Rural Access Index: MENA versus Comparator Regions, 2004 24 Figure 1.5 Some MENA Economies Are Well Below the Trend in Terms of Cellular Connections 26 Figure 1.6 A Framework for Addressing Spatial Disparities 29 Figure 2.1 Spatial Distribution of Disparities 38 Figure 2.2 Hypothetical Range of Outcomes between National Consumption Growth and Consumption Growth in Tunisia's Poorer Areas, 1985­2000 43 Figure 2.3 In Morocco, Provinces with Large Industrial Sectors Provide the Highest Returns to Industrial Investment 43 Figure 2.4 Consequences of Splitting MD 1 Billion Industrial Investments between Leading Three and Lagging Eleven Provinces in Morocco, 1996­2005 44 Figure 2.5 Agglomeration and Spatial Disparities in Well-Being: Two Separate Outcomes 51 x Contents Figure 2.6 Matching Policy Packages to Lagging Areas 57 Figure 3.1 MENA Is Agglomerated for Its Level of Development 66 Figure 3.2 Poverty Is Rural and Growth Is Nonagricultural, in MENA Economies 70 Figure 3.3 MENA's Agricultural Labor Force Is Growing Fast for MENA's Level of Development 71 Figure 3.4 MENA's Working-Age Population Has Been Booming, Compared with Other Countries at the Same Level of Development 73 Figure 3.5 High Poverty Rates Did Not Mean High Poverty Numbers in Iraq, 2007 77 Figure 3.6 Demography Matters for Intergovernorate Disparities in the Arab Republic of Egypt 85 Figure 3.7 Remove Demographic and Educational Differences, and the Arab Republic of Egypt's Urban-Rural Divide Shrinks Away, 2005 85 Figure 3.8 Djibouti's Interdistrict Disparities: Tadjourah and Obock Lag in a Different Way from Ali-Sabieh and Dikhil 87 Figure 3.9 Djibouti's Rural-Urban Gap, 2002 87 Figure 3.10 Intergovernorate Disparities Are Small and Largely Demographic in Jordan 88 Figure 3.11 Demography and Education Do Not Cause Interregional Disparities in Lebanon 89 Figure 3.12 Interprovincial Disparities Are Quite Modest and Not the Fault of Demography in Morocco 90 Figure 3.13 The Rural-Urban Gap Is Large and Cannot Be Explained Away by Demography or Education in Morocco, 1998 91 Figure 3.14 Lagging Governorates Are a Demographic Phenomenon; Education Is Not the Issue in the Syrian Arab Republic 92 Figure 3.15 Demography and Education Explain about a Third of Intergovernorate and Rural-Urban Inequality in Republic of Yemen 93 Figure 3.16 The Islamic Republic of Iran's Demographic Transition Began in Urban Areas 95 Figure 3.17 Inter-Ostan Disparities Shrink over Time in the Islamic Republic of Iran 100 Contents xi Figure 3.18 Regions Split into Rich and Poor Clubs, While Converging Overall in Tunisia 102 Figure 3.19 Proximity of Eastern Governorates to the Metropolis Does Not Explain Performance in Tunisia, 1980­2000 102 Figure 3.20 Two-Speed Development in Rural Upper Egypt, 1996­2006 104 Figure 3.21 The Mashreq's Lagging Area Populations Are Denser and Closer to Density than the Maghreb's 119 Figure 4.1 People in Smaller Settlements State More Confidence in their Government in the Arab Republic of Egypt 140 Figure 4.2 A Composite Index of Objective Measures of the Subnational Investment Climate in Algeria, the Arab Republic of Egypt, Jordan, Morocco, and the West Bank and Gaza 143 Figure 4.3 Public Investment and per capita Household Expenditure by Governorate in Tunisia 148 Figure 4.4 Transfers from the Center to Governorates Are Not Reducing Spatial Disparities in the Arab Republic of Egypt 150 Figure 4.5 Much Variation, but a Tendency for the Intergovernmental Transfers to Favor Lagging Ostans in the Islamic Republic of Iran 151 Figure 4.6 Variation in per Capita Expenditure of Local Government by Governorate in Jordan 152 Figure 4.7 Percentage of Adults Older Than Age 25 Who Completed Primary Education 154 Figure 4.8 Percentage of Adults Older Than Age 25 Who Completed Secondary Education 154 Figure 4.9 Percentage of Adults Older Than Age 25 Who Completed Tertiary Education 155 Figure 4.10 Rural-Urban Disparities in Mathematics Learning, 2007 TIMSS Mathematics Scores 157 Figure 4.11 Women's Education Reduces Fertility Rates in the Arab Republic of Egypt 164 Figure 4.12 Unusually High Urban-Rural Gaps in Access to Potable Water in Morocco and Iraq 172 xii Contents Figure 4.13 Rural-Urban Disparities in Sanitation Coverage 173 Figure 4.14 Commodity Subsidies: A Big Slice of MENA's GDP 181 Figure 4.15 Spatial Incidence of Cash Transfer Program, Iraq 184 Figure 4.16 National Population Growth Underlies Urban and Slum Population Growth in the Arab Republic of Egypt and Tunisia 192 Figure 4.17 Some of Cairo's Most Valuable Land Is Reserved for Agriculture 197 Figure 5.1 Paved Roads Density: MENA Compares Well with Other Middle-Income Countries 218 Figure 5.2 MENA Region Concentrates on Paved Roads More Than in Comparator Economies 219 Figure 5.3 Rural Access: MENA versus Comparator Regions 222 Figure 5.4 MENA Is a Region of Dangerous Roads 225 Figure 5.5 Average Rates of Road Crash Fatalities per Capita and Registered Vehicles in MENA Region and in Comparator Countries 226 Figure 5.6 The Spatial Coverage of ICT in Developing Regions' Rural Areas 236 Figure 5.7 Telecommunications: Who Is Ahead and Who Is Behind the International Trend? 239 Figure 7.1 A Stylized Portrait of the Multilevel State in the MENA Region 278 Figure 7.2 MENA's Local Governments Spend Far Less Than Those in Comparable Economies 288 List of Maps Map 1.1 A Policy-Oriented Spatial Typology of the Middle East and North Africa 12 Map 2.1 Spatially Concentrated Economic Activity in Libya (left) and Iraq (right) 41 Map 3.1 MENA's Pockets of Nonaridity 67 Map 3.2 Water-Based Settlement Patterns in Iraq: Rainfed and Riverine 68 Map 3.3 Interior and Saharan Algeria Contain Regional Pools of Demographic Vitality and Potential Population Growth 96 Contents xiii Map 3.4 Remittance Income Clusters in the Republic of Yemen 111 Map 3.5 A Policy-Oriented Spatial Typology of the Middle East and North Africa Region 114 Map 4.1 Petroleum Product Subsidies Exacerbating Spatial Disparities in Iraq 182 Map 4.2 Poverty Incidence and Poverty Gaps in the Syrian Arab Republic 187 Map 4.3 Poverty Gaps in the Republic of Yemen 189 Map 5.1 Trunk Road Connectivity Is Good Where the People Are: Travel Time in Hours to a City of More Than 500,000 Inhabitants, with Closed Borders 215 Map 5.2 Smooth Border Crossings Would Give Lagging Border Regions a Better Chance of Benefiting from Spillover Connectivity: Travel Time in Hours to a City of More Than 500,000 Inhabitants, with Zero-Hour Border Crossings 229 Map 6.1 The Islamic Republic of Iran's Placement of Economic Zones in Low-Density Areas 255 Map 6.2 Tunisia's Priority Development Zones 257 List of Tables Table 1.1 Matching Policy Packages to Lagging Areas 3 Table 1.2 Lower Enrollment Rates for Rural Children, Especially the Poor 16 Table 1.3 Old and New Paradigms of Regional Economic Development Policy 27 Table 2.1 Spatial Disparities in Economic and Human Development in MENA 37 Table 2.2 Investment in the Poorest Areas of Tunisia Gives the Poorest Returns 42 Table 2.3 Matching Policy Packages to Lagging Areas 56 Table 3.1 How Agglomerated Are the Economies of the Middle East and North Africa? 65 Table 3.2 Difference between Agricultural Labor Productivity in MENA and Spain: How Much Is Due to the Land-Labor Ratio? 71 Table 3.3 Mean versus Median per Capita Consumption 75 Table 3.4 Human Development Indicators in Slums 77 xiv Contents Table 3.5 Rural and Slum Populations in MENA 78 Table 3.6 Where Does the Urban-Rural Divide Count? 80 Table 3.7 Where Does the Interprovincial Divide Count? 81 Table 3.8 Explaining the Difference between per Capita Consumption in Upper and Lower Egypt 83 Table 3.9 Urban-Rural Differences in the Cost of Living Partially Cancel Out Spatial Disparities in Spending Power 98 Table 3.10 The Arab Republic of Egypt, 2005­08: Urban Convergence with Rural Divergence 103 Table 3.11 Urbanization without Deruralization or Metropolitanization 108 Table 3.12 Little Lateral Redistribution of Population toward Leading Provinces, Save in Tunisia 109 Table 3.13 Lebanon's Population Distribution Becomes Less Agglomerated 112 Table 3.14 Indicators of Population Concentration 116 Table 3.15 Proximity Comparators 117 Table 3.16 Geographic Character of Lagging-Area Populations in MENA 118 Table 3.17 Contrasting Broad Regions in the Arab Republic of Egypt 121 Table 3.18 Policy Implications of Multi-Level Spatial Analysis in Rural Upper Egypt 123 Table 4.1 Politics Affects the Size of a Country's Largest City 133 Table 4.2 Political Rights Reduce Agglomeration 134 Table 4.3 Nonmetropolitan People State More Confidence in Their Governments 140 Table 4.4 MENA's Commercial Banks Have Extensive Networks with Few Accounts 144 Table 4.5 Transfer Characteristics and Allocation Rules in Six MENA Economies 149 Table 4.6 Lower Enrollment Rates for Rural Children, Especially the Poor 153 Table 4.7 Indicators of Rural-Urban Disparities in Health 178 Table 4.8 Social Resource Transfers in MENA 180 Contents xv Table 4.9 Most Social Funds Target Lagging Areas Well 183 Table 4.10 Conditional Cash Transfers that Are Proven to Reduce Poverty 183 Table 4.11 Urban and Slum Growth in MENA, 1990­2001 191 Table 4.12 Housing Supply Indicators for Eight MENA Economies 193 Table 4.13 The Time and Cost Required to Purchase Property: How the MENA13 Match Up against Other Regions 195 Table 4.14 How MENA Is Upgrading the Institutional Environment for Urban Growth 203 Table 5.1 Most of MENA's Population Live Close to Agglomerations 215 Table 5.2 Percentage of Population Living More Than Three Hours' Road Travel from a City of at Least Half a Million Inhabitants 219 Table 5.3 Road Connectivity Problems in MENA, as Highlighted in EuroMed and World Bank Sector Reviews 220 Table 5.4 MENA's 2006­08 Logistics Performance Index (LPI) Benchmarked 229 Table 5.5 How ICT Can Benefit Lagging Areas 234 Table 5.6 Fixed-Line and Internet Market Status (According to Operational Licenses) in MENA 238 Table 6.1 Spatially Targeted Incentives for Business Are Prevalent in MENA 252 Table 6.2 Old and New Paradigms of Regional Economic Development Policy 258 Table 6.3 Agriculture Is Straining Water Resources in MENA 264 Table 6.4 Price Decreases Were the Rule for MENA's Fruit and Vegetable Farmers, 1995­2007 265 Table 6.5 Forms of Subsidization of Agriculture, 2005 267 Preface It is now well known that the clustering of production in major cities is a driving force of economic growth. However, the governments of the Middle East and North Africa (MENA) are determined to make sure that their citizens benefit from their national development no matter where they live. The principle of economic justice makes wide gaps in living standards between leading and lagging areas politically intolerable. So the key question is this: how can MENA reap the benefits of economic agglomeration while mitigating geographic disparities in well-being? It is hard for policy makers to find straightforward advice on lagging area development. This is not surprising, given the controversy that exists among the experts and the diversity of locations. Following the 2009 World Development Report titled Reshaping Economic Geography, the moment was right to draw together the World Bank's thinking on lagging areas in MENA. The main messages of Poor Places, Thriving People: How the Middle East and North Africa Can Rise Above Spatial Disparities can be summarized in four words: people, connections, clusters, and institutions. · The surest way to help a lagging area to catch up is to invest in its people. A lack of education (especially for girls) and high dependency ratios are the essence of the lagging region phenomenon. · MENA has a proximity premium: its lagging areas' populations are rarely far from its growth hubs. This report shows how smart invest- ments and policies in transport can connect poor places to the dynamic economies of their rich neighbors. There is also a wide- open field of opportunity for telecommunications to bring electronic proximity to lagging areas. xvii xviii Preface · Many countries have spent huge sums on subsidies to entice investors to lagging areas--usually without any sustainable impact. This report recommends that governments turn their efforts toward the new approach to local economic development, which is gaining ground around the world, and is based on economic clusters, local competi- tive advantage, private initiative, and public-private dialogue. · MENA's historical legacy is one of centralized and sectorally- compartmentalized institutions. However, as the region's govern- ments increase their focus on integrated local development, they are exploring new institutional models that are more suited to their needs. The report describes the state-of-play in territorial planning, public financial management, targeted programs, deconcentration, and decentralization, and it sketches some emerging lessons. The World Bank is committed to the development of MENA's lagging areas. In Sohag, one of the Arab Republic of Egypt's poorest gover- norates, for example, the Bank financed community-identified local infrastructure such as schools and sanitation systems. The Republic of Yemen's rainfed agriculture project is raising farmers' yields and incomes in the highlands and in the wadis of Tihama and the Eastern Plateau. In Tunisia, the Bank has partnered with the Agency for Rural Development in the northwest since 1982. Our analytical and advisory services to MENA governments have included reports on the international experi- ence of technology poles, briefings on the latest trends in local economic development policy, an options paper on the allocation of investment funds for public infrastructure in Saudi Arabia, and econometric analyses of the causes of spatial disparities in welfare. This report combines the insights of specialists in the majority of the World Bank's key sectors: agriculture, development economics, educa- tion, health, poverty analysis, social protection, and transport. It is the report's modest aim, if not to offer a single formula for reducing spatial disparities, at least to propose a range of policy options that the region's leaders can reflect on in the light of their national objectives. Shamshad Akhtar Vice President, Middle East and North Africa The World Bank Acknowledgments This report was prepared by a team led by Alex Kremer and including Loďc Whitmore, Ted Alcorn, Hayat Al-Harazi, Omar Aloui, Brian Blankespoor, Heba Elgazzar, Rose-Marie Esber, Luca Etter, Fatou Fall, Hassene Kassar, Hiromichi Katayama, Stephen Kennedy, Mourad Kha- ladi, Stephen King, Anat Lewin, Andrea Liverani, Abdoul Mijiyawa, Claudio Montenegro, Siddig Muneer, Charbel Nahas, Ziad Nakat, Mag- alie Pradel, Rachidi Radji, Gudivada Rao, Joana Silva, Estelle Som- meiller, Mehmet Tohsun, Michael Wilson, Serdar Yilmaz, and Ayman Zohry. The peer reviewers were Chorching Goh, Daniel Lederman, and Marisela Montoliu Muńoz. Many people provided advice, but the team gives special thanks to the representatives of Middle East and North African (MENA) countries at the consultations in 2008 and 2009 in Marseille, as well as Indermit Gill, Marisela Montoliu Muńoz, Nada Eissa, and the OECD Public Gover- nance and Territorial Development Group. Andrew Stone, Ruslan Yemtsov, and Vivian Hon provided valuable technical advice. Jeff Leck- sell of the World Bank's Printing, Graphics, and Map Design Unit assisted in the formulation and finalization of the maps. The team acknowledges the support of the multi-donor Gender Action Plan. xix Abbreviations ABB Activity-Based Budgeting AERI Agricultural Exports and Rural Income ARC Appalachian Regional Commission ASEAN Association of Southeast Asian Nations BRT bus rapid transport CCT conditional cash transfer CGDR Commissariat Général de Développement Régional (General Commissariat for Regional Planning) CIAT Interministerial Committee for Territorial Planning DALYs disability-adjusted life years DECRG Development Economics Research Group (of the World Bank) DHS Demographic and Health Survey DT Tunisian dinar EEI Egyptian Education Initiative ELMO index a measure of economic justice; acronym formed from the surnames of its inventors: Elbers, Lanjouw, Mistiaen, and Özler ERVET Ente Regionale per la Valorizzazione Economica del Territorio (Regional Institution for Territorial Economic Development) ESFD Economic and Social Fund for Development ESSBs Établissements de soins de santé de base (primary health care facilities) EU European Union EUN Egyptian Universities Network FAO Food and Agriculture Organization FAOSTAT FAO statistical database FNS Fund of National Solidarity GAEB General Authority for Education Buildings xxi xxii Abbreviations GCC Gulf Cooperation Council GDP gross domestic product GIS geographical information system GNI gross national income GRDP gross regional domestic product HIECS Household income, expenditure, and consumption survey HRM Human resource management ICT information and communications technology IFPRI International Food Policy Research Institute IHSES Iraq Household Socio-Economic Survey IMCI Integrated Management of Childhood Illnesses IMR infant mortality rate INDH National Human Development Initiative INS Institut National de la Statistique (National Statistics Institute) (Tunisia) ISSIP Integrated Sanitation and Sewerage and Infrastructure Project ITU International Telecommunications Union JEI Jordan Education Initiative LPI Logistics Performance Index MAUP Modifiable Areal Unit Problem MCIT Ministry of Communications and Information Technology MDCI Ministry of Development and International Cooperation MEHAT Ministry of Environment, Habitat, and Territorial Planning MENA Middle East and North Africa MENA13 Specifically Algeria, Djibouti, the Arab Republic of Egypt, the Islamic Republic of Iran, Iraq, Jordan, Lebanon, Libya, Morocco, the Syrian Arab Republic, Tunisia, the Republic of Yemen, and the West Bank and Gaza MERCOSUR Southern Cone Common Market (Mercado Común del Sur) MoE Ministry of Education MPO Management and Planning Organisation NAF National Aid Fund NAFTA North American Free Trade Agreement NHHS National Household Survey NRENs National Research and Education Networks NRRP National Rural Roads Program Abbreviations xxiii NSDP National School Development Plan ODESYPANO Office for Forestry and Livestock Development in the northwest OECD Organisation for Economic Co-operation and Development ORMVAs Offices Régionales des Mise en Valeur Agricole PCBS Palestinian Central Bureau of Statistics PEM public expenditure management PHC primary health center PISA Program for International Student Assessment PMT proxy means testing POS point-of-sale PPP purchasing power parity SDIs School Development Initiatives SEZs special economic zones SMEs small and medium enterprises SNAT T erritorial Master Plan (Schéma National d'Aménagement du Territoire) SSN Social Safety Net TFR total fertility rate TIMSS Trends in International Mathematics and Science Study TME total managed expenditure UK United Kingdom ULO Urban Land Organization UN United Nations UNDP United Nations Development Programme UN-WIDER United Nations World Institute for Development Economics Research USO universal service obligation WDI World Development Indicators WDR World Development Report WFP World Food Programme (of the UN) WUAs Water Users' Associations WVS World Values Survey ZDR Zones de Développement Régional All dollars are in US$ unless otherwise noted. CHAPTER 1 Overview This is a summary of the World Bank's 2010 Middle East and North Africa Development Report on spatial disparities. The full report has more in- depth analysis and examples from the region and beyond. It will be avail- able online and from booksellers in August 2010. For the international experience of spatial development and regional policy, readers may be interested in the World Bank's 2009 World Devel- opment Report, Reshaping Economic Geography. The Political Demand: Spatial Equity with Productivity Leaders in the Middle East and North Africa (MENA)1 are concerned about the development gap between leading and lagging areas. Spatially uneven development may be seen as a failure of economic justice, or as a betrayal of the social contract between the government and the gov- erned, and that migration may translate into urban discontent or changes in political and social relations. But MENA governments have another concern that is no less press- ing. Unemployment rates are on average more than twice as high as those in middle-income countries worldwide. Creating jobs requires competitiveness and growth. Can MENA raise living standards in lagging areas without compro- mising economic productivity? This report answers yes and proposes a policy framework for doing so. Spatial Concentration with Spatial Equity Everywhere in the world, economic activity is spatially concentrated for the simple reason that production is more efficient when it is agglomerated. 1 2 Poor Places, Thriving People As a result, 54 percent of the world's economic activity is concentrated in countries that make up 10 percent of its area; within countries, eco- nomic activity is squeezed into the major cities. In MENA, the same pattern exists: the productivity of capital and labor is measurably higher in the Arab Republic of Egypt, Morocco, and Tunisia's leading areas. But we have to distinguish two spatial patterns: · Agglomeration--the tight concentration of businesses in certain places · Spatial disparities--the gaps in living standards between people in dif- ferent places The broad pattern of spatial disparities in the Middle East and North Africa is typical of low- and middle-income countries: living standards are highest where economic activity is densest and where connections to economic density are closest. · Urban areas are better off than rural areas. · Areas close to urban centers are better off than isolated areas. · The same pattern of spatial disparities applies to indicators of both economic and human development. · The same pattern of disparities is found within subnational divisions. · Women and men experience spatial inequalities differently. However, living standards between rich and poor places can converge even as economic agglomeration grows. With effective policies in place, disparities in social indicators--such as poverty, health, and education-- can and do shrink. Spatial Disparities Are a Political Issue If spatial disparities are a by-product of economic growth, why are they of such concern to policy makers? From a sociopolitical perspective, spatial disparities can have worri- some effects. They can become a source of political grievance for resi- dents of low-income areas. In addition, by fueling migration, spatial inequalities bring political risks to the better-off areas where the migrants settle. The policy maker's challenge is, therefore, to sustain spatial and politi- cal balance without sacrificing the efficiencies gained from agglomeration. The aim of this report is to offer MENA's policy makers a framework for doing so. Overview 3 A Policy Framework for Spatial Equity with Economic Efficiency The solution to the spatial disparity problem lies in careful local analysis and an appropriate mix of policies tailored to the characteristics of each lagging area. Adapting the 2009 World Development Report, this study examines three policy packages (summarized in table 1.1): Package 1 levels the playing field and the opportunity for human devel- opment in lagging areas by allowing better access to social services and an improved environment for business. Policy Package 1 applies to all lagging areas because it reduces spatial disparities without com- promising the efficiencies of spatial agglomeration. Package 2 connects poor places to the poles of development. Policy Package 2 applies to lagging areas that are close to leading areas, where spillover effects can reduce spatial disparities. Package 3 underpins private sector interest in nonleading areas and does so without relying on subsidies. Policy Package 3 applies to lagging areas where competitive potential is manifested by private investor interest. The aim is encouraging local economic growth to reduce spatial disparities. Spatial Disparities in Living Standards: Uncovering the Truth The spatial concentration of MENA economies is apparent to all. But the reality of spatial disparities--differences in living standards-- requires careful examination. To match the policy package to the place, TABLE 1.1 Matching Policy Packages to Lagging Areas Policy package Area of applicability Package 1. Level the playing field All lagging areas and the opportunity for human development in lagging areas Package 2. Connect poor places to Lagging areas close enough to the poles of development leading areas to benefit Package 3. Underpin private Areas where the private sector might sector interest in nonleading areas see unrealized growth potential Sources: Study Team and World Bank 2009. 4 Poor Places, Thriving People policy makers must understand the unique characteristics of each lagging area. Some pointed questions can help indicate the way forward. First, is geography really a major contributor to socioeconomic inequality? Sec- ond, is an area lagging because of low economic opportunity or because of poor human development? Third, assuming a genuine spatial inequal- ity problem does exist, how has it evolved over time? Are lagging areas converging with leading areas, or are disparities becoming wider? Is redistribution of population from lagging to leading areas helping to solve the problem? This analysis will grapple with the problem of diagnosing spatial inequality, showing first how to identify degrees of spatial inequality and then how to separate lagging areas into distinctive types. Fast Urbanization but a Delayed Transition out of Agriculture MENA differs from other developing regions in at least one key respect. The region's populations are more agglomerated (urbanized) than one would expect for their level of economic development. Why is that? For one thing, most of the land in the MENA region is too arid or mountainous to support settlement and farming. Consequently, rural populations here have historically been concentrated in coastal strips, in mountain valleys, and along the rivers. By necessity, therefore, MENA's settled areas have had high population densities from an early date. One implication of this fact is that the region needs less additional population agglomeration to obtain the economies of agglomeration found in developed countries. MENA is urbanizing fast: on average, countries were 65 percent rural in 1960 but 65 percent urban in 2007. Normally, urbanization is accompanied by a shift of labor from agriculture to services and manufacturing. In MENA, that labor transition has been delayed. Agriculture has lost its share of value added, but rural areas remain centers of low-productivity employment and poverty. Moreover, MENA has an unusually high rate of agricultural employment growth (figure 1.1), and too many people on too little land has become a common feature of MENA's lagging rural areas. One rea- son is that the region has an unusually high rate of growth in the working-age population (figure 1.2). Given the land and water con- straints, the region's agricultural sector cannot continue to absorb surplus labor. A shift of labor out of agriculture is overdue; when it comes, it will shift population into the cities. Climate change is likely to accelerate this process. Overview 5 FIGURE 1.1 MENA's Agricultural Labor Force Is Growing Fast Relative to Its Level of Development 6 4 annual agricultural labor force growth (percent) ALG SYR 2 YEM JOR IRN TUN EGY 0 100 1,000 MOR 10,000 100,000 ­2 ­4 LEB ­6 ­8 ­10 GDP per capita 2000 ($ 2005) Source: World Development Indicators Database, accessed May 17, 2009. Note: ALG = Algeria; EGY = Egypt, Arab Rep.; IRN = Iran, Islamic Rep.; JOR = Jordan; LEB = Lebanon; SYR = Syrian Arab Republic; TUN = Tunisia; YEM = Yemen, Rep. FIGURE 1.2 MENA's Working-Age Population Has Been Booming Compared with Other Countries and Economies at the Same Level of Development 6 annual growth of population aged 15-64, 5 JOR YEM 4 SYR 1990-2005 (percent) ALG IRAN 3 DJI TUN LEB 2 EGY MOR 1 0 ­1 ­2 100 1,000 10,000 100,000 GDP per capita 2007 (constant 2005 international $) Source: World Development Indicators Database, accessed May 29, 2009. Note: DJI = Djibouti; MOR = Morocco. 6 Poor Places, Thriving People Spatial Disparities Are Smaller Than They Seem Geography is highly political, so it is essential for policy makers to have an objective understanding of how location affects household welfare. Things are not always as they seem at first glance. For example, eco- nomic elites bump up the average consumption levels in leading areas, but the living conditions of the vast majority of people in those areas may actually be similar to those in lagging areas. Urban slums may have lev- els of poverty and human development similar to or even worse than rural areas. Simply by using median instead of average values of house- hold welfare can alter the appearance of spatial disparities. It is also important to focus on absolute poverty numbers as opposed to poverty rates. Focusing on poverty rates can overlook the large absolute numbers of poor people living in high-density urban areas. Location matters more with respect to living standards in some countries than others. The spatial component of inequality is biggest in Morocco, fol- lowed by Egypt and the Republic of Yemen. Nevertheless, in no MENA country or economy for which data were available does rural-urban inequal- ity account for more than 20 percent of total inequality of household expen- diture. Overall, MENA's urban-rural and interprovincial divides are no greater than those in other developing regions of the world. Spatial Disparities Are Often Social Disparities Household characteristics explain a large share of spatial disparities. The following example will help to clarify the concepts at work. Although Upper Egypt has only 40 percent of the population of Egypt, it accounts for 60 percent of its poverty and 80 percent of its severe poverty. Eco- nomic geography, however, is not the main factor holding Upper Egypt back. In fact, demography explains 62 percent of the consumption gap between Upper Egypt and Lower Egypt. The problem is that Upper Egyptian workers have to share their earnings with more family mem- bers who are too young or too old to work. In other words, Upper Egypt has a high "demographic dependency"--the ratio of dependents to peo- ple of working age. Households with unfavorable demographic charac- teristics would be at a disadvantage wherever they lived. Likewise, a household's education level can influence its success. While leading areas may have better-educated people and lagging areas may provide inadequate educational opportunity, households with low educational achievement will do worse than better-educated households, regardless of location. Three factors explain why household characteristics in MENA's lag- ging areas differ so greatly from those in leading areas: fertility, sorting, and education. Overview 7 MENA's demographic transition has been and remains spatially uneven. Although fertility rates are falling across the region, they began falling in leading and urban areas first and in lagging and rural areas later. In the short run, therefore, MENA's demographic transition has caused spatial disparities in dependency ratios, and this disparity is a major cause of the region's rural-urban and interprovincial disparities in well-being. As the demographic transition reaches MENA's lagging areas, however, it will become a major contributor to spatial convergence in living standards. Many factors influence fertility rates, including female education, gender empowerment, female labor participation and income growth, and access to quality reproductive health services. Spatial considerations influence all of these determinants, which explain why MENA's demo- graphic transition has been geographically uneven. The available data on internal migration in MENA confirm the sort- ing effect: the people with the highest earning potential leave lagging areas. This is due partly to the tendency of people with greater education to place greater value on city amenities, and partly to the fact that the jobs that make education productive tend to be concentrated in cities. Finally, people in lagging areas are less able to acquire human capital, whether from the educational system or from family members and social networks. This is the education effect. These three effects--fertility, sorting, and education--are related and mutually reinforcing, and they account for much of what makes a house- hold rich or poor. Improving living standards in lagging regions is, there- fore, as much a question of developing people as it is of developing the place itself. The demographic and educational components of spatial disparities differ significantly from country to country, as shown in the following examples. · Egypt. Demography is a major contributor to Egypt's disparities. If all of Egypt's governorates had the same demographic profile, the rural- urban gap would be a full 17 percentage points lower than it is. Edu- cational disparities are also significant, albeit less so than demography. · Djibouti. Unusually, the demography of rural Djibouti mitigates the rural-urban gap. If rural and urban areas had the same demographic characteristics, the gap would be 4 percent higher than it is. Con- versely, educational disparities add 11 percentage points to the rural- urban gap. · Jordan. Average rural per capita consumption is 24 percent behind urban levels in Jordan. If demographic disparities are removed, this difference dwindles to an insignificant 12 percent. Controlling for educational disparities, the gap becomes just 3 percent. 8 Poor Places, Thriving People · Lebanon. In looking at what separates Mount Lebanon, the North, and Bekaa from Beirut, the differences in household characteristics and earnings per worker are not significant. Labor force participation and mobility are the key factors. · Morocco. Morocco features only modest interprovincial disparities but a significant urban-rural gap: per capita household consumption in rural areas is only 54 percent of that in Morocco's urban areas. Even if rural and urban areas had the same demographic characteristics, the rural-urban gap would still be 41 percent. And controlling for education, rural Morocco's per capita household consumption would still be 32 percent lower than urban Morocco's. · The Syrian Arab Republic. Syria's northeast regions lag far behind Damascus City in terms of average per capita household expenditure. Without demographic disparities, however, the northeast would not be a lagging area at all. Spatial disparities in education seem to play no role. · The Republic of Yemen. Without the demographic and educational components of intergovernorate disparities, the Republic of Yemen's poorest governorate would trail only 40 percent (instead of 65 per- cent) behind the city of Sana'a in terms of per capita consumption. The rural-urban gap would be 25 percent instead of 40 percent. The implication for policy makers is clear: one cannot assume that observed spatial disparities in economic welfare are due to spatial dispar- ities in economic opportunity. For many lagging areas, the population would be disadvantaged wherever they lived, so human development is likely to be a key component of effective spatial development almost everywhere in MENA. Convergence and Migration: Sometimes Spatial Disparities Resolve Themselves Once policy makers have a clear profile of spatial disparities, it can be useful to know whether those disparities are narrowing or widening over time (box 1.1). In some countries, the redistribution of population from lagging to leading areas can take place quickly enough to solve the lagging-area problem. If this is the case, then it makes sense to focus economic invest- ment wherever the return is highest, which is usually in agglomerations. In other countries, the population will not shift from lagging areas to leading areas, perhaps because lagging areas have higher fertility rates or because inflexible social structures limit migration. In this situation, the lagging areas may require spatially targeted economic development poli- cies to bring about an increase in earnings. Overview 9 BOX 1.1 Convergence Trends for the Islamic Republic of Iran, Tunisia, and the Arab Republic of Egypt The Islamic Republic of Iran. In the Islamic Republic of Iran, the convergence in living standards between leading and lagging areas was more a result of leading areas losing ground owing to economic decline than of lagging areas gaining from growth. The urban-rural convergence occurred during 1976­85, when per capita household expenditure fell much more in urban areas than in rural areas. Once the economy resumed steady growth in the early 1990s, however, the urban-rural gap widened once again. Tunisia. Between 1980 and 2000, Tunisia experienced significant interregional convergence. The rural-urban gap also narrowed, although it fluctuated from year to year, probably according to rainfall. Tunisia's convergence was uneven, however, as regions split into rich and poor clubs. The case of Tunisia shows that spa- tial convergence at the national level may take the form of a "breakout" by middle-ranking regions. For that reason, policy makers should resist the temptation to focus their convergence concerns solely on the most lagging regions. Egypt. The country appears to be facing widening spatial dispar- ities within lagging Upper Egypt, even though divergence is not taking place between Upper and Lower Egypt. Some villages in rural Upper Egypt are being left behind as others "break out" and reduce their poverty rates. Therefore, Egypt's current strat- egy of targeting social investments toward low-income villages in low-income governorates may indeed be appropriate. Source: Authors. Throughout the MENA region, people are moving from the country- side to the cities. The redistribution of MENA's population, however, tends to take the form of a gradual shift from rural to urban areas within provinces instead of a lateral migration from lagging provinces to leading provinces. Why is the redistribution of population from lagging to leading provinces slow in MENA? One reason is that fertility rates are higher in lagging provinces. The role of social networks may also be critical. In the absence of transparent job and housing markets, the opportunity to migrate may depend on the presence of friends and family in the desti- nation. So while the poorest may have the greatest desire to migrate, the best connected have the greatest opportunity to do so. Studies of Egypt 10 Poor Places, Thriving People confirm the importance of social networks in determining internal migration flows. Gender relations in MENA may also slow interprovin- cial migration. Social conventions constrain female mobility and labor market participation. From the preceding considerations, we can see that the population in MENA is urbanizing but not "metropolitanizing." Even though political concerns may focus on the specter of an uncontrolled human tide flooding the primary cities, this picture is not the reality of inter- nal migration in MENA. The implication is that lagging provinces' development will depend on local growth poles for some time to come. Three Ways to Lag, with Three Policies to Match Lagging areas' diverse geographical characteristics call for different pol- icy responses. For the purposes of spatial analysis, MENA's area can be classified along three dimensions. · Lagging versus nonlagging. The question of living standards · Dense versus sparse. The question of population density · Near versus distant. The question of travel time to a major city A map of the region that uses these classifications emphasizes MENA's varied geography (map 1.1). Much of MENA's population is clustered in islands of density. The 3 percent of MENA's surface area that is densely populated is home to 92 percent of the population. Thus, most lagging areas' populations are reasonably agglomerated. The proximity of populations to economic density is another distinctive characteristic of MENA. Once we know a lagging area's character, we can match it to the appropriate policy package. For example, in the lagging area of Djelfa in Algeria on the Trans- Saharan Highway, Djelfa looks like a fringe area: sparsely populated, dis- tant from economic density, and with economic potential constrained by a lack of agglomeration and resources. Djelfa may, therefore, be suited only for Policy Package 1 (level the playing field and the opportunity for human development in poor places). The Jordanian governorate of Al-Mafraq is another lagging area. Like Djelfa, it is sparsely populated and arid. Yet its provincial capital is only 50 kilometers from Jordan's capital. Al-Mafraq is, therefore, a belt area. It could benefit from Policy Package 1's level playing field for development, Overview 11 but it is also close enough to the capital to benefit from spillover con- nectivity. Therefore, Policy Package 2 might be appropriate. The urban governorate of Aleppo in Syria is a lagging area although it has a large population of educated people, several major employers, and an environment of fertile farmland. It would seem to be a pocket of unrealized economic potential. As well as being a candidate for Policy Package 1, therefore, it could be a candidate for Policy Package 3, a pack- age that is aimed at facilitating private investment in lagging areas. Analyzing a given lagging area at different levels of spatial disaggre- gation also reveals new policy-relevant information. Upper Egypt is a case in point. As a region within a country, rural Upper Egypt appears as a lagging area. Yet 80 percent of the difference in per capita consumption between Upper Egypt and Lower Egypt is due to demography. At the regional level, therefore, the policy emphasis should be on overcoming Upper Egypt's demographic handicap. As a governorate, Upper Egypt shows wide disparities from district to district in access to public services and health care, so it is crucial for authorities to allocate public social investment projects to priority areas. The government's logical policy response has, therefore, been to empha- size bottom-up, participatory planning of social infrastructure, with financial allocations targeting the poorest districts. As a collection of communities, Upper Egypt can benefit from the new technique of poverty mapping, which allows policy makers to analyze spatial disparities below the province level. In MENA, this technique reveals the disparities in welfare between neighboring villages, as well as geographically small pockets of dense poverty in large cities. At this level, policy responses should include village-level targeting of public social investment projects and targeting of dense poverty pockets in richer urban areas. With the importance of diagnosis for matching policies to places made clear, we are now ready to present the three policy packages described at the outset. Policy Package 1. Level the Playing Field and the Opportunity for Human Development in Poor Places All lagging areas have one thing in common: they deserve equal policy treatment--a level playing field for development, one that will reduce spatial disparities in living standards while letting the market determine the spatial distribution of economic activity, thus securing both the eco- nomic benefits of agglomeration and the political benefits of spatial 12 Poor Places, Thriving People MAP 1.1 A Policy-Oriented Spatial Typology of the Middle East and North Africa lagging dense sparse nonlagging population extremely low population within 3 hours of a major city Sources: Elevation, SRTM30 (2000); land cover, GLC2000 (2000); population, Landscan (2005); rail, VMAP0 (1997); rivers, CIA World Data Bank II (1980s); roads, Euro-med (2000) and VMAP0 (1997); slope, in degrees, SRTM30 (2000); urban areas, GPW3-GRUMP (2000); water bodies, Global Lakes and Wetlands Database layer 1 (2004). Note: Lagging areas represent 40 percent of the administrative units by country that have the highest rates of poverty, and nonlagging areas represent the remaining administrative units. Using Nelson's (2008) methodology to calculate a travel time model, "near" is defined as a travel time less than or equal to 3 hours from a city of 500,000, and "distant" refers to any travel time greater than 3 hours. According to population density data from Landscan (ORNL 2005), "dense" is defined as more than 50 people per square kilometer, and "sparse" covers the remainder. Areas with population density less than 1 are displayed in white. equity. This is the goal of our first policy package, which is applicable in all circumstances and the only package suitable for fringe areas. History Has Left an Uneven Playing Field for Development There is a strong statistical relationship between political institutions and spatial development: the less accountable a state, the more agglomerated its population. This finding might be because undemocratic states tend to neglect people in outlying areas. Overview 13 MENA's political history has left behind a spatial bias in favor of met- ropolitan areas. Many of the contemporary drivers of spatial bias go back to the colonial era. At independence, most of today's MENA states were founded on belief in the state's capacity to regulate social and economic relations. Governance under Ottoman and European rule was based on the devel- opment of strong central bureaucracies. Colonial tax systems encouraged a neglect of lagging regions and shifted a large share of tax revenues away from the provinces to the capital region. In the years immediately fol- lowing independence, central administrations expanded, and power remained concentrated at the capital. In catering essentially to the metropolis, colonial management avoided redressing regional inequalities. Under both Ottoman rule and European colonization, states served as providers to the metropolis. In favoring resource extraction over economic development, the colonial 14 Poor Places, Thriving People powers created an economic geography of "useful versus useless" regions. Disadvantaged people were divided along ethnic and religious lines to prevent their alliance against the government. The development of metropolis- and export-oriented economies subsequently encouraged coastal agglomeration. Empirical metrics suggest that the MENA states are less externally accountable than other countries at similar levels of development. With- out formal arrangements for spreading accountability, decision makers are most likely to be accountable to those with whom they deal directly in the capitals--giving rise to an urban bias. And recently, some countries' oil and gas revenues placed huge amounts of resources in the hands of the cen- tral administration. These factors encouraged centralization, accentuating features that modern MENA states inherited from their colonial past. MENA's governments are now seeking to reverse the political drivers of spatial disparities in response to several developments, including rising urban unemployment; the relatively poor performance of centrally planned economies worldwide; the limited ability of non-oil-exporting countries to subsidize public employment and production; and the spread of telecommunications, migration, and the broadcast media, which has made the populations of lagging areas more aware of spatial inequalities. To create a level playing field for development, MENA's policy mak- ers will, therefore, need to remove the historical disadvantages faced by populations in peripheral areas. The approach can be broken down into four action areas: · Improving the investment climate in lagging areas · Integrating the spatial dimension into public expenditure management · Making education, health, and safety nets serve lagging-area populations · Facilitating urban growth How to Level Subnational Disparities in the Investment Climate Data for the MENA Region suggest that disparities in the investment climate (consisting of the institutional context and infrastructural base) are affecting the spatial distribution of economic activity. Studies in Morocco and Egypt found an empirical link between spatial disparities in the investment climate and spatial disparities in firm productivity. We cannot be sure from the data that there is a systematic investment climate bias against lagging regions. It is clear, however, that MENA has signif- icant spatial disparities in the investment climate, which have the poten- tial to influence the location of economic activity (figure 1.3). Overview 15 FIGURE 1.3 A Composite Index of Objective Measures of the Subnational Investment Climate in Algeria, the Arab Republic of Egypt, Jordan, Morocco, and the West Bank and Gaza Région Ouargla Algeria Région B Boumerdčs Egypt, Arab Rep. Qalyubia Jordan Région D Oran Région H Ghardaďa Morocco Kafr el-sheikh West Bank and Gaza Gharbia Région C Blida Région A Algiers Beheira Région G Sétif Région F Bordj BA Beni-Suef Gaza Strip Port Said Amman Cairo Irbid Région E Annaba West Bank Suez Région Tlemcen Zarqa Minya Alexandria Sharqiya Giza Damietta Menoufia Rabat/Casablanca Tanger-Tétouan Dakahlia Ismailia Fčs Settat 0 10 20 30 40 50 60 70 80 90 100 Source: Staff calculations from World Bank investment climate surveys. How can policy makers begin to level the playing field for private investment across MENA' Possibilities include more transparency in the application of business regulations in lagging areas and the provision of critical infrastructure services in their cities. Not all lagging areas have the economic potential to benefit from improvements in the investment climate. Evidence from Egypt and Morocco, however, suggests that improvements in the institutional environment could be achieved at a relatively low cost and would strongly benefit enterprise productivity. Public Expenditure Management Has a Spatial Dimension Several countries outside MENA have systems and rules for tracking the spatial incidence of expenditure and for ensuring that intragovernmental 16 Poor Places, Thriving People transfers reflect different areas' needs. But MENA's spatial public expen- diture management systems are a clear priority for action. Most MENA countries or economies cannot monitor where central government spending takes place, which is in itself a problem for spatial policy. Moreover, the systems for allocating fiscal transfers to local gov- ernments are usually ad hoc, with the result being that poorer areas are often not compensated for their high level of needs or low resource base. For Egypt and Jordan, the data suggest that resource transfers and gov- ernorate GDP per capita were not correlated: the highest transfers were not given to governorates with the weakest economic base. Conversely, the data from the Islamic Republic of Iran suggest that the apportionment of trans- fers from the central government to ostans (provinces) does give priority to lagging areas. Correction of Spatial Disparities in Public Education Systems Rural-urban disparities in primary and secondary enrollment rates (and other indicators of educational attainment, such as school completion and math achievement) are an obvious feature of MENA's education pro- file (table 1.2). But spatial analysis of the determinants of educational performance is critical; it is not just a question of building more schools in lagging areas. What causes the spatial disparities in education? To what extent are they due to spatial disparities in children's socioeconomic status (demand-side factors) and to what extent to spatial disparities in educa- tion services (supply-side factors)? We examined the determinants of educational attainment (staying in school) in one country with high spatial disparities in attainment (Morocco) and in another with low disparities TABLE 1.2 Lower Enrollment Rates for Rural Children, Especially the Poor (Percent) Urban Rural Poor Nonpoor Poor Nonpoor Algeria 1995 Primary 96.0 95.0 89.0 89.0 Secondary 77.0 82.0 59.0 66.0 Egypt, Arab Rep. 1999 6­15 95.8 98.5 93.5 96.7 15­19 72.4 84.9 64.7 72.9 Iran, Islamic Rep. 2001 6­10 99.0 100.0 98.0 98.0 11­13 92.0 97.0 76.0 84.0 Morocco 1998 7­15 69.4 87.2 36.4 49.8 Tunisia 2000 6­18 79.4 82.2 67.0 70.7 Yemen, Rep. 1998 10­14 83.0 92.1 59.6 62.0 Source: World Bank 2008. Overview 17 (Jordan). The results for Morocco suggest that after the respondent's age and the household head's level of education were taken into account, the rural-urban factor explained only a very small percentage of attain- ment gaps. Instead, most of the disparities were explained by household characteristics. Similarly in Jordan, household economic status was the main determinant of attainment. Put simply, this finding means that the priority is to make it easier for poor households to keep their children in school, not to build more facilities in poor places. While spatial disparities in educational attainment are mostly due to household characteristics, spatial disparities in educational achievement (learning) are largely due to geographic differences in the quality of schooling. It seems that students in rural schools receive education of lower quality than do those in urban schools. Although there have been spectacular increases in enrollment in the MENA Region, the economies have been less effective than comparator economies in con- verting those investments and enrollments into retention and learning. The idea that spatial disparities in learning achievement might be largely due to geographical differences in educational quality reflects the same diagnosis, but this is at the subnational rather than the inter- national level of analysis. Gender disparities in education are also wider in rural areas in several MENA economies. Reducing these disparities will help reduce spatial disparities in development overall. These analyses have important implications for policy makers seeking to reduce spatial disparities in education. They lead to four priority axes to combat spatial disparities in education. · Reducing spatial disparities on the demand side · Reducing spatial disparities in quality · Reducing gender disparities in lagging areas · Using spatial planning techniques to deploy appropriate infra- structure Each theme is illustrated with examples from our case studies on Egypt, Jordan, Morocco, and the Republic of Yemen. Reducing Spatial Disparities through Demand-Side Interventions We have already seen that the low economic status of children in lagging areas may be more of an obstacle to educational attainment than are dis- parities in educational supply. Therefore, one strategy for reducing spatial disparities is to make it more economically attractive for families to keep their children in school. For example, in conditional cash transfer pro- grams, the transfer of a welfare benefit depends on the receiver's actions. 18 Poor Places, Thriving People The education components of such programs address demand-side obsta- cles to education by contributing toward the costs of sending a child to school. Data from Mexico, Colombia, and Nicaragua clearly show the potential of conditional cash transfer programs to improve enrollment. Reducing Spatial Disparities in Quality The spatial deployment of teachers and administrators creates inequities in the quality of education. In Morocco, difficult living conditions in rural areas lead young teachers to move to urban areas as soon as they can. In addition, levels of absenteeism are high, and teachers are underused, working less than the legal minimum of 24 hours per week. In Jordan, teaching posts are awarded to those who have waited the longest to get the next job in line. In Egypt, as in Morocco, teachers migrate from rural to urban areas as they gain experience. Finally, in the Republic of Yemen, teachers are haphaz- ardly placed on the basis of local influence, so schools in lagging areas are unlikely to obtain experienced teachers. It is clear that improving human resource management in public edu- cation services is critical for reducing spatial disparities in education qual- ity. Local recruitment of teachers and incentives such as improved living conditions and housing are necessary to draw experienced teachers and principals to lagging areas. And more formal, systematic, and needs-based posting procedures are needed to better match teachers to vacancies. "School development initiatives" have demonstrated great potential for improving education quality in targeted geographic areas. These initiatives are development plans managed at the school level by school boards, teachers, and parents. This approach relies on the capacity and motivation of the local school governance committee. It also presupposes that school principals have the skills and motivation to act as community leaders. Reducing Gender Disparities in Education in Lagging Areas The interaction of gender disparities and spatial disparities creates wide gender gaps in lagging areas. In Morocco, for example, the gender gap for literacy is 5 percentage points in urban areas, compared to 35 per- centage points in rural areas. Reducing spatial disparities in girls' educa- tion would also remove one of the causes of spatial disparities in fertility rates, which would, in turn, reduce spatial disparities in dependency ratios and living standards. Any intervention aimed at reducing spatial-gender disparities in edu- cation should be made with girls' unique needs in mind. For example, programs in the Republic of Yemen are improving living conditions for female teachers in lagging areas and recruiting women teachers locally. Conditional cash transfer programs can be designed to provide families Overview 19 with special incentives to keep girls in school. Spatial planning of public investment in educationsuch as shorter distances between schools and residences, and adequate water and sanitation can reflect girls' educa- tional needs. Using Spatial Planning Techniques to Improve the Spatial Efficiency of Infrastructure The race to achieve the Millennium Development Goal of full primary level completion has given strong emphasis to school construction across the MENA region, but the need for robust spatial planning of investments is becoming urgent. As the goal of universal primary enrollment and com- pletion comes close to achievement, policy makers are shifting their focus to secondary schools, which are more expensive and farther from home. In addition, population distributions are changing rapidly with urbanization and the spread of the demographic transition from leading to lagging areas. The situation in several MENA economies bears out the assertion that robust spatial planning is necessary. The Republic of Yemen faces a number of spatial imbalances in education. Despite Egypt's high popula- tion densities, there are significant spatial inequalities in access to educa- tional infrastructure. Jordan suffers from a serious spatial disequilibrium in the distribution of basic and secondary schools, leading to high main- tenance costs and low utilization rates. Health: Correcting Spatial Disparities in Outcomes Geographical differences in health status are a major concern for policy makers. MENA governments spend an average of 3.3 percent of GDP on health services, and, as a matter of equity, the impact of that spending should not depend on a person's location. Moreover, the performance of health systems in the MENA region will increasingly be measured in terms of each system's ability to reach marginalized people. Changing Health Services and Health Outcomes in Poor Places Large spatial disparities exist within the MENA economies in health out- comes and access to services. There are localized pockets of bad health indicators in countries such as Lebanon, Egypt, Tunisia, Jordan, and Morocco. Health care personnel are concentrated in urban areas, which are likely to offer higher wages, better living conditions, and greater opportunities for professional development; however, low-density areas have more facilities per capita. This situation may be due to the desire of policy makers to make facilities geographically accessible to all citizens, which inflates the proportion of facilities in areas of low density. If we put these two trends together, rural facilities are understaffed and underused 20 Poor Places, Thriving People compared with those in cities. Thus, building facilities in lagging areas is not the answer. Bringing Better Health to Lagging Areas in Different Ways How, then, can MENA governments reduce the health status divide between leading and lagging regions? The key is improving the quality of health service delivery, rather than its scale. When it comes to strengthening public health delivery facilities in lag- ging areas, successful interventions have focused on staff incentives and nonstaff recurrent funding. Morocco's Integrated Management of Child- hood Illnesses program, for example, focused explicitly on staff motivation, with training and free housing for personnel in lagging areas. Spatially adapted delivery systems have also proved cost-effective for low-density areas. For example, the Islamic Republic of Iran has strengthened com- munities through training of local health workers to provide family plan- ning and reproductive health care. As a result, service coverage in rural areas has increased from nearly 40 percent to 90 percent in 15 years. Empirical evidence from countries as disparate as Costa Rica, the Islamic Republic of Iran, Singapore, Sri Lanka, and the United King- dom reveals the efficacy of demand-side interventions in improving health status in lagging areas. A multitude of demand-side factors may be limiting the expansion of health services in lagging areas, notably limited insurance coverage. MENA's health insurance systems tend to be limited to wealthier urban residents. Strategies that aim to expand education, transportation, nutrition, water, sanitation, and incomes in lagging areas may have a stronger effect on spatial disparities in health outcomes than a simple emphasis on the stock of health facilities. In Egypt, programs have focused on increasing women's awareness of health issues and empowering women to take advantage of family plan- ning services. Expanding insurance plans into poor areas, as the Islamic Republic of Iran has done, can also encourage public and private providers to deliver appropriate care by stimulating demand. Improved management information systems can help health depart- ments ensure efficient targeting of resources. Implementation of better monitoring and IT systems in the Islamic Republic of Iran and the West Bank and Gaza resulted in strong improvements in resource distribution and usage rates of health facilities. How to Target Social Safety Nets to Mitigate Spatial Disparities Although many MENA economies have some kind of social safety net pro- gram, they are rarely considered a component of spatial policy. When well targeted, these programs can reduce spatial disparities, but poorly targeted programs actually increase the gap between rich and poor places. Overview 21 In terms of spending, energy (and food) subsidies are by far the major social transfers in the region; however, for policy makers concerned with spatial equity, a major drawback of commodity subsidies--and energy subsidies in particular--is that they are regressive. The people who consume the most, the nonpoor in leading areas, receive the lion's share of the subsidy. Conversely, social funds and conditional cash transfers have been shown to effectively target lagging areas. Geographical targeting dis- tributes benefits to people who are likely to be poor or vulnerable, depending on where they live. Proxy means testing uses easily observed household indicators to predict household income and to determine who is eligible for benefits. But these positive examples of transfer schemes are not prevalent in MENA, and many existing programs could contribute more to reducing spatial disparities if better targeted. How to Smooth the Path for City Growth The development of MENA's lagging rural areas depends on labor shifting out of farming into the cities. As MENA's farm workforce continues to grow, it becomes harder and harder for limited land and water resources to sustain its productivity. Two-thirds of the farm labor productivity difference between MENA and Spain is explained by MENA's higher ratio of labor to land. The data also show that agri- cultural productivity growth in MENA is driving farmgate prices down rather than farmer incomes up because supply is outstripping urban demand. In short, long-term rural development in MENA depends on urbanization. Urbanization helps to reduce spatial disparities by shifting labor from lagging to leading areas. Therefore, smoothing the path for urbanization is a recipe for spatial convergence in living standards. Urbanization is proceeding apace in MENA13, where the average rate of growth of the urban population is 3.3 percent per year, compared with 2.6 percent for national populations. The question, therefore, is not how to slow down the growth of cities, but rather how policy makers can ensure that urban- ization occurs as smoothly as possible. This approach involves three top- ics: land markets, urban public services, and institutional frameworks. Having Land Markets that Support Smooth Urban Growth Well-functioning land markets are critical for smooth urban develop- ment. In MENA, however, there are signs that land markets could be performing better. Land for housing. While the supply of housing in MENA compares favorably with other countries at a similar level of development, MENA's housing is expensive, and the affordability problem is most acute for 22 Poor Places, Thriving People low-income households. Although housing subsidies in countries such as Algeria, Egypt, the Islamic Republic of Iran, Jordan, Morocco, Tunisia, and the Republic of Yemen are sometimes massive, low-income house- holds may have difficulty making the investments that give them access to these subsidies. MENA's housing stock is also underused, with high vacancy rates. These factors act as a brake on rural-urban migration, thereby denying people in lagging areas a key pathway out of poverty. Land for business. Surveys of MENA entrepreneurs indicate that diffi- culties in accessing land are one of their main obstacles to competitive- ness. While MENA13 economies perform well with respect to the time required to purchase property, they are worse than any region of the world except Sub-Saharan Africa in terms of transaction costs. Land for the rural transition. Efficient agricultural land markets will be important to facilitate the shift of labor out of agriculture in the future and to increase the earnings of those who remain in farming. In MENA's agricultural land markets, however, it is often difficult for landowners to obtain and maintain registration. Therefore, the actual pattern of recognized ownership is not reflected in the official record. This lack can create problems in accessing credit or consolidating frag- mented land holdings. Planning Urban Services The integration of new arrivals into the city depends upon their ability to access public and private services. Here we illustrate a few key princi- ples for an agenda for MENA urban service development. Planning ahead, rather than reacting to in-migration. Public authorities have a tendency to delay the provision of services to new informal settle- ments because of perceptions that they are temporary or unwelcome. But these settlements instead should be seen as part of the housing solution rather than a problem. The earlier that services for emerging settlements are planned, the greater the efficiency of road, power, water, and sanita- tion networks. Urban planning can also limit MENA's vulnerability to climate change by minimizing construction in vulnerable areas and by ensuring investment in protective infrastructure. Responding to the "spatial flow" of market-led city growth. Service provision and new developments should follow people's settlement preferences rather than attempt to lead them. For example, the Islamic Republic of Iran cre- ated 18 new towns since the Revolution at a considerable distance from their parent cities with the aim of protecting farmland and preventing infor- mal settlements around major cities. These towns, however, became dor- mitory communities for salaried employees of the mother city and had no economic base of their own. Therefore, the authorities have refocused their attention on increasing the supply of housing within the metropolis itself. Overview 23 Similarly, in 1969, Egypt created distant satellite cities in the desert around Cairo intended to limit the city's growth. These cities suffered from the opposite problem as the Iranian new towns; they had significant business but few residents, and this created a pattern of reverse commut- ing. Meanwhile, the forces of agglomeration were accelerating the growth of Cairo's metropolis. By the early 1990s, therefore, the govern- ment had refocused its efforts on slum upgrading, using services to improve the quality of growth in the metropolis. In both the Iranian and the Egyptian cases, the authorities shifted their approach, using service provision to improve the quality of metro- politan agglomeration rather than trying to discourage it. Having Institutional Arrangements for Smooth Urban Growth The institutional capacity of municipalities to respond to urban growth through planning, regulation, and service provision differs greatly throughout the region. Various initiatives to strengthen those capacities are ongoing. Policy Package 2. Connect Poor Places to the Poles of Development Spillovers from agglomerations into their peripheries are one of the most powerful forces in spatial development, thereby offering the potential to reduce spatial disparities by yoking lagging regions to the growth of lead- ing regions. MENA's lagging areas have a proximity advantage: 61 percent of their population lives within three hours of a major city. How, then, can MENA connect its lagging areas to agglomeration hubs so that they maximize the benefit from spillover effects? The answer has three components: · Transport · Trade facilitation · Information and communication technologies Improvement of Spillover Transport Connections MENA's transport infrastructure is basically road-based. The exten- sive primary road network generally provides adequate intercity (that is, long-distance) connectivity. In measures for road network coverage, MENA performs well in comparison to other developing regions. 24 Poor Places, Thriving People Given that 61 percent of MENA's lagging-area populations already live within three hours of a city, MENA governments should not assume that long-distance connectivity is a major component of the lagging-areas prob- lem. The positive outlook on long-distance connectivity in MENA, how- ever, masks a problem: the predisposition toward expanding higher-class roads in MENA at the expense of other key transportation investments. MENA's broad pattern of public expenditure in transport could bene- fit from a reallocation of resources away from expansion of the domestic trunk road network toward maintaining the existing network, improving rural access, creating transborder networks, and improving traffic man- agement to reduce congestion and improve safety. Urban and peri-urban mass transit systems are critical if people from lagging areas are to bene- fit from demand for labor from nearby agglomerations. Degradation of road surfaces. The expansion of the trunk road network is taking place at the expense of regular maintenance. World Bank stud- ies identify road deterioration as a major problem in Algeria, Egypt, the Islamic Republic of Iran, Iraq, Libya, and the West Bank and Gaza. Rural connectivity deficits. MENA also suffers from a comparative neg- lect of rural accessibility. The rural index indicator for MENA, defined as the percentage of rural population with access to an all-season road, was an estimated at 60 percent in 2004 (figure 1.4), outperforming only Sub-Saharan Africa and South Asia. Recent rural road investments in Morocco and the Republic of Yemen, however, have delivered measura- ble and cost-effective impacts in lagging areas. Urban congestion. Too many private cars and poor public transporta- tion lead to congestion in MENA's cities. Motorization rates are set to FIGURE 1.4 Rural Access Index: MENA versus Comparator Regions, 2004 (percentage of rural population with access to all-season road) 100 80 percent 60 40 20 0 EAP ECA LAC MENA SA* AFR** Source: Roberts, Shyam, and Ragstogi 2006. Note: * Year 2003/04 **Year 1999­2004. AFR = Africa, EAP = East Asia and Pacific, ECA = Europe and Central Asia, LAC = Latin America and the Caribbean, MENA = Middle East and North Africa, SA = South Asia. Overview 25 increase with economic growth, but these rates lean toward private cars instead of public transportation. Mass transit in MENA consists largely of low-quality, publicly owned bus services. Unaffordable public invest- ment in light rail is often seen as the main alternative; however, other potential strategies include lifting barriers to private investment in urban buses and the development of bus rapid transport (BRT). Road safety. MENA has, by far, the developing world's most dangerous roads. The situation is particular worrisome in Morocco, Jordan, and the Islamic Republic of Iran. Low efficiency of road use. In some MENA economies, the return on road investment is being undermined by inappropriate regulation of commercial operators. Smoother Border Crossings Land borders create artificial lagging areas by restricting the movement of people and freight. Constructing transborder connections, improving logistics infrastructure, and easing border formalities, therefore, function not only as trade-promotion policies, but also as a part of national strate- gies to reduce spatial disparities. Electronic Proximity: A New Tool for the Integration of Lagging Areas Isolation from information is part of the lagging-area problem. As mobile phones and the Internet become the default communication media, those without access to them will become marginalized. Information and communication technologies (ICTs), therefore, give policy makers a new set of options for attacking spatial disparities. MENA's rural areas generally have poor access to mobile telephone and Internet services. In 2007, cellular connections in Egypt, the Islamic Republic of Iran, Lebanon, Libya, and Syria were lower than in other countries at similar levels of development (figure 1.5). MENA is the only world region in which the rate of rural mobile phone usage has not over- taken that of fixed lines. The challenge of ICT access in MENA's lagging areas has two com- ponents: the market efficiency gap and the access gap. · Closing the market efficiency gap. The market efficiency gap is the dif- ference between the current level of service penetration and what is achievable in a liberalized and stable regulatory environment. How- ever, although telecommunications liberalization in MENA is a work in progress, tariffs are generally not high (except in Lebanon and Morocco). So MENA's problem is possibly more one of . . . 26 Poor Places, Thriving People · Closing the access gap. The access gap is the difference between lead- ing and lagging areas remaining under efficient market conditions. Competitively awarded public-private partnerships are an efficient instrument for bringing ICT infrastructure and services into less profitable, low-density areas. Policy Package 3. Underpin Private Sector Interest in Nonleading Areas Although economically marginal areas usually miss out on private invest- ment because of the forces of agglomeration, there are exceptions to the rule. Tunisia's center-east region, the Tanger-TČtouan region in Morocco, and some governorates in the Islamic Republic of Iran are examples of lag- ging areas that converged with leading areas. What can governments do to facilitate this desirable process? In this section, we will begin by examining failed regional economic development policies in lagging areas around the world. We cast a simi- larly critical eye on regional economic development policies targeting lagging areas in MENA13, and we describe an alternative approach to regional economic development. Global experience has little positive to teach about the potential of big spending programs to encourage economic growth in lagging areas. FIGURE 1.5 Some MENA Economies Are Well Below the Trend in Terms of Cellular Connections 160 140 120 mobile cellular subscription per 100 people, 2007 100 ALG 80 JOR MOR TUN LIB 60 IRN EGY 40 20 SYR LEB 0 2.5 3.0 3.5 4.0 4.5 5.0 log10 GNI per capita, PPP (current international $), 2007 Source: ITU International Telecommunications Union Database. Note: ALG = Algeria; EGY = Egypt, Arab Rep.; GNI = gross national income; IRN = Iran, Islamic Rep.; JOR = Jordan; LED = Lebanon; LIB = Libya; MOR = Morocco; PPP = purchasing power parity; SYR = Syrian Arab Republic; TUN = Tunisia. Overview 27 While the traditional approach has been to focus public infrastructural investment on lagging areas, a number of studies indicate that returns to public investment are highest in leading areas, which offer the benefit of economies of agglomeration. Governments have also attempted to steer private investments toward lagging areas through financial incentives-- with typically poor results. Even where spatially targeted incentives have induced firms to invest in lagging areas, the gains are often short-lived. In MENA, the focus has been on using financial incentives to override economies of agglomeration, despite the global evidence that this approach is ineffective. In Algeria, regional economic development programs, which included investment in state-owned industries, attempted to spread eco- nomic activity beyond Algiers' sphere of influence. Yet both public and pri- vate investments remained centered on Algiers. Syria experienced similar difficulties during the 1990s in using industrial zones as incentives to encourage economic activity in lagging rural areas. Despite these efforts, economies of agglomeration prevailed, and the private sector continued to locate in major cities. Coordination Instead of Financial Incentives Some governments are now adopting a new approach to regional eco- nomic development policy. This approach accepts that the forces of agglomeration are too strong to be overridden by public money. The role of the government becomes that of facilitator rather than steerer. Table 1.3 compares the old and new paradigms of regional economic development policy. TABLE 1.3 Old and New Paradigms of Regional Economic Development Policy Old paradigm New paradigm Objectives Trying to compensate for Tapping underused potential locational disadvantages to enhance regional of lagging regions competitiveness Unit of intervention Administrative units Functional economic areas Strategies Sectoral approach Integrated approach Tools Subsidies and state aids Mix of soft and hard capital (capital stock, labor market, business environment, social capital, and networks) Actors Central government Different levels of government (with private sector) Source: OECD 2009. 28 Poor Places, Thriving People Infrastructure and education are effective only in combination. A recent analysis of the determinants of regional growth found that the infra- structural investment influences growth only when combined with human capital and innovation; infrastructure alone has little impact. If regional development is to be based on local potential, regions need a decen- tralized, bottom-up approach to policy design and implementation. Regional actors are better able to use information about local capacities and needs than are central authorities, and local actors are more likely to support projects they have helped design. Mobilizing local stakeholders may require decentralization and a structured dialogue between the different levels of government. A successful example is the EU's LEADER Com- munity Initiative, which aims to promote endogenous growth through decentralized local action groups. The public-private interface is a key source of regional productivity, growth, and innovation. Although governments cannot force agglomer- ation, they can play an important role in encouraging interchanges between regional actors and assembling the links on which agglomera- tions are built. For example, a Moroccan initiative identified four sec- tors as candidates for cluster development and fostered a continuous dialogue between public and private institutions aimed at identifying and coordinating investments and regulatory reform. Because cluster-driven growth is multisectoral, policy makers need to shift increasingly to a territorial perspective rather than a sectoral one. The "green clusters" approach for rural areas provides a good example. Realization of Agricultural Potential: Green Clusters Agriculture is still central to MENA's lagging area economies. So what kind of agricultural growth can lead development there? MENA's agricultural policy has long involved public spending on water diversion and price subsidies. Water diversion is reaching its eco- logical limit, however, and production subsidies can exacerbate inequality because the better-off residents of better-off rural areas receive a disproportionate share of the subsidy. Sector analysis sug- gests that the main potential engine of agricultural growth in MENA's lagging areas needs to be cluster development around high- value product supply chains. Therefore, we can apply the general principles of the new paradigm of regional economic development to agriculture: first, begin with a cross-sectoral territorial diagnosis that identifies latent potential; second, ensure that the full range of private stakeholders will have a leadership role, beginning in the diagnostic stage. Overview 29 Proximity for All: Public Institutions for Spatial Policies MENA governments can, therefore, beat the trade-off between agglom- eration economies and spatial inequality--if they tailor their policies to the specific characteristics of their lagging areas. The three policy pack- ages just described are designed to capitalize on those opportunities, as shown in figure 1.6. So how can the state best organize and mobilize itself to resolve spa- tial disparities? Many MENA governments have similar structures. Sector min- istries are centralized, which makes local coordination difficult, and subnational administrations have very limited resources and capacity. Global experience suggests that states can use five basic approaches to integrate the management of the various sectoral strands of spatial policy: territorial development planning, needs-based spatial alloca- tion of fiscal resources, area-based programs, deconcentration, and decentralization. Territorial development planning involves a central institution responsi- ble for monitoring and analyzing spatial development trends and feeding FIGURE 1.6 A Framework for Addressing Spatial Disparities Are spatial disparities in well-being really a significant and persistent problem? (consider spatial distribution of well-being, migration, convergence, evolution of household characteristics) YES Is the area close enough to economic density to benefit from spillovers? YES NO Package 1 Package 1 Is there private YES Package 2 Package 3 interest in locational Package 3 competitive potential? NO Package 1 Package 1 Package 2 Sources: Study Team and World Bank 2009. 30 Poor Places, Thriving People spatial orientations into the national public investment planning process. Algeria, Morocco, and Tunisia are currently implementing this approach. Needs-based spatial allocation of fiscal resources can take the form of untied grants or program funding. A formula-based approach to the spatial alloca- tion of fiscal grants is common in federal states, where the government's political credibility depends on transparent and equalizing resource trans- fers. In the absence of simple and transparent allocation rules, however, political bargaining is likely to prevail. As we have seen, there is much room for improvement when it comes to using needs- or formula-based fiscal allocation rules to support the development of lagging regions. Central governments can also steer fiscal resources toward poor areas by financing nationwide programs in which the level of activity is based on a location's need. Several MENA economies, including Morocco and Egypt, have programs of this type. Area-based programs bring the spatial dimension to policy by dedicat- ing funds and creating institutions for specific areas. The challenge lies in integrating the program with the activities of local governments with- out undermining them. Deconcentration is on the agenda in all the MENA13 economies. Key lessons for MENA's policy makers include the following: · Deconcentration requires a strong institutional champion and coordinator. · Deconcentration involves the delegation of decision making on activities, finance, and human resource management; it works best when all three are deconcentrated in parallel. · The central ministry's policy-making and oversight functions remain critical when policy execution is deconcentrated. · A clear definition for the respective mandates of elected local authori- ties and deconcentrated sector ministries becomes extremely important. Decentralization involves the transfer of resources and decision making to subnational administrations with political accountability to local citizens. It makes for a more efficient implementation of policies and enhances external accountability. A vast literature suggests that the transfer of power and resources to lower levels of government helps match resources to local priorities. The problem, however, is that fiscal decentralization increases the lagging area's influence only over its own (inadequate) resources, not over the national allocation of resources. Therefore, decentralization can widen spatial disparities because there is a risk (a) of a mismatch between fiscal responsibilities and spending mandates and (b) of the overloading of Overview 31 subnational governments' limited managerial capacity. Fiscal decentral- ization is usually a political response to spatial disparities or ethnic divi- sion, and so it is often implemented without sufficient attention to the potential negative economic consequences. Cross-country analysis reveals that decentralization in low- and middle-income countries is linked to, if anything, a faster increase in spatial disparities over time. So what is the verdict on decentralization and spatial disparities in MENA? We can sum up the evidence with three propositions: · There are good reasons to continue decentralization in MENA: to improve accountability, responsiveness, and service quality. · Decentralization will not necessarily reduce spatial disparities, and decentralizing mandates without resources and capacity will increase the disparities. · What is most important is creating transparent and predictable arrangements for equalizing intragovernmental fiscal transfers and building local governments' capacity in lagging areas. In conclusion, although most MENA governments are strongly committed to the development of lagging regions, few have asked what type of institu- tional platform is needed for a spatial approach to development policy. In Short . . . Seven Policy Axioms If we draw the previous arguments together, seven brief messages can sum up the report's policy implications for MENA. 1. Spatial development begins with clear institutional roles and processes. Spa- tial policy making is multisectoral and multi-institutional. The "process platform" for spatial development includes arrangements for monitoring and coordination, resource allocation, planning, and role definition. 2. Analysis is vital. MENA needs better spatially disaggregated statistics, especially broken down by gender. It is critical to check that spatial disparities really are a major development issue; very often they are not as important as politics portrays them. Diagnosing the lagging area will identify the right policy package(s). 3. Good development is good spatial development. First and foremost, lag- ging areas need a level playing field for development. Simply by fulfilling their existing mandates to provide a positive business environment and services to the poor, sector ministries will be mak- ing a full contribution to national spatial development objectives. Girls' education in lagging areas is the win-win-win-win priority for 32 Poor Places, Thriving People spatial development. And erasing spatial disparities in the business environment should offer a high benefit at a particularly low cost. 4. Bringing social services to lagging areas usually means (a) finding the incentives for high-quality personnel to work in underprivileged areas and for poor people to use public services and (b) tailoring delivery mechanisms to lower-density areas. Simply building more facilities will not solve the problem. 5. Commodity subsidies channel a lot of public money to leading areas. Tar- geted social protection schemes are more spatially equitable. 6. Short-distance spillover connectivity between leading and lagging areas is the transportation priority for MENA's lagging areas. Depending on a lagging area's specific geography, this priority could mean investment in rural feeder roads, road maintenance, intraurban public transportation, border logistics or peri-urban radial connec- tions, and better regulation of transport operators. It is up to MENA governments to encourage private investment and innova- tion in electronic connectivity, which offers new potential for lag- ging areas to obtain the benefits of proximity. 7. The best practice approach to local economic development emphasizes the coordination of multiple actors to facilitate private sector interest in an area's latent potential. Attempts to use financial incentives to shunt pri- vate investment toward lagging areas have a very bad track record. In short, agglomeration may be an inevitable part of growth, but gov- ernments can mitigate spatial disparities. It is a matter of policy choice. Endnote 1. This study covers the MENA Region, as follows: Algeria, Djibouti, the Arab Republic of Egypt, the Islamic Republic of Iran, Iraq, Jordan, Lebanon, Libya, Morocco, the Syrian Arab Republic, Tunisia, the West Bank and Gaza, and the Republic of Yemen. We use the acronym MENA (or MENA13 when necessary) to signify these 13 economies. The members of the Gulf Cooper- ation Council (GCC) are not covered in this report because their geographic and economic characteristics are so different from those of the rest of the region. References ITU (International Telecommunications Union) database. ITU, Geneva. Overview 33 Nelson, A. 2008. Travel Time to Major Cities: A Global Map of Accessibility. Ispara, Italy: Global Environment Monitoring Unit, Joint Research Centre of the European Commission. http://gem.jrc.ec.europa.eu/. OECD (Organisation for Economic Co-operation and Development). 2009. "Investing for Growth. Building Innovative Regions." Policy Report, Meeting of the Territorial Development Policy Committee at the Ministerial Level, OECD, Paris. ORNL (Oak Ridge National Laboratory). 2005. Landscan Global Popula- tion Database. http://www.ornl.gov/landscan. Roberts P., K. C. Shyam, and C. Ragstogi. 2006. "Rural Access Index: A Key Development Indicator." Transport Papers 10, World Bank, Washington, DC. World Bank. 2008. Islamic Republic of Iran: Spatial Patterns of Poverty and Economic Activity. Washington, DC: World Bank. ------. 2009. World Development Report 2009. Reshaping Economic Geog- raphy. Washington, DC: World Bank. World Bank Investment Climate Surveys. World Bank, Washington, DC. World Development Indicators Database. World Bank, Washington, DC. CHAPTER 2 The Political Demand: Spatial Equity without Compromising Productivity In Morocco in 1999, His Majesty King Mohammed VI begins his first Revolution Day speech by asking how rural development can stem rural-urban migration.1 In 2006, the Arab Republic of Egypt's prime minister, Ahmed Nazif, promises members of Parliament from the south that "The aim is to direct close to 40 per cent of the budget to Upper Egypt. . . . We will do this despite the fact that Upper Egypt is not where 40 percent of Egyptians live. The president is seeking to make up for the decades-long bias against the south." (Al-Ahram 2006) In the Islamic Republic of Iran, President Mahmoud Ahmadinejad states that 40 percent of the 2008 fiscal year budget will be spent in rural areas (Christian Science Monitor 2007). All Middle East and North Africa (MENA) governments state that the spatial distribution of popu- lation requires change. Nearly all have policies that explicitly aim to reduce rural-urban migration. Twelve out of 13 also have policies that specifically aim to reduce migration into urban agglomerations (United Nations 2006). This is where our report begins: with the concerns of national leaders and politicians. Leaders are concerned about high poverty rates in out- lying areas, concerned that uneven development may translate into a flood of migrants into the cities, concerned that migration may translate into urban discontent, and uneasy that spatially uneven development may be seen as a failure of economic justice or a betrayal of the social contract between the government and the governed. But MENA governments have another concern no less pressing. Average unemployment rates in MENA are more than twice those of middle-income countries worldwide. Creating jobs requires competi- tiveness and growth. So can MENA raise living standards in lagging areas without com- promising economic productivity? This report answers, "Yes." 35 36 Poor Places, Thriving People MENA's Spatial Disparities in Living Standards: Six Facts From the start we have to distinguish between two different spatial patterns: · Agglomeration, the tight concentration of businesses in certain places, and · Spatial disparities, the gaps in living standards between people in dif- ferent places. These may coincide. In fact, they often do. But they are not the same thing. The pattern of MENA's spatial disparities (table 2.1) is typical of low- and middle-income countries: living standards are highest where eco- nomic activity is densest and where connections to economic density are closest. The next chapter will offer a richer exploration of the geography of inequality. For the moment, therefore, five simple facts can usefully summarize the geography of MENA's spatial disparities: · Urban areas are, on average, better off than rural areas. · Areas close to urban centers are better off on average than isolated areas. · The same broad spatial pattern of disparities usually applies both for indicators of economic welfare, such as mean per capita consumption or the poverty rate, and for human development indicators, such as educational attainment and child mortality. · The same pattern of disparities exists within subnational divisions: provinces, cities, and even rural districts have economic hubs that are better off and peripheral areas that are worse off. · Women and men experience spatial disparities differently. One more simple fact is required to set the scene: the entire MENA region is undergoing urbanization. The average MENA country or economy2 was 65 percent rural in 1960 but 65 percent urban by 2007. But so what? All countries are full of inequalities: between the educated and the uneducated, the healthy and the sick, the lucky and the unlucky, the inheritors and those whose parents had nothing, the powerful and the powerless. Economic injustice is everywhere. What is so special about spatial disparities that policy makers care so much? TABLE 2.1 Spatial Disparities in Economic and Human Development in MENA Mean per capita consump- Mean per tion capita Access to Access to Under-5 Under-5 (capital Poverty Poverty consump- improved improved mortality mortality region/ rate rate tion (urban/ sanitation sanitation Stunting Stunting rate rate national) Year (% urban)a (% rural)a Year rural) Year (% urban) (% rural) (% urban) (% rural) (urban) (rural) Algeria -- -- 8.9 19.3 1995 -- -- 98 87 10 12 40 56 Djibouti 1.24 2002 -- -- -- 1.69 2002 70 19 32 44 95 73 Egypt, Arab Rep. 1.75 2005 10.1 26.8 2004/05 1.92 2005 85 52 16 18 39.1 56.1 Iran, Islamic Rep. 1.54 2005 0.8 7.2 2005 1.87 2005 86 78 11 22 36.8 34.6 Iraq 1.02 2007 16.1 39.3 2007 1.58 2007 98 82 19 24 41 41 Jordan 1.54 2002 12.9 18.7 2002 1.32 2002 92 76 12 13 22 27 Lebanon 1.34 2004 8 > 25 1995 -- -- 100 87 -- -- -- -- Libya -- -- -- -- -- -- -- 99 98 4 6 27.5 36.3 Morocco 1.16 1998 4.8 14.5 2007 2.17 1998 98 58 13 24 38 69 Syrian Arab Republic 1.55 2004 7.9 14.5 2004 1.52 2004 100 94 22 23 19 23 Tunisia 1.33 2000 1.6 8.3 2000 1.86 2000 96 64 4 10 -- -- West Bank and Gaza -- -- 16 15 2003 1.18 2005 100 93 11 9 29.7 29.9 Yemen, Rep. 1.70 2005 30.8 45 1998 1.67 2005 92 34 44 56 56.7 86.1 Sources: Poverty: Algeria, World Bank 1999; Egypt, World Bank 2007; the Islamic Republic of Iran, World Bank 2008a; Iraq, Shlash 2009; Jordan, World Bank 2004; Lebanon, Haddad 1996; Morocco, Haut Commis- sariat au Plan, 2007; Syria, El Laithy and Abu-Ismail 2005; the West Bank and Gaza, World Bank 2003; Republic of Yemen, World Bank 2002. Consumption: Djibouti, Egypt, the Islamic Republic of Iran, Iraq, Lebanon, Jordan, Morocco, Syria, the Republic of Yemen: national household surveys, years as specified (Egypt July­September only); Tunisia, INS 2000; West Bank and Gaza, PCBS 2006. Sanitation, stunting, and under-five mortality: Alcorn 2009. Note: -- = not available. a. Poverty lines vary between countries, so poverty rates in different countries cannot be compared from this table. 37 38 Poor Places, Thriving People Politics Makes Geography Important Spatial disparities begin to matter when politics enters the equation. Consider two countries: Country A and Country B. Both countries have the same poverty rate, 20 percent (figure 2.1). Poor people have annual per capita income of $600. Nonpoor people have twice as much. Both countries, therefore, have a GDP per capita of $1,080 per year. Both countries have the same income distribution and a Gini coefficient of 0.44. Both have the same poverty gap. Each country could design the same social program to target the poor and probably achieve approximately the same results at approximately the same cost. From the point of view of individuals' economic welfare, Country A and Country B are identical. Seen from this strictly individu- alistic viewpoint, therefore, the question of place is of zero relevance: it is literally a nonissue. From a broader sociopolitical viewpoint, however, the two countries are materially different. Whereas the poverty rate in Country A is the same in all regions, Country B has a poverty rate of 50 percent in the southeast, but only 10 percent in the other provinces. This spatial FIGURE 2.1 Spatial Distribution of Disparities a. Country A b. Country B N P N P N N N N N N N N N P N N N N N N N P N P N N N N N N N N N N N N N N P N N W E N N N N S P N P P N P N N P N N P N N N N N P P N N N P N N N N N N N N N N P N N Source: Study Team. Note: P = poor, N = nonpoor. The Political Demand: Spatial Equity without Compromising Productivity 39 clustering of low-welfare people could lead to a number of sociopolitical phenomena: · Perceived spatial injustice. The residents of Country B's southeastern province may conclude that their birth location has deprived them of opportunity, through no fault of their own effort or ability. · Perceived spatial agency of the state. It may be inferred that Country B's government and those who hold political power are in some way responsible for the lagging area status of the southeastern province. · Spatial political association. Residents of Country B's southeastern province may form geographically based political ties in order to fur- ther their perceived common grievances. Ethnic and religious divi- sions may provide a framework for political association, in which case the spatial disparities become particularly worrisome (Kanbur and Venables 2005). · Spatial social and political reorganization. Natives of Country B's south- east province may migrate north and west in order to share in the higher incomes of other regions. Internal migration might involve the reconfiguration of family and social networks; informal social protection systems; patronage systems and political allegiances; a shift in power; and the spatial redistribution of ethnic, religious, and age groupings. Any of these rearrangements of the existing order could be perceived as a source of sociopolitical stress in Country B's four provinces. As the UN-WIDER study on spatial inequality (Kanbur and Venables 2005) concluded, "Even when spatial units do not represent ethnic or other cleavages, but command the allegiance of the population as politi- cal entities, increasing disparities in group averages may lead to tensions and conflict." These sociopolitical phenomena are real. MENA politi- cians' concern about them is genuine.3 Tackling them is a political imperative, as we shall see. But tackling them in the wrong way can impose an unacceptable cost upon national development. MENA Needs Economic Agglomeration MENA needs the spatial concentration of productive activity. No coun- try has ever achieved developed-country status without urbanizing. Even the populations of great agricultural exporters such as Australia, Argentina, Chile, New Zealand, and Uruguay are more than 85 percent urbanized. 40 Poor Places, Thriving People Among the earliest thinkers to identify the relationship between agglomeration and productivity was the North African father of eco- nomics, Ibn Khaldun (1332­1406). Ibn Khaldun observed that people concentrate in high-density areas in order to achieve economic specialization and a division of labor. The larger and denser the agglom- eration, he wrote, the greater its level of prosperity (Al-Alwani 1982). Economic activity is spatially concentrated for a good reason: pro- duction is more efficient when it is agglomerated (Henderson, Shalizi, and Venables 2001; World Bank 2009a, 128­29). 1. There is a "home market effect"; firms keep transport costs low by locating close to their customers. These firms' employees then become customers themselves, attracting more businesses into the area (Krugman 1991). A firm can keep its transportation costs down by locating near other firms if it is a producer of intermediate goods. So consumers in agglomerations attract firms that employ workers who become more consumers. 2. "Localization economies" arise from having many companies in the same sector in the same place. They can learn from each other, keep in touch with the market, and share the same pool of suppliers and skills. 3. "Urbanization economies" appear when different industries gather together. They exchange technologies, information, and skills. Places grow economically by producing and diffusing knowledge (Romer 1994). These "economies of agglomeration" are the founding principles of the new economic geography, a subdiscipline of economics that explains where growth takes place. Economies of agglomeration attract investment to places where economic activity is already densest. So 54 percent of the world's economic activity is concentrated in countries with 10 percent of its area (Henderson, Shalizi, and Ven- ables 2001). Within countries, economic activity is squeezed into the major cities. Ten metropolitan areas in Mexico account for a third of the country's population and 62 percent of its output. Vietnam's cities have 30 percent of the nation's population but produce 70 percent of its output (World Bank 2009a). Shanghai has only 1.2 percent of China's population but produces 12.5 percent of its economic value (UN Habitat 2004). Seven centuries later, the current spatial distribution of economic activity in MENA confirms Ibn Khaldun's predictions. Map 2.1 shows the agglomeration of economic activity in Libya and Iraq, estimated by gridding the distribution of employment by sector. Tripoli and Baghdad are the prime foci of production. They dwarf the secondary poles of The Political Demand: Spatial Equity without Compromising Productivity 41 MAP 2.1 Spatially Concentrated Economic Activity in Libya (left) and Iraq (right) Spatially concentrated economic activity Libya Iraq Source: http://gecon.yale.edu/data.php. Accessed August 18, 2009. Methodology described in Nordhaus 2006. Benghazi, Basra, and Kirkuk, as investment and labor are drawn into the metropolis by the economies of agglomeration. The evidence from MENA confirms that investment and production are indeed more efficient in areas of dense economic activity. In Egypt (World Bank 2009c) researchers broke down the determi- nants of labor productivity in industrial firms. They looked at the companies' sector, years of operation, size, and so on. Controlling for all these variables, they found that a firm in Upper Egypt, Egypt's lag- ging area far from Cairo, had labor productivity 44 percent lower than elsewhere. The Tunisian government collects investment data and household consumption data by governorate. It is, therefore, possible to see how much investment it took in each governorate to generate the same increase in household consumption. The calculations confirm the power of agglomeration economies. It takes much more investment to support a given value of consumption growth in the lagging areas of the south and the northwest than it does in the leading areas of Tunis, the north- east and the center-east (table 2.2). The Tunisian figures may understate the spatial disparities between the rates of return on investment in different regions. First, firms some- times register their investment at the location of their head office, even if the actual physical investment takes place elsewhere. Investment in the Tunis area may, therefore, be overstated. Second, remittances usually flow from rich areas to poor areas, so spatial disparities in income are bigger than disparities in consumption. But conversely, the ratios do not capture the nonconsumption benefits of investment in social services. 42 Poor Places, Thriving People TABLE 2.2 Investment in the Poorest Areas of Tunisia Gives the Poorest Returns (in millions TD, 2000 prices) Increase in annual Public and private Investment needed to raise consumption, 1985­2000 investment, 1986­2000 consumption by 1 m TD/year Tunis and Northeast 2,018 23,594 11.6 Northwest 409 6,455 15.8 Center-west 395 5,316 13.5 Center-east 1,843 15,591 8.4 South 483 9,225 19.2 Sources: INS 2000;. Enquęte nationale sur le budget, la consommation et le niveau de vie des ménages [National survey of household budgets, consumption, and living standards], Commissariat Général de Développement Régional (CGDR). Therefore, the location of investment in Tunisia can contribute to the sociopolitical objective of raising living standards in the south, north- west, and center-west lagging areas on the one hand, or the economic objective of maximizing growth nationwide on the other. Figure 2.2 shows the hypothetical range of outcomes if investment were to be redis- tributed between leading and lagging areas while earning the same aver- age return by region in terms of household consumption increases that it did in 1985­2000. The cost of increasing consumption in the lagging regions by 1 million Tunisian dinars (TD) per year is a reduction in nationwide consumption of TD 610,000 per year. Morocco publishes industrial production and investment figures by province and prefecture, which allows one to calculate the how much investment it takes to produce a dirham of additional production in each region. Figure 2.3 illustrates the agglomeration economies: regions with more industry have tended to achieve greater increases in production rel- ative to the cost of investment. This finding seems to confirm that there is a trade-off between investing to achieve the political goal of shifting production to lagging areas and investing for national growth. Figure 2.4 illustrates the hypothetical trade-off between the objective of increasing the geographical spread of industrial activity in Morocco and promoting national growth. It takes less investment to generate Moroccan dirham (MD) 1 million of industrial growth in Morocco's three most industrial- ized provinces--Grand Casablanca, Doukala-Abda, and Tanger- Tétouan--than it does in the other eleven. Applying the average provin- cial returns on investment for 1996­2005, it seems that the same investment could either generate a MD 1 million increase in industrial production in the 11 nonleading provinces or a MD 1.2 million increase in national industrial production, depending on where it takes place. A measure of the power of economies of agglomeration is that they are the main factor determining where firms locate across the developing The Political Demand: Spatial Equity without Compromising Productivity 43 FIGURE 2.2 Hypothetical Range of Outcomes between National Consumption Growth and Consumption Growth in Tunisia's Poorer Areas, 1985­2000 Increase in consumption 2,000 actual 1985­2000 Tunisian dinars, millions, per year, observation 1,500 poor areas 1,000 500 0 4,800 5,000 5,200 5,400 TD, millions per year, nationwide Source: Table 2.2. Note: The poor areas are the northwest, center-west, and south. FIGURE 2.3 In Morocco, Provinces with Large Industrial Sectors Provide the Highest Returns to Industrial Investment 1.2 Tanger-Tétouan industrial production increase 1996­2005 1.0 relative to 1996­2005 investment 0.8 Grand Casablanca 0.6 0.4 0.2 0 100 1,000 10,000 100,000 ­0.2 ­0.4 1996 industrial production by region (millions of DH, 2000 prices) Source: Annuaires Statistiques du Maroc (Statistical Directory of Morocco), 1998­2007. Note: All figures converted to 2000 prices. world (Deichmann, Lall, Redding, and Venables 2008). Tehran province's density of economic activity (value added per square kilometer) is six times higher than that of the next-densest province (World Bank 2008a). In Morocco, industrial employment growth is focused on Casablanca and then on spillovers from that metropolis into adjoining areas (Catin, 44 Poor Places, Thriving People FIGURE 2.4 Consequences of Splitting MD 1 Billion Industrial Investments between Leading Three and Lagging Eleven Provinces in Morocco, 1996­2005 250 increase in annual industrial production 30% invested in in the 11 provinces with least industry, 1996­2005 (millions of MAD/year) leading provinces 200 150 100 70% invested in 50 leading provinces 0 0 100 200 300 400 500 600 increase in annual national industrial production, 1996­2005 (millions of MD/year) Source: Annuaires Statistiques du Maroc [Statistical Directory of Morocco], 1998­2007. Note: All figures converted to 2000 prices. Hanchane, and Kamal 2007). In Tunisia, agglomeration is one of the most easily identifiable determinants of a region's growth (Karray and Driss 2008). However, there are exceptions to the general rule that economic activ- ity becomes increasingly concentrated. Some countries have "break- away" regions that pulled away from the peloton of middle-ranking regions to achieve accelerated growth. In the Islamic Republic of Iran, it was the southwestern ostans of Bushehr and Ilam that broke out. In Tunisia, it was the center-east, and in Morocco it was Tanger-Tétouan. So, while economies of agglomeration may be primordial in determining spatial competitiveness, we must also ask what conditions make for a "breakaway region" and what role governments can play in facilitating breakaways. We will explore this question in later sections. Nonetheless, breakaway areas are the exception, not the general rule. It is a statement of fact that agglomeration is a major source of produc- tivity and economic growth in MENA and elsewhere. Spatial Disparities Create Political Stresses In MENA as elsewhere, spatial disparities are a hot issue because they put pressure on governments. The Political Demand: Spatial Equity without Compromising Productivity 45 The regional disparities that developing countries inherited at inde- pendence were a challenge to postindependence nation building. Tackling inherited spatial inequalities was a crucial objective for the state-building agenda. By highlighting the developmental divides within national boundaries, lagging areas encouraged the preservation of a traditional asabiyyah--a form of group identity and solidarity promoting regionally based agendas and hampering the development of fully integrated national communities (Hudson 1997; Owen 1992, 67­68; Esman and Rabinovitch 1988). (See box 2.1.) Moreover, by fueling migration, spatial inequalities bring political risks to the better-off areas. MENA has experienced rural-urban migration, BOX 2.1 Ibn Khaldun and the Political Economy of Spatial Disparities Ibn Khaldun (1332­1406), the North African polymath, is considered one of the fathers of modern social science and economics. His career took him to many of the great cities of the Arab world: Granada, Fčs, Tunis, Cairo, Alexandria, and Damascus. His master- piece, The Muqaddimah, elaborates a theory of the political economy of the urban-rural divide (Omran 1980). Ibn Khaldun observed that the stability and prosperity of society depended on the existence of asabiyyah, which has been translated as "social cohesion," "social capital," or "group solidarity." Asabiyyah is strong among rural populations and provides the foun- dation for a stable society within which urbanization, economic specialization, and economies of agglomeration can be constructed. Ibn Khaldun argues, however, that the economic growth resulting from agglomera- tion carries the seeds of political tension. Increasing disparities between cities and rural areas lead to corruption, conspicuous consumption, and a breakdown of asabiyyah and of the legitimacy of the state in the eyes of the governed. These in turn lead to political instability and maybe a change of government. Although Ibn Khaldun speaks to us from six centuries ago, his observations about how political stability depends upon the perceived fairness of the relationship between rich places and poor places are echoed in the concerns of MENA's politicians today. However, MENA's twenty-first-century politicians have policy instruments that fourteenth-century rulers would not have dreamed of: public schools and health centers, targeted social protection schemes, planning capacity, and infrastructure budgets. Mod- ern governments can deploy these services to reinforce asabiyyah--social cohesion--at the same time that market forces are building agglomeration. So perhaps if Ibn Khaldun were alive today, he might be more optimistic. 46 Poor Places, Thriving People urban unemployment, and degrading living conditions in slums and shantytowns, and finally political radicalization in urban areas. Spatial disparities are, therefore, more than economics; seen through the lens of MENA's recent history, they are also a challenge to the equilibrium of the state. Therefore, the question is how to sustain this spatial-social equilibrium without sacrificing the economic efficiency of agglomeration. Women and Men Experience Spatial Inequalities Differently In MENA, as in other parts of the world, women's lives are conditioned by society's expectations and norms. The traditional gender paradigm in MENA, combined with gender-specific legal provisions, determines to a great extent women's access to and interaction with the state and the public sphere. Key elements of the traditional gender paradigm in MENA include centrality of the family, establishment of the man as the principal breadwinner and as head of household, the "code of modesty," and unequal decision making and economic power in the private sphere (World Bank 2008b, 94, 132). Thus, living in a lagging area affects women differently from men. Their experience of distance differs according to their mobility. Their experience of disparities in services differs according to their ability to interact with officialdom. Their experience of migration differs accord- ing to their access to the job market and social networks. This study, therefore, commissioned field surveys in Egypt, Saudi Arabia, and Tunisia to investigate how spatial inequality and gender roles interact (Esber 2008). They confirmed that spatial inequality sometimes can combine with gender inequality to create a double perception of mar- ginalization. The main findings from the surveys are summarized below. Women are particularly affected by spatial disparities in access to services. Gender disparities in access to public, social, and administrative services tend to be higher in lagging areas. This finding is borne out by MENA's data on education access and outcomes; in Morocco, the literacy gap between men and women is 5 percentage points in urban areas and 35 percent in rural areas. In Tunisia, the gender literacy gap in 2004 was 5 percentage points in Ettadhamoun, a low-income suburb of Tunis, and was 12 percentage points in Kasserine, a provincial town in a lagging gov- ernorate. Two-thirds of women in the lagging area town of Kasserine cited improvement in relations with public administrations and govern- ment authorities as a priority for improvement in their living conditions, a higher score than any other issue such as education or access to credit. The Political Demand: Spatial Equity without Compromising Productivity 47 When men from lagging areas migrate in search of work, the female members of the family who remain sometimes have difficulties dealing with male officials (see box 2.2 for an example). This issue is particularly important for the design of MENA's agricultural extension services. Many rural women assume day-to-day responsibility for the household's farming activities when men of working age migrate in search of employ- ment, but these women are socially and physically constrained from obtaining advice and services from male extension agents. Women are especially affected by the lack of appropriate transport options in peripheral areas. Social and physical constraints on women's use of trans- portation are, of course, particularly important in low-density areas. Women in lagging areas of Saudi Arabia have particular difficulty in accessing all levels of education and services, in addition to limited social interactions, because of their partial mobility. The transportation dispar- ity exists but is less pronounced in Egypt and Tunisia. In the Republic of Yemen, 56 percent of rural women, but only 25 percent of men, reported that transportation options were unresponsive to their needs (World Bank forthcoming). Rural girls have to pay 50 percent more than boys for transportation to school; boys can hitch a lift, but girls have to pay for a private car for themselves and their muhrams (male companions). The costs and benefits of migration are different for women. Displacement from home results in physiological, psychological, and social stress (Shami 1993), but men and women may have very different experiences of the pros and cons of moving to the big city. This finding was clear from our survey of residents of Manshiyet Nasser, a low-income area of Cairo, who were asked to compare life in the capital with Qena, a lagging BOX 2.2 A Migrant's Wife Has Difficulties with the Local Authorities "My husband worked in the Sahel for a few years. At the time, I was entitled to receive social security from the authorities but often didn't receive it. The shaykh [local author- ity] did not want to give it to us, because I sent him away from my home. I was a woman alone with my children, and if the neighbors saw that I received him at home, they would say bad things about my conduct, and I would have problems when my husband returned. Unfortunately, the shaykh considered it his right to go to people's homes; it is not the case for me [to permit it in order] to have social security. I should have made a complaint against him to his superiors, and I would have won my case." Source: Khadija, a 40-year-old woman from Kasserine, a lagging area of Tunisia. 48 Poor Places, Thriving People area town in Upper Egypt. The economic motive for migrating to Cairo appeared more positive to men than to women: 82 percent of men but only 36 percent of women saw job opportunities as an advantage of living in Cairo compared with Qena. For women, the main advantage of living in Cairo was the ability to build strong social networks and was cited as an advantage of Cairo by 55 percent of women but only 36 percent of men. When asked about the advantages of Qena com- pared with Cairo, men and women again had very different view- points. For men, Qena's primary positive points were transportation and the ease of obtaining official papers. For women, Qena's main attractions were the low crime rate and the healthier environment. In Tunisia, however, men and women both saw the advantage of the capital city over a lagging area town (Kasserine) in the same light: income and employment. And both men and women saw the lagging area town having the same advantage over Tunis: greater family solidarity, respect for tradition, and freedom from crime. More than 80 percent of women said that they would support their son's or daughter's migration. Women were particularly pragmatic in their attitude that economic factors justi- fied migration. A majority of female respondents quoted the saying, "Your country is where you earn your bread." One cannot be certain why the Tunisian women surveyed seemed much less worried by the negative social effects of migration than their counterparts in Egypt; however, it may be that the comparative weakness of Tunisia's social restrictions on women makes migration less of a chal- lenge when it comes to mobility, to the labor market, and to establishing social networks. Therefore, the social roles apportioned to women in MENA mean that they experience the lagging region syndrome of distance and isolation even more acutely than men. The gender dimension of spatial disparities may not be as visible as the political dimension, but it is, nonetheless, an element to be taken into account when policy makers attempt to balance the sociopolitical with the economic. Spatial Equity with Spatial Efficiency So we begin with two firm facts. First, investment and employment are more productive when they take place in areas of economic agglomera- tion in leading areas. Second, the strongest political demand from gov- ernments and the governed is for improvements in living standards in lagging areas. The purpose of this report is to suggest how MENA policy makers can turn the tension between spatial politics and spatial economics into a The Political Demand: Spatial Equity without Compromising Productivity 49 win-win: a policy package that satisfies both the political argument for equity and the economic argument for efficiency. · First, by analyzing MENA's lagging areas, we will see that MENA economies can get greater improvements in living standards in lagging areas than they think for any given economic cost. The reason, as our detailed diagnosis of MENA's lagging areas will show, is that many of the factors holding lagging areas back are social and institutional rather than economic. This means that MENA governments can address the root causes of spatial disparities through administrative reforms and human development interventions, such as education, and health and social protection programs, rather than relying on the doubtful and costly strategy of luring economic activity to lagging areas. · Second, we will show how governments can maximize the benefit-cost ratio of policies to reduce spatial disparities by tailoring them to different lagging areas. Later sections of this report will explore, in turn, the potential of human development, connectivity, and targeted regional economic development to narrow MENA's spatial disparities. · Finally, the report will examine how some MENA economies are attempting to reverse the historical legacy of centralization and build a new institutional framework for managing territoriality. We will then discuss whether such reforms will give MENA politicians more room to maneuver in the future, setting up a new and more propi- tious trade-off between spatial economics and spatial politics. Matching Policy Packages to Lagging Areas: A Three-Way Framework It is a cliché that development policies must fit the local context, and this is especially true for spatial development. Lagging areas (see box 2.3) come in many different varieties; an arid rangeland is not a peri-urban slum. Which policies, then, are appropriate in which places? With some adjustment, the framework of the 2009 World Development Report, Reshaping Economic Geography (World Bank 2009a), provides a starting point for matching policy packages to places. Drawing on the World Development Report, we propose three pol- icy packages for MENA: · Policy Package 1. Level the playing field and the opportunity for human development in lagging areas. · Policy Package 2. Connect lagging areas to the poles of development. 50 Poor Places, Thriving People BOX 2.3 Mind Your Language! This report will use the word "lagging" nearly 500 times, so we need to agree what it means. Our definition of a "lagging area" combines the spatial distribution of economic activ- ity and living standards. It uses the phrase "lagging area" to denote a place with low pro- duction per capita and low living standards compared with other places in the same country. If a subnational measure of GDP were available for a lagging area, it would be significantly lower than the national average. Welfare measures such as per capita con- sumption and human development indicators will also be low. A lagging area can be big or small. In the context of a country, it might be a region. In the context of a region, it might be a district or an agro-ecological zone. In the con- text of a district, it might be a village. The phrase "spatial disparities," however, refers solely to the gaps in living standards between lagging areas and the rest of the country. It says nothing about the distribution of economic activity. This terminology underlies the critical distinction illustrated in figure 2.5. Overcom- ing spatial disparities is about raising people's living standards in lagging areas, but it does not necessarily involve a change in the spatial distribution of economic activity. · Policy Package 3. Underpin private sector interest in nonleading areas. The three packages are not mutually exclusive; they can overlay each other. Policy Package 1 applies everywhere. Governments can choose to add packages 2 or 3 or both if the lagging area meets specific criteria. Policy Package 1. Provide a Level Playing Field and the Opportunity for Human Development in Lagging Areas Breaking the link between the spatial concentration of production and spatial disparities. Just because economic activity is attracted to agglomerations (density) does not mean that people living in peripheral areas must suf- fer from growing spatial disparities in well-being. Unfortunately, it is very difficult to analyze the spatial distribution of economic activity in MENA, as authentic subnational GDP estimates are scarce.4 However, the experience of developed countries shows that economic agglom- eration does not necessarily mean greater spatial disparities in living standards (World Bank 2009a). The Political Demand: Spatial Equity without Compromising Productivity 51 In France, for example, the concentration of economic activity around the Paris metropolitan area grew more and more pronounced between 1982 and 2002. At the same time, however, the distribution of per capita income became more and more equal (Martin 2005), as the taxation and benefits systems transferred the value of metropolitan areas' production to poorer people in the provinces. In the United Kingdom (UK), conversely, both economic activity and people's incomes became more concentrated during the 1990s (Duranton and Monastiri- otis 2002). The reason was a growing gap in education levels between the leading southeast and the poorer parts of the UK. In the United States during the twentieth century, the gap in posttax incomes between the poorer and the richer states narrowed whenever the tax rate increased (World Bank 2009a) and widened when it fell again. Long-term conver- gence between richer and poorer states in the United States was achieved by investments in education (Lall and Yilmaz 2000). A study of Eastern Europe and Central Asia found that the correlation between a region's GDP and its level of poverty was very strong in some countries and weak in others (Dillinger 2007). In short, the spatial distribution of production and spatial disparities in welfare are different outcomes. The decisions of governments deter- mine how far agglomeration translates into spatial disparities in living standards (figure 2.5). The implication is that policy makers who are concerned about spa- tial disparities in people's well-being should focus first and foremost on FIGURE 2.5 Agglomeration and Spatial Disparities in Well-Being: Two Separate Outcomes outcome 1 spatial market forces distribution of economic activity outcome 2 spatial policy distribution of environment people's well-being Source: Authors. 52 Poor Places, Thriving People developing the people who live in poor places. Such policies could include improving educational services, arranging health services and social protection, and making it easier to take advantage of migration opportunities. The core principle of Policy Package 1 is that governments should try to give poor places the same institutional environment and opportunities for human development as those in rich places--in other words, a "level playing field for development." Leveling the playing field involves understanding how the historical accretion of centralizing policies has created implicit spatial biases in policy, and then setting out to correct them. As we shall see in later sec- tions, a legacy of MENA's colonial and postcolonial history was often an implicit prometropolitan bias in the application of policies. The institu- tional environment; patterns of government expenditure; pricing poli- cies; and education, health, and social protection systems were most favorable to metropolitan, urban, or leading areas. For example, energy subsidies, which obviously favor urban and metropolitan areas the most, averaged 7.1 percent of GDP for all MENA countries, even before the fuel price increases of 2008. The policy package of investing in people and leveling the playing field for development can be applied to any area with lagging development indi- cators. Clearly, it does not depend upon the lagging area having specific geographic characteristics, such as proximity to agglomeration, or on a specific source of economic potential, such as natural resources or access to markets. It is based upon the universally applicable principles that insti- tutions and human development are critical to spatial development and that nondiscriminatory treatment of different target groups is nondistorting. The two policy packages that follow, however, are applicable only to areas that meet defined criteria. Policy Package 2. Connect Lagging Areas to the Poles of Development Economic growth spreads out and spills over from cities to hinterlands. And if a lagging area is close enough to a leading area, linking the two can help prosperity to spill over from the latter to the former. Being con- nected to markets will increase consumer welfare in the lagging area, as residents face lower prices because of market integration. Workers and producers will also realize higher value added as they benefit from better access to jobs and markets. Therefore, connectivity infrastructure is always a backbone of spatial development. In Vietnam, for example, a highway from Hanoi to Hai Phong cre- ated a corridor of growth and poverty-reduction penetrating deep into The Political Demand: Spatial Equity without Compromising Productivity 53 a lagging area (ADB, JBIC, and World Bank 2005). Rural road invest- ments raised GDP by 30 percent between 1993 and 1998, as well as expanding school enrollment and improving health service utilization. In India, investment in rural roads was one of the most effective forms of public expenditure in terms of promoting growth and poverty reduc- tion at the subnational level (Fan, Hazell, and Thorat 1999). MENA has a special advantage over other world regions when it comes to using connective infrastructure to develop lagging areas. MENA's settle- ment patterns are compact. Lagging areas are often close to leading areas; only 39 percent of its lagging areas' populations live more than three hours from a major city. The benefits of linking lagging areas to nearby hubs are proven. A study by the World Bank found that improved rural feeder roads in Morocco generated substantial positive effects compared with areas without new roads for transport, agriculture, health, education, and gen- der. Farmers' production, productivity, and output values increased, and they shifted toward higher-value crops. And enrollment in primary edu- cation increased much faster, particularly for girls (Levy 2004). However, connectivity investments do have a side effect: cheaper trans- port often means more economic agglomeration. As transportation costs between the leading and the lagging area fall, producers in the lagging area can be undercut by competition from the leading area, which has lower costs because of its economies of agglomeration (Venables 2005). Looking at the arrival of new interstate highways in rural America, Chandra (2000) found that rural-based companies that had been trading nationally expanded after the arrival of the roads, presumably because they could now access their distant markets more cheaply. Companies that had been trading within the local region, conversely, tended to shrink as their home market was left more exposed to competition from other areas. Italy's intervento straordinario, which mobilized billions of dollars per year to build roads and railways for Italy's lagging southern region, is now credited with having led to the out-migration of two million workers from the target area in the 1950s (World Bank 2009a). The literature cited earlier suggests that effects of short-distance con- nections (such as rural feeder roads) are more likely to be unambiguously positive for the lagging area than the effects of trunk roads. This finding may be because the artisans and farmers in the small villages that benefit from feeder roads are usually not in competition with larger towns. The village benefits from cheaper consumer goods and farm inputs, as well as better access to agricultural markets and services and the possibility of commuting to larger settlements. The availability of feeder roads can be a major influence on farmers' costs to market and decision to invest in cash crops (Lall, Wang, and Munthali 2009); however, it should be noted 54 Poor Places, Thriving People that it is difficult to measure the impact of rural roads programs ex-post, because it is hard to know whether a new road in a place led to its development or whether a place's high level of development made the government decide that it needed a new road (Van de Walle 2009). Transportation is no longer the only critical mode of connectivity. Improving information and communication technology (ICT) connec- tivity is now of growing importance in linking lagging areas to national economic growth. Not only can cheap and efficient communications improve access to market information, but also they can support the delivery of education. They help people from poorer areas to migrate and integrate into leading areas by acquiring contacts and "cultural markers" from more advanced parts of the country. This report will, therefore, assess the state of connectivity in MENA and will identify thematic priorities for lagging areas in the transport and ICT sectors. Policy Package 3. Underpin Private Sector Interest in Nonleading Areas We suggested earlier that the people-based approach is the most appli- cable solution for all lagging areas. Sometimes, however, the people-based approach alone is not enough. Sometimes there are places whose populations need new economic opportunities nearby to sustain their livelihoods. Governments will look to place-based economic development interventions under the following three circumstances: · There is no prospect of growth spilling over from leading areas fast enough. · There is no prospect of migration solving the lagging area problem fast enough. · There is genuine private investor interest in the lagging area, requir- ing only that government puts the necessary facilitating institutions and infrastructure in place. The critical point is that governments should underpin private sector interest, not seek to steer it toward lagging areas. The ability of the population to redistribute itself from lagging to leading areas is a key criterion for deciding whether place-based approaches are appropriate (World Bank 2009a). Clearly, the easier it is for people to move to leading areas with high economic potential, the less sense it makes to allocate productive investment to lagging The Political Demand: Spatial Equity without Compromising Productivity 55 areas with low economic potential. In this report we will see that most of MENA's populations are urbanizing but not redistributing from lag- ging to leading provinces. MENA's policy makers, therefore, cannot always rely upon metropolitan growth and migration toward leading provinces to take care of underemployment in lagging provinces. This finding gives some weight to the case for place-based economic develop- ment policies, perhaps targeting lagging areas' secondary agglomerations. However, place-based regional economic development policies that aim to secure spatially balanced growth have earned a bad reputation, and rightly so. The problem is that governments have usually tried to use financial incentives to create investor interest where it did not really exist. In fact, the forces of economic agglomeration are so strong and so persistent that governments' attempts to combat them with incentives to invest in lagging areas have usually been overwhelmed. Even the great- est catastrophes, such as the near-destruction of German and Japanese cities by Allied bombs at the end of World War II, seem unable to re- direct the spatial distribution of economic activity in the long term (Davis and Weinstein 2002; Brakman, Garretsen, and Schramm 2004). The history of policies aimed to stimulate investment in lagging areas is, therefore, littered with costly failures, as this report will describe in detail. When nonleading areas move to the front, it is not because government has deployed subsidies and incentives to influence firms' location. It is because the market has responded to potential locational competitiveness, whether it be human resources (Ireland, southern United States), natural resources, or market access (Tanger-Tétouan and center-east Tunisia). This finding, therefore, means a new paradigm for the state's role in promoting regional development. The government's role is no longer to use incentives to steer investment toward lagging areas by distorting market forces; it is to help ensure that the private sector's market-led deci- sions are underpinned by the necessary infrastructure, information, net- works, and regulations. The key is not incentivization but coordination. However, the corollary of the new paradigm of regional economic development is that it just cannot apply to all lagging areas. It needs pri- vate investor interest in a lagging area's unrealized commercial potential. So it is likeliest to succeed in middle-ranking regions or in areas with a strong skills or natural resource base. Conversely, it is least likely to work in very poor regions without such assets. This report will examine in more detail the lessons learned from global practice on this "new paradigm" for regional economic develop- ment. It will examine MENA experience and derive principles for MENA policy makers. 56 Poor Places, Thriving People How to Put It All Together: Three Policy Approaches So we have three policy packages, each applicable to a different set of lag- ging areas (table 2.3 and figure 2.6). First, all lagging areas need fair access to decent institutions and human development. Second, lagging areas near to economic poles can gain from improved spillover connec- tivity. Third, places with private sector interest in unrealized economic potential could benefit from the government's coordinating the under- pinning of private investment with the necessary infrastructure, infor- mation, networks, and regulations. Figure 2.6 restates this framework and also serves as a map for the rest of this report. The process begins by checking the real size and tra- jectory of spatial disparities, because the spatial dimension of inequality is sometimes not as important at it appears. But if policy makers deem that the spatial disparities problem is serious enough to warrant inter- vention, the next stage is to identify the characteristics of the lagging areas in question. We will discuss these diagnostic stages in detail. The following three sections explore the three policy packages themselves: level the playing field and the opportunity for human development in lagging areas, connect lagging areas to the poles of development, and underpin private sector interest in nonleading areas' economic poten- tial. The last section concludes by exploring how MENA's institutional structures can become more supportive of such territorial approaches. TABLE 2.3 Matching Policy Packages to Lagging Areas World Development Policy packages Report 2009 terminology Why? Where? Package 1. Level the "Institutions" Spatial disparities in well- Applicable to all lagging playing field and the ("spatially blind policies") being can narrow even areas opportunity for human while production is development in agglomerating, so long as lagging areas policies are appropriate. MENA has a historical legacy of policies' widening disparities. Package 2. Connect "Infrastructure" ("spatially For growth to spill over Lagging areas close lagging areas to the connective policies") from economic poles into enough to density poles of development lagging areas, they have to benefit to be connected. Package 3. Underpin "Incentives" Sometimes nonleading Areas where the private private sector interest in ("spatially focused areas do move ahead sector sees unrealized nonleading areas policies") to leading-area status growth potential (e.g., Tunisia's center-east, Morocco's Tanger-Tétouan) Sources: Study Team and World Bank 2009a. The Political Demand: Spatial Equity without Compromising Productivity 57 FIGURE 2.6 Matching Policy Packages to Lagging Areas Are spatial disparities in well-being really a significant and persistent problem? (Consider spatial distribution of well-being, migration, convergence, and evolution of household characteristics.) YES Is the area close enough to economic density to benefit from spillovers? YES NO Package 1 Package 1 Is there private YES Package 2 Package 3 interest in locational Package 3 competitive potential? NO Package 1 Package 1 Package 2 Sources: Study Team and World Bank 2009a. What This Report Covers and What It Doesn't From Examples to Recommendations Spatial analysis needs spatially disaggregated data, but spatially disaggre- gated data in MENA are scarce. Household consumption surveys are generally representative at the province-governorate level, but poverty maps are usually required to access finer levels of disaggregation. Other subnational data in MENA are generally a by-product of sectoral analy- ses. Rural-urban breakdowns are sometimes available and provincial breakdowns are rarer. Where subnational data do exist, the use of admin- istrative boundaries for demarcation makes cross-country comparison a hazardous exercise. The approach of this report has, therefore, been to illustrate general principles from specific examples. In the chapters that follow, key spatial development issues are exemplified by the available data. The message to the policy maker is always the concept and the analytical method rather than a locality-specific conclusion. In addition, this report contains little spatial econometrics. We could have spent a lot of time, money, and effort analyzing the forces of spatial agglomeration in MENA, but we would have concluded only what MENA's policy makers know well already: that economic activity is drawn inexorably toward existing agglomerations. We could have attempted to quantify the returns to public spending on spatial development, but with 58 Poor Places, Thriving People such analyses showing inconclusive results even in countries with abundant data (Glaeser and Gottlieb 2009), there was little hope that they would have been more productive in MENA. Indeed, even if we had had more data, there was the risk that deeper analysis would merely have led to the truisms that every locality's experience is unique and that yet more disag- gregated analysis was needed. This report, therefore, emphasizes operational implications. We hope to illustrate some simple questions that policy makers can pose in order to anatomize the spatial disparities situation in their country. Every section concludes with an "Ideas for Policy Makers" box--a handful of suggestions for taking strategy toward implementation. Country Coverage: The MENA13 This report does not cover Israel, Malta, or Middle Eastern and North African economies that are members of the Gulf Cooperation Council (GCC). The reason is that the GCC states have very different geo- graphical and economic characteristics from the rest of the region: their economic geography results from the absence of widespread seden- tary agriculture because of their aridity and their abundance of mineral resources. So, while the GCC states do indeed face a number of critical spatial development issues, they would require a separate and distinctive analysis. The study, therefore, covers the following Middle Eastern and North African economies: Algeria, Djibouti, Egypt, the Islamic Repub- lic of Iran, Iraq, Jordan, Lebanon, Libya, Morocco, the Syrian Arab Republic, Tunisia, the West Bank and Gaza, and the Republic of Yemen. The report uses the abbreviation "MENA" to signify these 13 economies. Sometimes, for example, to distinguish them from the broader set of MENA economies, it will refer to them as the "MENA13." When a secondary source refers to "MENA" as the whole region, including the GCC, we will say so. The report uses national definitions of "rural" and "urban" areas. Unless otherwise stated, when the report states an indicator for "lagging areas," this refers to the worst-off 40 percent of governorates or provinces. Endnotes 1. Speech of H. M. King Mohammed VI of Morocco on the occasion of the forty-sixth anniversary of the revolution. 2. Algeria, Djibouti, Egypt, the Islamic Republic of Iran, Jordan, Lebanon, Libya, Morocco, the Syrian Arab Republic, Tunisia, the West Bank and Gaza, and the Republic of Yemen. 3. In an unscientific poll, the report team surveyed 27 MENA policy makers in April 2009 concerning spatial development policies in their countries. The Political Demand: Spatial Equity without Compromising Productivity 59 Seventy-eight percent said that government had favored the metropolis at the expense of peripheral areas in the past, though half of them thought that this bias had been corrected. Two-thirds said that spatial policy was primarily based on political or social considerations; one-third said it was based on efficiency considerations. 4. Egypt's National Human Development Reports (HDR) publish estimates of GDP per capita by governorate. These figures are obtained by applying a con- stant multiplier to consumption data. GECON's publicly available datasets assume that a given sector's labor productivity is equal in all locations within the same country. Morocco issued its first provincial GDP estimates after this report went to press. References ADB (Asian Development Bank), JBIC (Japanese Bureau for Interna- tional Cooperation), and World Bank. 2005. Connecting East Asia: A New Framework for Infrastructure. Washington, DC: World Bank. Al-Ahram. 2006. "Focus on Upper Egypt." Al-Ahram Weekly Online. March 23­29. Al-Alwani, M. 1982. "The Arabs and the Science of Population Distri- bution." Population Bulletin of ECWA 22: 159­77. Alcorn, E. 2009. "Regional Overview of Spatial Disparities in Health Care Access and Health Outcomes." Unpublished. Brakman, S., H. Garretsen, and M. Schramm. 2004. "The Strategic Bombing of German Cities during World War II and Its Impact on City Growth." Journal of Economic Geography 4 (2): 201­18. Catin, M., S. Hanchane, and A. Kamal. 2007. "Structure Industrielle, Externalités, Dynamique et Croissance Locale au Maroc." Région et Développement 25: 1­18. Chandra, A. 2000. "Does Public Highway Infrastructure Affect Economic Activity? Evidence from the Rural Interstate Highway System." Regional Science and Urban Economics 30: 457­90. Christian Science Monitor. 2007. "Ahmadinejad: Rock Star in Rural Iran." December 7. Davis, D., and D. Weinstein. 2002. "Bones, Bombs, and Break Points: The Geography of Economic Activity." American Economic Review 92 (5): 1269­89. Deichmann, U., S. Lall, S. Redding, and A. Venables. 2008. "Industrial Location in Developing Countries." World Bank Research Observer 23 (2): 219­46. 60 Poor Places, Thriving People Dillinger, W. 2007. Poverty and Regional Development in Eastern Europe and Central Asia. Europe and Central Asia Chief Economist's Regional Working Paper Series. Washington, DC: World Bank. Duranton, G., and V. Monastiriotis. 2002. "Mind the Gaps: The Evolution of Regional Earnings Inequalities in the U.K., 1982­1997." Journal of Regional Science 42: 219­56. El Laithy, H., and K. Abu-Ismail. 2005. "Poverty in Syria: 1996­2004." Unpublished. Esber, R. 2008. "Gender and Spatial Disparities in the Middle East and North Africa: A Study of Saudi Arabia, Egypt and Tunisia." Unpublished. Esman, M. J., and I. Rabinovitch. 1988. Ethnicity, Pluralism and the State in the Middle East. Ithaca, NY: Cornell University Press. Fan, S., P. Hazell, and S. Thorat. 1999. Linkages between Government Spending, Growth and Poverty in Rural India. Washington, DC: IFPRI (International Food Policy Research Institute). Glaeser, E. L., and J. D. Gottlieb. 2009. The Economics of Place-making Policies. Washington, DC: National Bureau of Economic Research. Haddad, A. 1996. Poverty in Lebanon. Beirut: UN Economic and Social Council for Western Asia (ESCWA). Haut Commissariat au Plan. 1998, 1999, 2000, 2007. "Annuaires Statis- tiques du Maroc" (Statistical Directory of Morocco). Rabat, Morocco. ------. 2007. Activité, Emploi, et Chômage. Rabat: Direction de la Statis- tique. Henderson, J. V., Z. Shalizi, and A. Venables. 2001. "Geography and Development." Journal of Economic Geography 1 (1): 81­105. Hudson, M. 1997. Arab Politics: The Search for Legitimacy. New Haven, CT: Yale University Press. INS (Institut National de la Statistique). 2000. Enquęte Nationale sur le Budget. La Consommation et le Niveau de Vie des Ménages. Tunis: INS. Kanbur, R., and A. Venables. 2005. Rising Spatial Disparities and Develop- ment. Helsinki: World Institute for Development Economic Research (UNU-WIDER). Karray, Z., and S. Driss. 2008. Regional Growth in Tunisia: Economic Geog- raphy Forces and Industrial Structure. Cairo: Economic Research Forum. The Political Demand: Spatial Equity without Compromising Productivity 61 Krugman, P. 1991. "Increasing Returns and Economic Geography." Journal of Political Economy 99 (3): 483­99. Lall, S., H. Wang, and T. Munthali. 2009. "The Where of Infrastructure Investments in Malawi: Insights from a Spatial Analysis." Unpub- lished background paper for World Bank. Lall, S., and S. Yilmaz. 2000. "Regional Economic Convergence: Do Policy Instruments Make a Difference?" Annals of Regional Science 35 (1): 153­66. Levy, H. 2004. "Rural Roads and Poverty Alleviation in Morocco." Paper presented at Scaling Up Poverty Reduction: A Global Learning Process and Conference, Shanghai, May 25­27, World Bank. Martin, P. 2005. "The Geography of Economic Inequalities in Europe." Swedish Economic Policy Review 12: 83­108. Nordhaus, W. D. 2006. "Geography and Macroeconomics: New Data and New Findings." PNAS 103 (10): 3510­17. Omran, A. R. 1980. Population in the Arab World: Problems and Prospects. New York: United Nations Fund for Population Activities. Owen, R. 1992. State, Power, and Politics in the Making of the Modern Middle East. London: Routledge. PCBS (Palestinian Central Bureau of Statistics). 2006. The Palestinian Expenditure and Consumption Survey 2005. Ramallah: PCBS. Romer, P. 1994. "The Origins of Endogenous Growth." Journal of Economic Perspectives 8 (1): 3­22. Shami, S. 1993. "The Social Implications of Population Displacement and Resettlement: An Overview with a Focus on the Arab Middle East." International Migration Review 27 (1): 4­33. Shlash, A. 2009. Iraq High Committee for Strategy Reduction. Draft Strategy Review presentation. June. Amman. UN Habitat. 2004. The State of the World's Cities 2004/5: Globalization and Urban Culture. New York: United Nations. United Nations. 2006. World Population Policies 2005. New York: United Nations, Department of Economic and Social Affairs, Population Division. Van de Walle, D. 2009. "Impact Evaluation of Rural Roads Projects." Journal of Development Effectiveness 1 (1): 15­36. 62 Poor Places, Thriving People Venables, A. 2005. "Spatial Disparities in Developing Countries: Cities, Regions, and International Trends." Journal of Economic Geography 5: 3­21. World Bank. 1999. Democratic and Popular Republic of Algeria. Growth Employment and Poverty Reduction. Washington, DC: World Bank. ------. 2002. Yemen: Poverty Update. Washington, DC: World Bank. ------. 2003. Deep Palestinian Poverty in the Midst of Crisis. Washington, DC: World Bank. ------. 2004. Jordan: Poverty Assessment. Washington, DC: World Bank. ------. 2007. Arab Republic of Egypt: Enhancing the Performance of the Road Sector. Washington, DC: World Bank. ------. 2008a. Islamic Republic of Iran: Spatial Patterns of Poverty and Economic Activity. Washington, DC: World Bank. ------. 2008b. Gender and Development in the Middle East and North Africa. Washington, DC: World Bank. ------. 2009a. World Development Report 2009: Reshaping Economic Geography. Washington DC: World Bank. ------. 2009b. Arab Republic of Egypt. Upper Egypt: Pathways to Shared Growth. Washington DC: World Bank. ------. 2009c. "Egypt Investment Climate 2009: Accelerating Private Enterprise-led Growth." World Bank, Washington, DC. ------. Forthcoming. Gender and Transport in the Middle East and North Africa Region: Case Studies from West Bank and Yemen. Washington, DC: World Bank. CHAPTER 3 Diagnose to Prescribe: Uncovering the Truth about Spatial Disparities Lagging areas are different from each other. Upper Egypt is not the Moroccan Rif, and the Moroccan Rif is not Baluchistan. The moment one starts to dig into the data, one discovers that every backward area has its own individual character. To misquote Tolstoy, all lagging areas are lagging in their own way. Recognizing the different species of lagging area is, therefore, the key to adopting the right package of policies. The symptoms might be the same, but the causes--and the required interventions--are very different. So, what questions must we ask when we diagnose a lagging area? First of all, does the country in question really have a spatial dispari- ties problem in the first place? The immediate answer may be, "Obvi- ously, yes." As the analysis below will show, however, spatial disparities are not always a significant part of overall socioeconomic inequality. If location is not a major determinant of household welfare, then it makes more sense for government to focus on general poverty reduction than on the development of specific places. The next question to ask is whether a location has a low level of wel- fare because of its lack of economic opportunity or because of the popu- lation's characteristics. Even if household welfare is linked to location, this link might be because certain areas contain people with low earning potential. If household characteristics are the main factor behind spatial inequality, then it makes more sense to focus on improving the potential of targeted groups of people--"investing in people"--than on developing the economy in targeted places. Supposing that we have proved that we have a genuine spatial inequal- ity problem, it is then necessary to characterize the lagging area in ques- tion. How is its level of development changing over time? Is it converging 63 64 Poor Places, Thriving People with leading areas? Is its population dwindling as people migrate in search of better opportunities? These trends will tell policy makers whether the lagging area problem is resolving itself over time without government intervention. Is it a high-density or a low-density area? Is it well connected to poles of economic activity, or is it isolated from them? As we shall see, the answers to these geographic questions will indicate to policy makers how the different options for action might work. What Kind of Spatial Inequality Does the Country Have? This section will take the reader through the process of spatial inequal- ity diagnosis. A problem thus defined clearly is the first step to selecting an appropriate policy response. Where Is the Country, on the Path to Agglomeration? The measurement of agglomeration has traditionally depended on indi- cators of urbanization. Measuring urbanization, however, is not straightforward. The first problem is that there is no standard definition of "urban" and "rural," which obstructs cross-country comparisons. The second problem is that areas change from rural to urban, which obstructs comparisons over time. Egypt is a case in point. Its census dis- tinguishes urban and rural population according to an administrative definition that labels most peri-urban settlements as rural, even if their appearance, habitat, and lifestyle are clearly urban (box 3.1). According to national and United Nations (UN) statistics, Egypt's population was 42.5 percent urban in 2000. If Egypt adopted the definition used by India, however, which defines urban areas as communities with 5,000 or BOX 3.1 The Nile Delta: Urban, Rural . . . or Rurban? "Follow the Nile north out of Cairo on the old agricultural road, and you find it hard to pinpoint where the city ends and the lotus-shaped Delta begins. Carpeted with redbrick apartment blocks and spliced with streets in every direction, the lush greenery of the Nile's splintered arteries is almost impossible to appreciate in isolation. This is where the urban and the rural get lost in each other, with livestock living in doorways and workers camping out in fields. In the past, literary giants venerated the Delta's wild marshlands; today, any clear-cut divisions between the metrop- olis and the countryside have long faded away." Source: Shenker 2009. Diagnose to Prescribe: Uncovering the Truth about Spatial Disparities 65 more inhabitants, then 86 percent of Egypt's total population would be classified as urban (World Bank 2009b). Then, as cities grow and over- flow into rural areas, population spreads from the city centers into new urban settlements, which might explain the curious statistic (Wahba 2007) that urban-rural migrants outnumber rural-urban migrants by a ratio of three to one. Researchers have recently surmounted this problem of definitions by creating an index of agglomeration that depends only on the distribution of population across space (Uchida and Nelson 2008; World Bank 2009a), so it is not affected by different countries' administrative definitions of urban and rural areas. This "agglomeration index" measures the propor- tion of the population living in an area with a population density of 150 people per square kilometer, which is also no more than an hour's travel time from a city (table 3.1). The index, therefore, provides an objective measure of agglomeration (although it seems to underestimate urbaniza- tion in desert city-states such as Djibouti or the United Arab Emirates). Using the agglomeration index, we can see that MENA's populations are more agglomerated than one might expect for their level of economic development; most of them are well above the trend line (figure 3.1). Egypt, Jordan, Lebanon, and Libya have levels of agglomeration that would normally be found in the most agglomerated of the richest developed countries. The populations of the Syrian Arab Republic, Morocco, and TABLE 3.1 How Agglomerated Are the Economies of the Middle East and North Africa? Share of urban population Agglomeration Economy (UN database) Rank index Rank Algeria 59.8 9 56.9 9 Djibouti 83.4 2 50.4 12 Egypt, Arab Rep. 42.5 12 92.6 1 Iran, Islamic Rep. 64.2 7 62.8 7 Iraq 67.8 6 72.2 6 Jordan 80.4 4 79.4 5 Lebanon 86 1 86.8 3 Libya 83.1 3 83.4 4 Morocco 55.1 10 53.6 10 Syrian Arab Republic 50.1 11 59.1 8 Tunisia 63.4 8 51.9 11 West Bank and Gaza 71.5 5 90 2 Yemen, Rep. 25.4 13 25.5 13 Source: Uchida and Nelson 2008. Note: The index is the percentage of population living in areas with a population density of at least 150 people/km2 who are no more than an hour's travel time from a city of 50,000 people. 66 Poor Places, Thriving People FIGURE 3.1 MENA Is Agglomerated for Its Level of Development 120 100 Lebanon agglomeration index, 2000 Egypt, Arab Rep. 80 Libya Jordan Syrian Arab Republic Iran, Islamic Rep. 60 Algeria Djibouti Morocco Tunisia 40 China Yemen, Rep. 20 0 2.0 2.5 3.0 3.5 4.0 4.5 5.0 log of GNI per capita, ppp, current international $, 2000 (Libya 2007) Sources: World Bank, World Development Indicators database, Uchida and Nelson 2008. Note: GNI = gross national income; PPP = purchasing power parity. Djibouti are agglomerated to the extent found in countries 50 percent richer. Only the Republic of Yemen is below the trend line, while Algeria and Tunisia are on it. Why is MENA a region with such high agglomeration of its popula- tions? One possible reason is water. Before humans learned to build long water conveyance systems or to drill deep tubewells, the only places fit for settlement and farming were areas with good soils plus plentiful rain- fall, surface water, or shallow aquifers. Most land in the MENA region does not meet these conditions; it is either too arid or too mountainous to support settlement and farming. One can take the percentage of a country's land that is either arable or forested as an indicator of the area with adequate water for agrarian settle- ment. In the European Union (EU), this ratio is usually above 50 percent and is never lower than 30 percent (Cyprus). In MENA, this ratio never rises above 28­30 percent (Lebanon, Morocco, and the Syrian Arab Republic) and is lower than 5 percent in Algeria, Djibouti, Jordan, Libya, and the Republic of Yemen (World Bank 2009a). So historically, MENA's rural populations were concentrated in coastal strips, in mountain valleys, and along rivers such as the Nile, Tigris, and Euphrates. As populations grew, there was little new fertile rainfed land to bring under cultivation. By necessity, therefore, MENA agriculture was from an early date characterized by high labor-land ratios, and settled areas had high population densities. The famous Middle Eastern cities of Diagnose to Prescribe: Uncovering the Truth about Spatial Disparities 67 antiquity were often either riverine commercial hubs for irrigated farming areas, such as Jericho and Babylon, or coastal hubs for maritime trade, such as Alexandria and Tyre. Map 3.2 illustrates how water availability has shaped settlement patterns in Iraq. In the mountainous areas of northern Iraq, where rainfed agricul- ture has been practiced for centuries, settlements are evenly spread, inter- rupted only by high mountain ranges, the arched bands visible to the north of the map. Average population per settlement is relatively low in these scattered villages (Ameen 2008). In the rest of Iraq, however, agriculture has historically been based on irrigation from surface water. Its population is, therefore, concentrated in riverine bands, with a higher level of popula- tion per settlement. But high-density agriculture is not the whole story of MENA's high level of agglomeration. (See map 3.1.) MENA's political and gover- nance history has also played a role in driving population concentra- tions. The politics of agglomeration is a question that we will examine in chapter 4. Whatever the cause, why does it matter to policy makers to know that their country is more agglomerated than other countries at the same level of development? Put simply, it may be an indication that a country does not need significant additional population agglomeration to obtain developed-country economies of agglomeration. Rather, the policy MAP 3.1 MENA's Pockets of Nonaridity IBRD 38169 NOVEMBER 2010 HYPER-ARID < 0.05 SEMI-ARID 0.2 ­ 0.5 ARID 0.05 ­ 0.1 DRY-SUBHUMID 0.6 ­ 0.65 ARID 0.1 ­ 0.2 HUMID > 0.65 This map was produced by the Map Design Unit of The World Bank. The boundaries, colors, denominations and any other information 0 500 1000 shown on this map do not imply, on the part of The World Bank Group, any judgment on the legal status of any territory, or any KILOMETERS endorsement or acceptance of such boundaries. Sources: World Bank Spatial Analysis Team and DECRG. 68 Poor Places, Thriving People MAP 3.2 Water-Based Settlement Patterns in Iraq: Rainfed and Riverine IBRD 38170 NOVEMBER 2010 POPULATED PLACE 0 50 100 150 200 250 KILOMETERS This map was produced by the Map Design Unit of The World Bank. The boundaries, colors, denominations and any other information shown on this map do not imply, on the part of The World Bank Group, any judgment on the legal status of any territory, or any endorsement or acceptance of such boundaries. Source: National Geospatial Intelligence Agency, http://earth-info.nga.mil/gns/html/cntyfile/iz.zip, accessed July 2009, reproduced in Ameen 2008. Note: A dot = a settlement. emphasis should be on managing urban density to avoid excessive con- gestion costs and overloading of infrastructure and services. The Republic of Yemen's level of population agglomeration, however, is the exception. It is far below what one would expect for its level of GDP per capita. This finding suggests that Yemeni policy makers should expect, plan for, and certainly not attempt to discourage, urbanization. A Case of Delayed Agricultural Transition? The shift of labor from agriculture into manufacturing and services is part of the agglomeration process. In a low-income, farm-based society, people have to live where the fertile land is. Even the first urban centers grow as "market towns" within farming areas. But as the economy grows and households become richer, agricultural output becomes a smaller share of total consumption. Manufacturing and services begin to Diagnose to Prescribe: Uncovering the Truth about Spatial Disparities 69 dominate the economy, and most enterprises are no longer tied to the location of a natural resource. Therefore, they cluster into urban cen- ters to take advantage of economies of agglomeration. So economic growth, the shift out of farming, and agglomeration are all intercon- nected strands of spatial development. The transition out of agriculture is a two-stage process (World Bank 2008a). First comes the change in the economy's sectoral composition, with manufacturing and services overtaking farming as the main source of value added. Then comes the shift in employment, with labor moving away from the land into workshops and offices. In between, there is an interim period, when agriculture has lost its share of value added, but rural areas are still a reservoir of low-productivity employment and poverty. If agriculture has lost its share of value added but has not yet lost its share of employment, this inevitably means that the labor productivity gap between the farm and nonfarm sectors is temporarily even bigger than normal. Then eventually the higher labor productivity in manufacturing and services will draw labor out of agriculture and rural areas into the urban agglomerations to work in manufacturing and services. As figure 3.2 shows, most MENA13 countries are in that interim period--the bottom right-hand quadrant--when agriculture has ceded its share of value added, but rural areas have retained their share of poverty. This is the time of transition when rural-urban disparities are most acute, because the gap in labor productivity between rural and urban areas is at its highest. If MENA economies are to follow the nor- mal path, the way forward is for labor to shift out of farming, raising land-labor ratios, labor productivity in agriculture, and rural incomes. If MENA is to raise agricultural incomes, much of the gains will probably have to come from a shift of labor out of farming, so that the area of fertile land per worker can increase. Table 3.2 breaks down the difference between labor productivity in MENA agriculture and labor productivity in a developed country with similar agro-climatic condi- tions (Spain). It shows that most of the difference in labor productivity is due to Spanish farmers having more land to work with, rather than Spanish land being economically more productive than MENA land. The exception is Iraq, which has comparatively high land-labor ratios but low land productivity. Improving land productivity through improved irrigation efficiency, higher-value cropping, slicker marketing chains, and agronomic improvement, will of course always be important for MENA; nevertheless, rare are the countries that have reduced the productivity gap between farm and nonfarm labor without workers leav- ing agriculture. But something very special seems to be happening to MENA's agri- cultural labor force. At middle-income levels of GDP per employed 70 Poor Places, Thriving People FIGURE 3.2 Poverty Is Rural and Growth Is Nonagricultural, in MENA Economies 80 agriculture-based countries agriculture's contribution to growth, 1990­2005, (%) 60 40 20 YEM EGY MAR TUN 0 transforming countries urbanized countries ­20 0 0.2 0.4 0.6 0.8 1.0 rural poor/total poor, 2002 actual poverty data predicted poverty data Source: World Bank 2008a. Note: The contribution of agriculture to growth is defined as the agricultural growth rate times the sector average share over the period divided by the GDP growth rate. Rural shares in poverty marked with a circle are from Ravallion, Chen, and Sangraula (2007), using the $2.15-per-day poverty line. Rural shares of poverty marked with a diamond are predicted with an estimated regression of the rural share of poverty on rural share of population, agricultural share in GDP, log of GDP per capita in 2000 US$, and regional dummies. EGY = Egypt, Arab Rep., MAR = Morocco, TUN = Tunisia, YEM = Yemen, Rep. person, it is usual for employment in agriculture to stabilize or shrink as workers move out of farming into manufacturing and service jobs. Figure 3.3 presents data for 162 countries on the relationship between a country's overall labor productivity and the growth rate of employment in agriculture: the richer the country and the higher labor productivity, the lower the rate of growth of agricultural employment. Countries usually have a declining farm labor force when they reach the level of development of the Syrian Arab Republic, Egypt, Jordan, Algeria, Tunisia, or the Islamic Republic of Iran. However, in MENA there is growing agricultural employment--and quite rapid growth too: more and more people are trying to make a liv- ing out of MENA's limited land and water resources. Between 1990 and 2005, the number of people working in farming grew by 32 percent in Diagnose to Prescribe: Uncovering the Truth about Spatial Disparities 71 TABLE 3.2 Difference between Agricultural Labor Productivity in MENA and Spain: How Much Is Due to the Land-Labor Ratio? percent Share due to difference Share due to difference in land productivity (%) in land per worker (%) Egypt, Arab Rep. ­64 164 Djibouti ­47 147 Jordan 2 98 Syrian Arab Republic 15 85 Iran, Islamic Rep. 17 83 Yemen, Rep. 19 81 Tunisia 24 76 Algeria 26 74 Morocco 27 73 Iraq 90 10 Source: Log-divisia decomposition, using WDI data accessed May 17, 2009. FIGURE 3.3 MENA's Agricultural Labor Force Is Growing Fast for MENA's Level of Development 6 annual agricultural labor force growth, 1995­2005 4 Algeria Syrian Arab Republic 2 Jordan Iran, Islamic Rep. Tunisia 0 Egypt, Arab Rep. Yemen, Rep. ­2 ­4 Lebanon ­6 ­8 ­10 100 1,000 10,000 100,000 GDP per capita, 2000 (constant 2005 international $) Source: WDI. the West Bank and Gaza, 58 percent in Jordan, 24 percent in Tunisia, 44 percent in the Republic of Yemen, 43 percent in the Syrian Arab Republic, 60 percent in Algeria, 13 percent in Egypt, and 24 percent in the Islamic Republic of Iran. 72 Poor Places, Thriving People Morocco, Lebanon, and Libya are the only countries whose agricultural workforce is following the usual (stabilizing or downward) trend. Farm employment in Lebanon and Libya, however, is dominated by foreigners, which may affect the statistical reporting. Lebanese nationals are thought to provide only about a fifth of the country's agricultural workforce (World Bank 2007a), the remainder being Syrians and Palestinians. The retention of labor in agriculture is reducing land-labor ratios and farm sizes, which, in turn, depresses the productivity of agricultural labor. Between 1990 and 2005, the land-labor ratio fell by at least 15 per- cent in Algeria, Djibouti, Jordan, the Syrian Arab Republic, Tunisia, the West Bank and Gaza, and the Republic of Yemen, with smaller decreases in the Islamic Republic of Iran and Morocco.1 It is predicted that climate change will aggravate the decline in land-labor ratios in agriculture, as marginal land becomes unsuitable for cultivation. One estimate (FAO 2009, 15) predicts that MENA (including the Gulf Cooperation Council [GCC]) will lose 7 percent of its arable land by 2050, compared with a 9 percent increase worldwide. Tunisia is a case in point of the delayed transition of labor out of agri- culture (World Bank 2006a). Agricultural employment grew by 20 per- cent between 1993 and 2002, but there was no trend increase in labor productivity in agriculture. During the same period, productivity in manufacturing and services rose by 4.4 percent and 4.7 percent per year, respectively. As Tunisia's agricultural labor force grows, therefore, farm sizes shrink. The share by number of landholdings less than 5 hectares rose from 41 percent in 1976 to 53 percent in 2005. Many people trying to live off little land is a recipe for a lagging rural area in MENA. If one divides the Republic of Yemen's rural districts into three groups according to their poverty rates, the high-poverty districts have 24 people per hectare of arable land--three times as many as the low-poverty districts (World Bank 2009c). If one compares (richer) rural Lower Egypt with (poorer) Upper Egypt, the difference in agricultural incomes cannot be blamed on inefficient farming in Upper Egypt. In fact, Upper Egypt makes 36 percent more net income from a feddan (approximately 0.42 hectares) of land than Lower Egypt (World Bank 2009b). Upper Egypt's problem is that it has 0.17 feddan per capita in an average village instead of Lower Egypt's 0.32 feddan (World Bank 2009b). So the retention of labor in farming is a major correlate of lagging area status in MENA's rural areas. One reason why MENA has an unusually high rate of agricultural employment growth is that MENA has an unusually high rate of growth in its working-age population (figure 3.4). There is a trend relationship between population growth and development: richer countries have lower population growth. But MENA economies tend to have had higher-than-expected rates of growth of the population ages 15­64 for Diagnose to Prescribe: Uncovering the Truth about Spatial Disparities 73 FIGURE 3.4 MENA's Working-Age Population Has Been Booming, Compared with Other Countries at the Same Level of Development 6 annual growth of population ages 15­64, 5 Jordan Yemen, Rep. 4 Syrian Arab Republic 1990­2005 (percent) Djibouti Algeria Iran, Islamic Rep. 3 Egypt, Arab Rep. Lebanon 2 Morocco Tunisia 1 0 ­1 ­2 100 1,000 10,000 100,000 GDP per capita in 2000 (constant 2005 international $) Source: WDI. their levels of GDP. Other reasons for MENA's high rate of agricultural employment growth may include the following: · Disappointing employment growth in manufacturing · Limited alternative opportunities for women in rural MENA societies · Increasing recognition, by statistical offices, of women's participation in the labor force as unpaid labor on the family farm Moreover, environmental degradation and climate change might increase urbanization pressures. Climate models predict that the MENA region will be particularly hard hit by climate change, with more fre- quent droughts, heat waves, and heavy storms. Between 1990 and 2002, all economies of MENA for which data were available experienced increased drought frequency.2 A continued deterioration of weather pat- terns would reduce the productivity of agriculture relative to other activ- ities, and thus--other things being equal--further encourage rural-urban migration. Cline (2007) forecasts that agricultural production in Algeria, the Islamic Republic of Iran, Iraq, and Morocco will fall by about 30 per- cent between now and 2080. The International Food Policy Research Institute (IFPRI) (Nelson and others 2009) predicts drops in production of 25­35 percent and 11­24 percent, respectively, for MENA's irrigated and rainfed wheat between 2000 and 2050. It has also been estimated that a one-meter rise in sea level, which may be brought about by the melting of polar ice, would compromise 13 percent of Egypt's agricul- tural land. 74 Poor Places, Thriving People Syria's drought of 2007­09 provided a stark illustration of a link between weather and movement to the cities, and the potential for climate change to accelerate future urbanization. According to a UN study, an estimated 200,000­250,000 farmers and their families were forced to abandon their land due to this drought. Peri-urban areas close to the cities of Damascus, Aleppo, and Hama received the majority of the displaced (Oweis 2009; IRIN News 2009). MENA's burgeoning farm workforce and stressed natural resource base matter for spatial policy. The region's agricultural sector has already done an exceptional job in absorbing surplus labor. Given the region's land and water constraints, a shift of labor out of agriculture is, in a sense, overdue. With a shift of labor out of agriculture will come a shift of pop- ulation into urban settlements. Are Poverty and Inequality Really a Lagging Areas Question? Without exception, all MENA governments see lagging area develop- ment as important for inclusive development. But how much attention should actually be given to the spatial dimension in the design of policies for socioeconomic inclusion? At one extreme, poverty could be completely nonspatial, with the mix of people of different standards of living being the same at all locations across the territory. Obviously in this case, it would make little sense for national or sectoral development policies to focus on specific locations. At the other extreme, however, poverty could be a highly spatial phe- nomenon, with all the worse-off people living in certain locations and all the better-off in others. In this case, location-based development policies would be more justified. In reality, of course, every country will be somewhere in between these two impossible extremes. In the words of Sherlock Holmes, "There is nothing more misleading than an obvious fact." As we have seen earlier, strong political forces are pushing decision makers to focus their attention and resources on some places more than others. It is therefore essential for policy makers to have an empirical, evidence-based understanding of the degree to which location really does affect household welfare. Or, to put it in other words, what if spatial disparities are actually less important than they seem? If we put the spatial dimension of household welfare into its proper perspective, three principles are worth bearing in mind: · Mistrust averages · Focus on the absolute numbers in poverty · Break down household welfare into its spatial and nonspatial components Diagnose to Prescribe: Uncovering the Truth about Spatial Disparities 75 Mistrust Averages: Don't Let a Few Very Rich People Distort the Picture of Spatial Inequality Welfare inequalities are usually greater in leading areas than in lagging areas. This inequality is what one would expect: the country's economic elite usually lives in the richest places. This means that a few very well- off households can bump up the mean development indicators of leading areas, even though the living standards of many people in the leading area might be low. Using median instead of average values of household welfare therefore stops spatial disparities from looking bigger than they really are. In Syria in 2004, for example, the average per capita household consumption in the richest province, Damascus, was 90 percent higher than in the poorest province, El Quneitra. But the difference between consumption per capita in the median households in the two provinces was significantly narrower: 64 percent (WDR 2009). Table 3.3 shows how average per capita household consumption typically gives a stronger impression of spatial disparities than the median. This is not merely a statistical curiosity; it indicates the very real tension between first impressions and underlying reality in spatial diagnostics. Focus on the Absolute Numbers in Poverty It is important to focus on the absolute number of poor people living in different areas. MENA's poverty rates tend to be highest in the semi-arid, desert, and mountainous areas, where population density is lowest. The risk of focusing on poverty rates, therefore, is that one might overlook the large absolute numbers of poor living in high-density urban areas. Libya provides a good example of the poverty mass being located where poverty rates are low. According to the 2003 household survey, TABLE 3.3 Mean versus Median per Capita Consumption Coefficient of variation of area per capita household consumption Number of Country Average Median areas Djibouti (2002) 0.33 0.29 5 Egypt, Arab Rep. (2005) 0.35 0.30 24 Jordan (2002) 0.19 0.14 12 Morocco (1998)a 0.20 0.19 13 Syrian Arab Republic (2004) 0.20 0.18 14 Yemen, Rep. (2005) 0.34 0.26 21 Source: National household surveys, years as specified, Egypt July­September. Note: a. Excluding data from Guelmine Es-Semara. 76 Poor Places, Thriving People the three Shabiyat (administrative subdivisions) with the highest poverty rates are Al Kufrah, Ghat, and Mizdah.3 Al Kufrah is a massive desert area in the southeast of the country, bordering Egypt and Sudan on the east and Chad on the south. Its capital, Al Jawf, receives less than 3 mm of rain per year. Ghat is another desert province, in the far southwest, bordering Algeria. Mizdah is in a hilly area further north. Between them, these three inhospitable Shabiyat had a combined pop- ulation of 81,000 in the 1995 census, of which around 25,000 were living below the poverty line. The coastal capital province, Benghazi, on the other hand, has an estimated 50,000 people living in poverty, despite having a much lower poverty rate. This is not to claim that Benghazi is a lagging area. The point is that a focus on lagging areas should not mean that one overlooks the larger absolute numbers of poor people in leading areas. The Iraqi experience shows how poverty rates and absolute numbers both enter into policy making. In 2009, the High Committee for Iraq's Poverty Reduction Strategy Paper was pondering the implications of the Iraq Household Socioeconomic Survey (IHSES), Iraq's first postwar nationwide household income and consumption survey. The survey had revealed huge disparities in poverty rates between governorates, from 10 percent in Kirkuk and 13 percent in Baghdad right up to Sala'alddin, Babel, and Muthanna governorates, all with poverty rates above 40 per- cent. The High Committee's analysis of the poverty data rightly focused on the high poverty rates in the lagging rural governorates, and the first pillar of its Poverty Reduction Strategy placed emphasis on raising poor people's earnings from agriculture. But the High Committee also noted that the numbers of poor people were concentrated in the big cities: 13 percent of the poor lived in the capital and another 11 percent in the port city of Basra (figure 3.5). In fact there were a third more poor people in Iraq's four richest governorates than in the four poorest governorates because of the higher populations of the rich governorates. Therefore, another pillar of the High Committee's poverty strategy focused on housing, a critical issue for the poor in Iraq's major cities. Similarly, although the urban-rural divide may show urban areas to be better off, this should not obscure the existence of poor places within cities. Urban slums, for example, may have levels of poverty and human development similar to or even worse than rural areas. Table 3.4 shows a selection of human development indicators for slums in Egypt, Morocco, and the Republic of Yemen. Child health seems to be as poor in slums as in rural areas. Literacy rates are higher, which may be to due to the tendency of educated people to migrate toward the cities. In Egypt, however, the net primary education enrollment rate for boys and girls combined is about 7 percentage points higher in rural areas than in slums (UN Habitat 2006, 122). Diagnose to Prescribe: Uncovering the Truth about Spatial Disparities 77 FIGURE 3.5 High Poverty Rates Did Not Mean High Poverty Numbers in Iraq, 2007 60 Muthanna 50 % of population that is poor Sala'addin 40 Babil (poverty rate) Basra 30 20 Baghdad 10 Kirkuk 0 0 200,000 400,000 600,000 800,000 1,000,000 number of poor people Source: Shlash 2009. TABLE 3.4 Human Development Indicators in Slums % of malnourished children under five % of births attended % of children under years (children by skilled health five years with Female literacy underweight) personnel diarrhea rate (%) Country Urban Rural Slum Urban Rural Slum Urban Rural Slum Urban Rural Slum Egypt, Arab Rep. 6.8 9.6 10.7 86.7 59.0 76.6 16.8 20.2 23.7 73.7 42.9 60.2 Morocco 6.5 13.9 13.9 85.3 39.5 39.1 11.5 12.4 13.4 67.5 24.6 53.4 Yemen, Rep. 24.5 30.7 29.5 45.8 10.0 35.1 25.7 35.8 32.6 37.4 6.2 24.0 Source: UN Habitat 2006. In 8 of 10 MENA13 countries, the data show that the ratio of the slum population to the rural population is on the increase (table 3.5). The exceptions are Tunisia, which has implemented a successful program of slum improvement, and Egypt, where the slum population data need some other explanation. In the Islamic Republic of Iran, Iraq, Lebanon, and Libya, the slum population already matches or exceeds the rural population. In short, the problem of poor people in rich areas is catch- ing up with the problem of poor people in poor areas. Break Down Household Welfare into Its Spatial and Nonspatial Components: Check That Spatial Disparities Are Really a Big Part of the Inequality Story If we are to know for certain how important spatial disparities are as a part of overall inequality, it is necessary to have some objective, statistical 78 Poor Places, Thriving People TABLE 3.5 Rural and Slum Populations in MENA (thousands) Slum Slum Rural Rural Ratio rural Ratio rural population population population population to slum to slum Country 1990 2005 1990 2005 1990 (%) 2005 (%) Algeria 1,508 2,370 12,110 12,057 803 509 Egypt, Arab Rep. 14,087 11,015 31,152 41,816 221 380 Libya 1,242 1,867 1,060 1,361 85 73 Morocco 4,457 6,054 12,470 3,564 280 224 Tunisia 425 188 3,433 3,480 808 1851 Iran, Islamic Rep. 17,094 21,763 23,773 22,868 139 105 Iraq 6,825 9,992 n.a. 5,610 n.a. 56 Jordan 388 741 881 1,174 227 158 Lebanon 1,142 1,811 503 537 44 30 Syrian Arab Republic 629 1,012 6,500 8,842 1033 874 Yemen, Rep. 1,787 3,803 9,740 14,999 545 394 Sources: UN Habitat 2006; World Bank World Development Indicators. Note: n.a. = not available. measure of how inequality breaks down into its spatial and nonspatial components. One well-tested technique for doing so takes a measure of overall inequality, the Theil index, and divides it into two parts: that which is attributable to inequality within locations and that which is attributable to inequality between locations (Bourguignon 1979; Shorrocks 2005). Comparing different countries' spatial components of inequality raises methodological complications (box 3.2); nonetheless, there are good reasons to trust the main messages from the exercise. Applying the decomposition to data on per capita household con- sumption in MENA shows some striking differences between economies. In some, spatial inequalities are a big part of the inequality story, but in others they are not. Table 3.6 gives the picture for disparities between urban areas and rural areas. It looks at two different measures of inequality in households' economic welfare. The National Household Survey (NHHS) data mea- sures per capita household expenditure. The Demographic and Health Survey (DHS) data measure a wealth index based on the household's ownership of physical assets (Rutstein and Johnston 2004). The second column of table 3.6 gives the overall Theil index for households' economic welfare from each survey. The third column gives the component of the overall Theil index explained by urban-rural dis- parities. The fourth column is the one to focus on: it shows the share of total inequality that takes the form of urban-rural disparities (the third column divided by the second column). The results show dramatic differences between MENA countries. In Jordan and Djibouti the rural-urban issue is not at all an important Diagnose to Prescribe: Uncovering the Truth about Spatial Disparities 79 BOX 3.2 Decomposing MENA's Inequality into Its Spatial and Nonspatial Components: A Worthwhile Exercise, Despite the MAUP The problem with comparing different countries' inequality decompositions is that com- parisons depend on the geographical unit of analysis. This "Modifiable Areal Unit Prob- lem" (MAUP) consists of a "Scale Problem" (the size of the unit of analysis) and an "Aggregation Problem" (where the boundaries are drawn). The effect of the Scale Prob- lem is that the smaller the geographical unit, the more important spatial inequality appears to be. Imagine an analysis, for example, in which the units were so small that each contained only one household. In this case, inequalities would appear to be 100 percent spatial. At the other extreme, if all the households in the study were listed in the same location, the spatial component of inequalities would appear to be zero. However, despite this flaw, the decomposition analysis for MENA may still be worth presenting. First, the two-way rural-versus-urban analysis standardizes the breakdown at two very large populations, reducing intercountry scale effect differ- ences. Second, the interprovincial analysis produces nearly the same ordering of coun- tries as the rural-versus-urban analysis, suggesting that there is more to the ranking than scale effects. Third, despite Morocco's high number of provinces and a large rural-urban divide, the decomposition portrays Morocco's interprovincial inequality as a very low share of total inequality. This is the opposite of what one would expect if the Scale Effect were the main determinant of the decomposition results. Source: Openshaw 1984. component of overall inequality, which is what one might expect for small, urbanized countries. In Egypt and Morocco, on the other hand, the rural-urban divide explains 16­29 percent of total inequality, depend- ing on which survey one considers. There is a straightforward policy implication here. Policies that target the development of lagging (rural) areas are not likely to make a sizeable impact on socioeconomic disparities in countries toward the top end of table 3.6, such as Jordan, Djibouti, the Republic of Yemen, and Syria. Policies that target poor households, whether in urban or rural areas, would stand a better chance. Spatially targeted development initiatives for rural areas might make more sense for countries toward the bottom of table 3.6, such as Egypt and Morocco. Even so, in no country in the table does rural-urban inequality account for more than a fifth of total inequality of household expenditure. Does this mean that spatial inequality is relatively unimportant? Some authors (Elbers 2007) argue that this Theil decomposition calculation understates the problem. The main reason is that society's aim is usually not to abolish 80 Poor Places, Thriving People TABLE 3.6 Where Does the Urban-Rural Divide Count? Maximum ELMO index = actual Share of Theil possible urban-rural disparity Overall Urban-rural index explained urban-rural versus maximum Households' economic Theil component by urban-rural component of possible urban-rural welfare, by urban and rural index of Theil index disparities (%) Theil index disparity (%) Jordan DHS 2002 0.20 0.00 2 0.12 3 Jordan NHHS 2002 0.29 0.01 2 0.18 4 Djibouti NHHS 2002 0.32 0.01 3 0.16 7 Yemen, Rep. NHHS 1999 0.24 0.01 5 0.15 8 Syrian Arab Republic NHHS, 2004 0.30 0.02 7 0.15 14 Yemen, Rep. NHHS 2005 0.38 0.03 9 0.17 20 Egypt, Arab Rep. DHS 2005 0.24 0.04 16 0.14 28 Egypt, Arab Rep. NHHS 0.29 0.05 17 0.13 39 Morocco NHHS 2001 0.31 0.05 17 0.16 34 Morocco NHHS 1999 0.28 0.05 19 0.15 37 Morocco DHS 2004 0.33 0.10 29 0.19 51 Lebanon NHHSa -- -- -- -- -- Source: Sommeiller 2008, using household data from the sources listed in the first column, years as specified, Arab Republic of Egypt July­September only. Note: -- = not available. DHS = Demographic and Health Survey, NHHS = National Household Survey. a. No urban-rural variable in the Lebanon dataset. inequality altogether but to give different social groups the same oppor- tunities. In other words, the goal is economic justice, not economic equal- ity. So what matters is not whether rural-urban disparities account for a high proportion of total inequality, but whether rural people and urban people have the same chance of being at the same position in the socioe- conomic hierarchy.4 The "ELMO index,"5 therefore, serves as a measure of economic jus- tice. It takes the existing income distribution, imagines that all the richest people lived in urban areas and all the poorest people lived in rural areas, and calculates the Theil measure of rural-urban disparities on this basis. This is the "maximum possible urban-rural component of Theil index," in the fifth column of table 3.6, and represents a situation of "maximum economic injustice" between urban and rural areas. It represents the situ- ation in which the rural population gets all the worst places on the socio- economic ladder. Then the ELMO index in the last column compares the actual urban-rural disparity with the maximum urban-rural disparity. Using the ELMO index, we can see that the ranking of countries is more or less the same. Jordan, Djibouti, the Republic of Yemen, and Syria have ELMO scores of 20 percent or less, meaning that their rural-urban divide could be five times bigger given the current level of Diagnose to Prescribe: Uncovering the Truth about Spatial Disparities 81 overall economic inequality. Egypt and Morocco again stand out as the countries in which rural-urban disparities are a significant part of the inequality picture. MENA is not particularly badly off, compared with the rest of the world, when it comes to the urban-rural divide. To put MENA's rural- urban disparities into perspective, one can compare them with other developing country regions. Rural-urban disparities in Sub-Saharan Africa, East Asia and the Pacific, and Latin America and the Caribbean average 14­18 percent of overall disparities (world region averages). Their ELMO indexes average 25­28 percent. By this measure, Morocco is at or just above the global norm, while several other MENA economies are well below it. The differences among MENA economies in the importance of spa- tial disparities are again quite dramatic when one considers interprovin- cial disparities rather than urban-rural disparities. Table 3.7 repeats the exercise for administrative subdivisions, such as provinces, gover- norates, and mohafazat. Interprovincial inequality is a very small com- ponent of overall inequality for Syria, Jordan, and the Republic of Yemen. This inequality means that "lagging area" policies targeting spe- cific provinces might be less effective at attacking inequality than policies that target disadvantaged sections of society nationwide. In Lebanon and TABLE 3.7 Where Does the Interprovincial Divide Count? ELMO index = actual Maximum interprovincial Share of Theil possible disparity versus Overall Interprovincial index explained interprovincial maximum possible Households' economic Theil component by interprovincial component of interprovincial welfare by provincea index of Theil index disparities (%) Theil index disparity (%) Syrian Arab Republic NHHS, 2004 0.31 0.01 3 0.21 5 Jordan DHS 2002 0.20 0.01 4 0.17 4 Yemen, Rep. NHHS 1999 0.24 0.01 5 0.21 6 Morocco NHHS 1999 0.28 0.02 8 0.22 10 Jordan NHHS 2002 0.29 0.03 9 0.21 12 Yemen, Rep. NHHS 2005 0.38 0.03 9 0.24 14 Morocco NHHS 2001 0.32 0.03 10 0.23 14 Egypt, Arab Rep. DHS 2005 0.24 0.04 17 0.22 19 Djibouti NHHS 2002 0.32 0.06 17 0.26 21 Egypt, Arab Rep. NHHS 0.29 0.06 21 0.19 31 Lebanon NHHS 0.28 0.06 21 0.22 27 Morocco DHS 2004 0.33 0.10 29 0.29 33 Source: Sommeiller 2008, using household data from the sources listed in the first column, years as specified, Arab Republic of Egypt July­September only. Note: a. To calculate the interprovincial component of inequality, provinces are divided into four sets, from richest to poorest. 82 Poor Places, Thriving People Egypt, however, interprovincial inequality accounts for more than a fifth of total inequality. In Djibouti, where the rural-urban divide was insignif- icant, interdistrict inequality is about a sixth of overall inequality. In Morocco, the situation is very interesting. Although rural-urban dispar- ities were quite important in Morocco, the household expenditure sur- vey data from 1999 and 2001 revealed that interprovincial inequality was a very small part of overall inequality. And yet the wealth index of the Demographic and Health Survey makes out that a full 29 percent of overall economic inequality is accounted for by interprovincial dispari- ties. This finding means that interprovincial disparities in ownership of assets such as housing, livestock, land, and consumer durables are impor- tant, but interprovincial disparities in household expenditure are not. Disparities that May Not Be as Spatial as They May Seem The bottom line is that socioeconomic disparities might not be as impor- tant in measured living standards as they are in politics. We have seen earlier that politics tends to "spatialize" issues. Therefore, it is all the more important for policy makers to perform some cold, objective analy- sis of the real role of location in determining people's life-chances. Low Economic Opportunity . . . Or Low Human Development? As we saw in table 3.7, Egypt's interprovincial disparities account for around a fifth of total inequality. The lagging area is the Nile Valley south of Cairo, also known as Upper Egypt. Although Upper Egypt has only 40 percent of the country's population, it has 60 percent of its poverty and 80 percent of its severe poverty. Here is a striking fact: demography explains 62 percent of the con- sumption gap between Upper and Lower Egypt (World Bank 2009b) (table 3.8). Much thought has gone into identifying which geographical handi- caps have blocked Upper Egypt's development: its distance from Cairo and major ports, a lack of transportation infrastructure, or a shortage of land and other productive natural resources. The fact is, however, that Upper Egyptian workers earn only 8 percent less on average than their Lower Egyptian counterparts and have only a slightly higher employ- ment rate, so economic opportunity might not be the main factor hold- ing them back. The problem is that when Upper Egyptian workers return home, they have to share their earnings with more nonworking family members who might be too young or too old, or excluded from the labor market by social constraints on women. In other words "demo- graphic dependency," the ratio of dependents to people of working age, is higher in Upper Egypt than in Lower Egypt. Unfavorable household Diagnose to Prescribe: Uncovering the Truth about Spatial Disparities 83 TABLE 3.8 Explaining the Difference between per Capita Consumption in Upper and Lower Egypt Farming is Farming is a No farming, but All main subsidiary agricultural Nonfarm households income source income source wages activities only Total gap ­18% ­13% ­19% ­23% ­21% of which: of which: of which: of which: of which: Demographic dependency ­11% ­12% ­13% ­11% ­9% Earnings per worker ­8% ­2% ­8% ­13% ­12% Employment rate +1% +1% +2% +1% +1% Percentage of total gap 100 100 100 100 100 Demographic dependency 62 95 67 46 46 Earnings per worker 45 13 41 57 57 Employment rate ­6 ­8 ­9 ­4 ­3 Source: World Bank 2009b. characteristics are therefore causing Upper Egypt's purchasing power to lag. Such households would be worse off wherever they lived. Various measures of a household's composition need to be considered as part of the demography question: · The size of the household: the more household members there are, the lower consumption per capita will be for any given amount of earnings · The number of people of working age in the household · The number of females in the household, because there are cultural limitations on women's earning opportunities · The gender of the head of the household, because female and male households heads have different opportunities to access markets and services on behalf of their household members · The age of the head of the household, because a person's earning power typically peaks in midlife before falling again as old age approaches6 Education is another--nondemographic--household characteristic that affects a household's chances of succeeding. A household with more edu- cation will tend to fare better than a household with less education, wherever it lives. Leading areas usually have more than their share of well-educated people, so part of the difference between the well-being of leading and lagging areas might be because of the differing education levels of their populations. Spatial disparities in education levels might be because educational services are better in leading areas, because parents 84 Poor Places, Thriving People in leading areas are rich enough to afford better education, or because educated people in lagging areas migrate to leading areas. One can, therefore, break down spatial disparities in households' con- sumption levels into three components: · Spatial disparities caused by spatial differences in economic opportunity · Spatial disparities caused by differences in household demography · Spatial disparities caused by differences in education The relative importance of these three components should determine the policy response. If demography and education are important causes of spatial inequality, then it makes sense to focus on human development. Investments in education, opportunities for women, and reproductive health are then a win-win policy; they reduce spatial disparities without the economic inefficiency of pushing production into lagging areas. One way of distinguishing between the influences of human develop- ment and economic opportunity is to work out what the economic wel- fare of households with the same characteristics would be in different geographical locations. The separation of spatial disparities into their demographic and educational components is presented next for 8 of the 13 MENA economies: Djibouti, Egypt, Iraq, Jordan, Lebanon, Morocco, Syria, and the Republic of Yemen. The findings are powerful, inasmuch as they show important differences in the form of spatial dis- parities among different countries of the MENA region. Each set of data tells a unique story and points each economy toward its own specific pol- icy direction. Before looking at each economy, it is important to mention a key point. The analysis only uses the limited range of household characteris- tics mentioned earlier, and for education, it measures only the household head's level of schooling. If more household characteristics had been included, these characteristics would have appeared even more impor- tant compared with spatial differences in economic opportunity. The statistical technique described next works only with average con- sumption values and not with medians. Otherwise, medians would have been used to give a more representative picture of spatial inequalities. Figures 3.6 and 3.7 confirm that demography is a critical cause of Egypt's spatial disparities. If all of Egypt's governorates had the same demographic profile, the coefficient of variation of mean governorate per capita household consumption would be 0.27 instead of 0.35. The rural- urban gap would be two-thirds of what it actually is, a full 17 percentage points lower. A lagging governorate in Upper Egypt, like Fayoum, would be 36 percent behind Cairo instead of 55 percent. Assiut, deeper into Upper Egypt, would be 59 percent behind Cairo instead of 71 percent Diagnose to Prescribe: Uncovering the Truth about Spatial Disparities 85 FIGURE 3.6 Demography Matters for Intergovernorate Disparities in the Arab Republic of Egypt 120 mean per capita household consumption 100 as % of richest area 80 60 40 20 0 ex iro rt a id M Giza ar uh a mi z a Ka ly ya Sh a Da eikh uf a Fa yya um Ke r As na Be an M ira ni ia ar f So a As g t xo Sh Sue siu Da Sue Po ndri yy Ism ett El bi on li y ha Sa Be en Ka aily ky Gh ro M kah w Al Ca he fr ou yo Lu be e at a cities real life standardized demography standardized demography and education Source: World Bank staff calculations, from 2004­05 HIECS, July­September. FIGURE 3.7 Remove Demographic and Educational Differences, and the Arab Republic of Egypt's Urban-Rural Divide Shrinks Away, 2005 60 48 50 rural-urban gap (%) 40 31 30 22 20 10 0 real life standardized standardized demography demography and education Source: World Bank staff calculations from 2004­05 HIECS, July­September only. 86 Poor Places, Thriving People behind. The governorates that share Cairo's favorable demography are either urban (Alexandria, Port Said, and Giza) or new developments (S. Sinai). Overall, education is important too, albeit less so than demography, as illustrated by the gap between the top line and the middle line in figure 3.6 and between the middle and right-hand columns in figure 3.7. The metropolitan governorates of Cairo, Alexandria, and Giza share the same educational advantages. If there were no spatial disparities in demography and education, the rural-urban gap might be less than half of what it is (figure 3.7). In Djibouti, different lagging districts have different types of spatial disparity (figure 3.8). The districts of Tadjoura, Obock, Ali-Sabieh, and Dikhil all appear at first sight to be in the same situation, with per capita household consumption well below Djibouti City's, but the breakdown of causes is different. In fact, disparities between Tadjoura and Obock on the one hand, and Djibouti City on the other, increase when one takes away their demographic differences, meaning that these two outlying districts have more favorable demographic characteristics than the capi- tal. For Ali-Sabieh and Dikhil, demographic differences do play a role in creating spatial disparities, as witnessed by the upward curling tail in the demography-adjusted curve. For Dikhil, demography is particularly cru- cial. If all of Djibouti's districts had the same demography, mean per capita consumption in Dikhil would be 65 percent of that in Djibouti City as opposed to the observed 46 percent, and Dikhil would be the richest district outside Djibouti City instead of the poorest. For education it is the other way around; there are significant educa- tional differences between Djibouti City on the one hand and Tadjoura and Obock on the other. For Ali-Sabieh and Dikhil, however, educa- tional disadvantage is not a major factor after demography has been taken into account. If one compares urban (mostly Djibouti City) with rural Djibouti (figure 3.9), the demography of rural Djibouti helps to mitigate the rural-urban gap. If rural and urban areas had the same demographic characteristics, the gap would be 4 percentage points higher. This is a very unusual finding, as we shall see when we look at the other coun- tries. Education, however, has the usual prodisparity effect. With demographic characteristics accounted for, education disparities add 11 percentage points to the rural-urban gap. Jordan is an interesting case, because at the time of this writing, the country is considering a strategy of regionalization and regional develop- ment to combat spatial inequality. In fact, Jordan has low intergover- norate disparities (figure 3.10). The coefficient of variation of governorate means per capita household consumption is only 0.19, compared with Diagnose to Prescribe: Uncovering the Truth about Spatial Disparities 87 FIGURE 3.8 Djibouti's Interdistrict Disparities: Tadjourah and Obock Lag in a Different Way from Ali-Sabieh and Dikhil 100 90 mean per capita household consumption 80 70 as % of richest area 60 50 40 30 20 10 0 Djibouti city Tadjourah Obock Ali-Sabieh Dikhil real life standardized demography and education standardized demography Source: 2002 National Household Survey. FIGURE 3.9 Djibouti's Rural-Urban Gap, 2002 50 45 45 41 rural-urban gap (% of urban per 40 34 capita consumption) 35 30 25 20 15 10 5 0 real life standardized standardized demography demography and education Source: 2002 National Household Survey. 88 Poor Places, Thriving People FIGURE 3.10 Intergovernorate Disparities Are Small and Largely Demographic in Jordan 100 90 mean per capita household consumption 80 70 of richest area 60 50 40 30 20 10 n a lqa id a sh n a ela n q k ra ma ab ab lou rq aa ra Irb ra Ba fe Ka Za af ad Aq M Am Je Aj Ta M M real life standardized demography standardized demography and education Source: World Bank staff calculations from 2002 National Household Survey. 0.35 in Egypt and a full 0.46 in Morocco. And if there were no demo- graphic disparities in Jordan, the coefficient of variation would be only 0.12. The poorest governorate, Al-Mafraq, would be only 35 percent behind Amman instead of 49 percent behind. Apart from Maan and Zarqa, every other governorate would be within 25 percentage points of Amman. Moreover, this result is before the dispersing effect of educa- tional disparities is taken into account. The importance of demography for the rural-urban comparison is even more dramatic. Average rural per capita consumption is 24 percent behind urban levels. Take away demo- graphic disparities, and this difference dwindles to an insignificant 12 per- cent. Allow for education, and it becomes a mere 3 percent. In short, urban-rural disparities in economic opportunity should not be considered a priority policy challenge for Jordan. Lebanon, another small country, is a total contrast from Jordan.7 At 0.29, the intermohafaza (province) coefficient of variation of mean mohafaza per capita household consumption is moderate. Lebanon, however, is the only country in the sample for which there is no spatial disparity in demography or education to account for the gap between Beirut and the rest of the country (figure 3.11). (Data are not available for Nabatieh or S. Lebanon.) One possibility is that opportunities to Diagnose to Prescribe: Uncovering the Truth about Spatial Disparities 89 FIGURE 3.11 Demography and Education Do Not Cause Interregional Disparities in Lebanon 120 mean per capita household consumption 100 as % of richest mohafazat 80 60 40 20 0 Beirut Mt. Lebanon North Bekaa and Dahyieh Lebanon real life standardized demography standardized demography and education Source: World Bank staff calculations, from 2002 National Household Survey. relocate in Lebanon are more limited than in other countries for socio- cultural reasons, even though the distances involved are quite small. Another possibility is that congestion and the high cost of living and housing in Beirut is encouraging high-potential households to stay in the provinces. One can divide Lebanon into a center (Beirut and Mount Lebanon) and a periphery (the north, south, Bekaa, and Nabatieh). The apparent gap between the center and the periphery is big enough--a 63 percent difference in per capita household consumption (UNDP 2008)--but the cost of living in the center is 20 percent higher than in the periphery. Adjust for that, and the gap falls to 36 percent. And the periphery has a lower percentage of its population economically active. Beirut, for example, has a 48 percent higher labor force participation rate than Nabatieh. Adjust for that, and per capita consumption per economically active person is a trivial 3 percent. The implication is that a spatial development strategy for Lebanon's periphery should lean toward improving labor mobility and increasing lagging areas' labor force participation rate. Disparities in earnings per worker, education, and family size, however, are not the major issues. Morocco's interprovincial spatial disparities are modest (figure 3.12): the coefficient of variation of provincial mean per capita household 90 Poor Places, Thriving People FIGURE 3.12 Interprovincial Disparities Are Quite Modest and Not the Fault of Demography in Morocco 100 90 mean per capita household 80 consumption as % of richest province 70 60 50 40 30 20 10 0 a r aa e a n l il n l uz te ta ila ae nc an igh ua af se na ao ar ien Az r-Z s-T bla Hs lem to rd -D ou IH Té Or ou dla ua ne sa sa ni Ta Ch Bou I H ift-A e- mm Be -O ek Ca as Ta a- ng uia M s- a- eim s-M G. ns Ze Ta Fč rd ao Te oc alé ra us Ch So t-S b- -A ba ar za Ra Gh Ta real life standardized demography standardized demography and education Source: World Bank staff calculations, from 1998 household survey. Note: Guelmime Es-Semara is not inculded. consumption is only 0.20.8 However, Morocco's urban-rural gap9 is quite large: 54 percent of mean urban per capita household consump- tion compared with 48 percent in Egypt and 40 percent in the Repub- lic of Yemen. The education of the household head and demography are factors to be considered but of less importance than in some of the other countries under consideration. Rabat-Salé Zemmour-Zaer, Fčs- Boulemane, Meknes-Tafil, and Tensift-Al Haouz have practically the same demographic composition as the leading area of Grand Casablanca, and elimination of the demographic disparities would not bring any province more than 9 percentage points closer to Casablanca's per capita consumption. Even if rural and urban areas had the same demographic characteristics, the rural-urban gap would still be 41 percent. Controlling for the level of education of the household head, rural Morocco's per capita household consumption would still be 32 percent lower than urban Morocco's. So Morocco is a case of significant rural-urban differences in economic opportunity, which household characteristics cannot explain away (see figure 3.13). In Syria, the issue of the north and northeast regions weighs heav- ily upon policy makers. Average per capita household expenditure in the northeast governorates of El Hassakeh, El Quneitra, and Al Raqqa Diagnose to Prescribe: Uncovering the Truth about Spatial Disparities 91 FIGURE 3.13 The Rural-Urban Gap Is Large and Cannot Be Explained Away by Demography or Education in Morocco, 1998 60 54 50 41 rural-urban gap (%) 40 32 30 20 10 0 real life standardized standardized demography demography and education Source: World Bank staff calculations, from 1998 household survey. is 45­53 percent lower than in Damascus City. Poverty is particularly high among the Kurdish population, which is concentrated in the north and northeast (UNHCR 2009) and includes some 300,000 stateless or unregistered persons (Lowe 2006). The northeast is also home to some of the poorest Iraqis to settle in Syria. Partly in order to support living standards and employment in the northeast, the government has maintained subsidies for cotton prices worth 0.9 per- cent of GDP (World Bank 2008b). Three years of drought, starting in 2006­07, have spurred migration out of the northeast toward Damascus. So what can the analysis tell us about the plight of Syria's lagging areas? It is striking that Syria's intergovernorate disparities can in no way be linked to spatial disparities in the level of education of the household head (figure 3.14); when one simulates the equalization of education levels across governorates, it makes no difference to per capita house- hold consumption. Even when one compares the urban population with the rural population, disparities in the level of education of the house- hold head do not count for much. Controlling for the education vari- able reduces the urban-rural gap only by 4 percentage points, which compares with a reduction of 7 to 11 percentage points in the other countries in the sample. This difference could be because educational attainment is evenly spread across the territory or because the returns to educational attainment are low due to low wage differentials. However, demographic disparities do matter. Outside Damascus, Syria's governorates can be divided into two groups. There are the 92 Poor Places, Thriving People FIGURE 3.14 Lagging Governorates Are a Demographic Phenomenon; Education Is Not the Issue in the Syrian Arab Republic 1.2 mean per capita household consumption as % of richest province 1.0 0.8 0.6 0.4 0.2 0.0 y us kia ms a ur ma s a po eh eb a a cit cu aid ra itr qq zo rto ep ak Idl ta Ho Ha Da as ne Ra us Ez Su ss La Al Ta am Qu sc ir Ha Al El De ma lD El El ra Da Ru real life standardized demography standardized demography and education Source: World Bank staff calculations from household survey, 2004. governorates of the central band (Tartous, Latakia, Homs, El Suaida, and Aleppo), which have a nearly identical demographic makeup to Damas- cus City, even though they lag 20 to 40 percent behind Damascus City in terms of per capita consumption. But then there is a second set of gov- ernorates to the north and south of this central belt (Deir Ezzour, Hama, Rural Damascus, Daraa, El Hassakeh, El Quneitra, and Al Raqqa) where demography is a major source of lagging area status. This group of gov- ernorates includes the northeast region. In fact, the demography factor is so strong that without it the northeast would not be a lagging area at all. If all of Syria's governorates had the same demographic profile (figure 3.14), El Hassakeh would be no worse off than Tartous, the leading governorate outside Damascus City--and significantly better off than the port city of Latakia and the urban provinces of Homs, Hama, Aleppo, and El Suaida. The northeastern lag- ging governorates of El Quneitra and Al Raqqa would be on a par with the leading governorates of Latakia and Homs. The northeast, therefore, may be a lagging region, but it is not nec- essarily a region lacking in economic opportunity. Its households have a demographic composition that would fare badly wherever it was Diagnose to Prescribe: Uncovering the Truth about Spatial Disparities 93 located (and household heads have a level of education similar to that in other areas). In the Republic of Yemen (figure 3.15), demography is worth a 15 percentage-point handicap, on average, for governorates outside Sana'a City. The demographic disadvantage is particularly strong (20 percentage points) for the Sana'a Region governorate (Sana'a City's peri-urban and rural fringe), for Al-Maharh, and for Sa'adah (the location of a religious insurgency and approximately 80,000 internally displaced people). Without the demographic and educational components of inter- governorate disparities, the poorest governorate, Amran, would only be 40 percent behind Sana'a City in terms of per capita consumption, instead of 65 percent. The vast, arid, and sparsely populated gover- norate of Al-Maharh would be 17 percent richer than Sana'a City instead of 20 percent poorer; thanks to its position on the Oman bor- der, it offers its inhabitants more economic opportunity than any other governorate. Without differences in education and demography, the rural-urban gap would be 25 percent instead of 40 percent. FIGURE 3.15 Demography and Education Explain about a Third of Intergover- norate and Rural-Urban Inequality in Republic of Yemen 1.4 mean per capita household consumption as % of richest province 1.2 1.0 0.8 0.6 0.4 0.2 0.0 -M ity h Dh an M ar eb b dr ion Al T t od z Sa ida Re ah Al mah ab e Al H ah ah jja Al eet Ab a La n Al heg n Am f ou aw -H ai ar 'a lb Sh hal aid ya ra am Al 'a C Ad ar 'ad w -M a Ha Reg e ah am w -J -D -B na Sa na Sa real life standardized demography standardized demography and education Source: World Bank staff calculations from 2005 household survey. 94 Poor Places, Thriving People The implication for MENA's policy makers is clear: before assuming that observed spatial disparities in welfare are the fault of spatial disparities in economic opportunity, one should check whether the households in the lagging areas are actually able to take advantage of economic oppor- tunities. As the data previously have shown, the situation varies from coun- try to country. For many lagging areas, however, the population of lagging areas consists of households who would have low per capita consumption wherever they lived--because of their household composition or their edu- cational level. This observation means that human development is a key component of effective spatial development in MENA. Where Does Lagging Areas' Demographic Disadvantage Come From? Why do MENA's lagging areas have household characteristics that are so different from those of leading areas? The differences can be broken down into three effects: · The fertility effect--Women in lagging areas have more children. · The sorting effect--People with high potential are the most likely to move out of low-potential areas. · The education effect--Children in lagging areas have less opportunity to acquire human capital. The fertility effect comes from MENA's demographic transition being spa- tially uneven. Fertility rates are falling across MENA, but they fell in lead- ing and urban areas first and in lagging and rural areas last. Lagging areas with a history of higher fertility have higher dependency ratios now. Figure 3.16 tells the story of the Islamic Republic of Iran's demo- graphic transition. In 1986, the rural total fertility rate10 was 1.1 points higher than the urban rate. (In other words, a typical rural woman was having 1.1 more children over her lifetime than a typical urban woman.) By 1990, the gap had widened to 1.7 points, as the urban rate fell faster than the rural rate. By 2000, however, the rural rate had caught up, and the gap had closed to 0.4 points. The spatially uneven demographic transition has been similar in other MENA economies. The rural-urban fertility rate gap in Tunisia closed from 2.3 in 1988 to 0.8 in 2006. In Jordan, the gap was 1.7 in 1990 but had practically vanished at 0.1 by 2007. In Egypt, the rural-urban fertil- ity gap suddenly began to shrink around 1992; since then, it has fallen from 2.0 to 0.7. Egypt had massive spatial disparities in contraceptive use in the mid-1990s. The highest rate of use was in urban Lower Egypt, while the rate in rural Lower Egypt (53 percent) was more than twice as high as in rural Upper Egypt (24 percent) (Chelala 1996). In Morocco Diagnose to Prescribe: Uncovering the Truth about Spatial Disparities 95 FIGURE 3.16 The Islamic Republic of Iran's Demographic Transition Began in Urban Areas 8 7 6 total fertility rate 5 4 3 2 1 0 86 88 90 92 94 96 98 00 19 19 19 19 19 19 19 20 year urban rural Source: Sutton 2001. the fertility rate gap was widening until 1992, when it hit a massive 3.0, but it was as low as 0.9 in 2003. In Algeria, the fall in fertility spread spatially like a wave, starting in coastal and urban wilayate (governorates) and only coming later to the high plateaus and then to the Saharan and Saharan Atlas wilaya (Sutton 2001). The result was that "interior and Saharan Algeria contain regional pools of demographic vitality and potential population growth" (map 3.3); according to the 1995 house- hold survey, the ratio of people of nonworking age to people of working age was 25 percent higher in rural areas than in urban areas (World Bank 1999). In the Republic of Yemen, however, the trend is not yet clear. The gap narrowed from 2.6 to 2.0 between 1991 and 1997, but then widened again to 2.2 by 2006. This "radiating demographic transition" conveys both bad and good news. In the short run, it is the cause of spatial disparities in dependency ratios that, as we have seen, are a major cause of MENA's rural-urban and interprovincial disparities in well-being. As the demographic transition reaches MENA's lagging areas, however, it will become a major long-term force for spatial convergence in living standards. Therefore, it is possible that a significant fraction of the spatial disparities "problem" may be already in the process of mitigating itself. Fertility rates are known to be influenced by many factors: female education, gender empowerment, female labor participation, and income growth, as well as access to quality reproductive health services (World Bank 2007b). All of these determinants of fertility vary from place to place, which is precisely why the demographic transition has been 96 Poor Places, Thriving People MAP 3.3 Interior and Saharan Algeria Contain Regional Pools of Demographic Vitality and Potential Population Growth IBRD 38171 NOVEMBER 2010 BIRTHS PER THOUSAND, 1998: 9.40 ­ 17.00 17.00 ­ 19.00 19.00 ­ 21.55 21.55 ­ 25.00 25.0 ­ 28.40 This map was produced by the Map Design Unit of The World Bank. 0 250 500 The boundaries, colors, denominations and any other information shown on this map do not imply, on the part of The World Bank Group, any judgment on the legal status of any territory, or any KILOMETERS endorsement or acceptance of such boundaries. 0 50 100 150 200 250 KILOMETERS ANNABA ALGER JIJEL BOUIRA CONSTANTINE CHLEF AIN DEFLA MEDEA TISSEMSILT ORAN MASCARA TIARET SAIDA NAAMA Source: Sutton 2001. geographically uneven. For example, in Tunisia (Ayadi and Amara 2009) the relationship between household size and poverty is strongest in coastal (leading) areas, suggesting that the motivation to limit one's num- ber of children might be highest there. As MENA governments develop population policies, they could, therefore, focus on the spatial dimension of fertility (see map 3.3). The sorting effect means that people with the highest earning opportunities leave lagging areas. This is because people with more education set greater store by city amenities, and because the jobs that make education productive tend to be concentrated in cities (Henderson 1986). In Morocco in the 1980s, people with secondary or higher education were 120 percent more common among rural- urban migrants than among the rural population overall (Centre d'Études et de Recherches Démographiques 1989). Lifetime rural- urban migration rates in Egypt are twice as high for university grad- uates as for people without intermediate-level education (Wahba 2007). People with more education, moreover, tend not to migrate out of large cities (McCormick and Wahba 2005). In Syria, a male Diagnose to Prescribe: Uncovering the Truth about Spatial Disparities 97 with secondary education is twice as likely to migrate as one with less than primary education (Khawaja 2002). The education effect is about people in lagging areas being less able to acquire human capital, whether from the educational system or from family members and social networks. Given the centrality of education for spatial development, therefore, later sections will discuss the spatial dimension of education and how MENA governments can use education as an instrument for achieving their spatial development objectives. The three household characteristic effects--fertility, sorting, and education--are interrelated, inseparable, and mutually reinforcing. Women who have received less primary and secondary education tend to have more children at any given level of household income (Akin 2005). Educated people who leave the lagging area take their education with them, rather than transmitting their human capital to the lagging area's next generation. Those who remain behind are more likely to be female and less educated--that is, the section of the population most associated with child-rearing. Developing People to Take Opportunities The conclusion, then, is that lagging areas' poor economic situation is linked to household characteristics. This point is of critical importance for policy makers. It means that stimulating economic development in MENA's lagging areas is as much a question of developing people to take opportunities as of developing places to make opportunities. Delivering quality education in lagging areas is critical, not just because educated people can take more opportunities, but also because education is a key factor in the demographic transition toward lower fertility rates and lower dependency ratios in lagging areas. What Are the Differences in the Cost of Living? It costs less to live in lagging areas than in leading areas. Housing is less expensive, and food can sometimes be cheaper too. When measuring spatial disparities in living standards, therefore, it is important to allow for differences in the cost of living. One way of gauging spatial disparities in the cost of living is to look at the difference between urban and rural poverty lines, which is the cost of maintaining a minimum standard of living for one person. However, one cannot assume that the rural-urban gap in the average cost of living will be exactly the same as the rural-urban gap in the poverty line, because poor people's spending has a distinct composition. In MENA, the urban poverty line averages 10 percent higher than the rural poverty line (table 3.9). This is a smaller gap than for most other 98 Poor Places, Thriving People TABLE 3.9 Urban-Rural Differences in the Cost of Living Partially Cancel Out Spatial Disparities in Spending Power Country or region Year of survey(s) Ratio of urban to rural poverty lines Middle East and North Africa 1.10 Egypt, Arab Rep. 1995, 1999/00 1.09 Iran, Islamic Rep. 1994, 1999 1.13 Jordan 2002­03 1.13 Morocco 1990­91, 1998­99 1.29 Tunisia 1995, 2000 1.18 East Asia and the Pacific 1.30 Europe and Central Asia 1.05 Latin America and the Caribbean 1.44 South Asia 1.30 Sub-Saharan Africa 1.29 Source: Ravallion, Chen, and Sangraula 2007. regions; however, it still cancels out a large part of the apparent urban- rural disparity in living standards. For example, per capita expenditure in urban Morocco in 1998­99 was 54 percent higher than per capita expen- diture in rural Morocco, but the poverty line in urban Morocco was 29 percent higher than in rural Morocco. In Jordan, the urban poverty line in 2002 was 13 percent higher than the rural poverty line. This gap is actually higher than what the urban-rural gap in per capita per con- sumption would be if rural and urban Jordan shared the same demo- graphic characteristics. Moreover, spatial disparities in demography could reinforce spatial disparities in prices, if the cost of living for a dependent is lower than for a worker--for example, in terms of food, travel, and clothing expenses. Spatial Disparities Are Not as Wide as They Seem The implication of all the earlier discussion is that the political econ- omy of space, sorting effects, fertility effects, education effects, and price differences all combine to magnify the appearance of spatial inequalities in economic opportunity. A key recommendation of this report is, therefore, to check that inequalities in economic opportunity are really as big as one imagines. Does Spatial Convergence Come in Many Qualities? Once policy makers have a clear profile of spatial disparities at a moment in time, it would be useful to know how they are changing across time. The crucial question is, are living standards converging spatially? If they Diagnose to Prescribe: Uncovering the Truth about Spatial Disparities 99 are, governments may decide that it is not necessary to focus resources on resolving an issue that is on its way to a solution. The 2009 World Development Report (World Bank 2009a) shows how spatial disparities follow a general pattern as countries experience economic growth: · The living standards gap between leading and lagging areas first grows, as industrialization gathers pace, and then shrinks. The ini- tial divergence happens as economies of agglomeration concentrate economic activity in leading areas and cities. The subsequent con- vergence results from migration out of lagging areas, making labor scarcer and pushing up wages, from investments spilling out of con- gested cities into peri-urban areas and from greater spending on public social services in lagging areas (2009a, 62). The same goes for disparities between subnational regions and rural-urban disparities (2009a, 65). The divergence-convergence cycle can be a long-term process; countries such as Sweden, Spain, the United Kingdom, the United States, and Japan took 40 to 100 years for the income gap to rise and fall back to preindustrialization levels (2009a, 86). But the trends are not always smooth or simple. Massive divergence can hap- pen very quickly, as it did in India, the Russian Federation, and Vietnam in the 1990s (Shankar and Shah 2003). Countries can converge and then diverge, as China did, or can diverge and then converge, like Brazil. · It is possible for disparities in social indicators, such as poverty, health, and education, to shrink at the same time that economic agglomeration is growing. Reducing social spatial disparities depends on the state; it requires a progressive taxation system and for lagging areas to receive quality social services (2009a, 238). The MENA examples below are specimens of the kind of convergence narrative that could inform policy decisions. They examine the Islamic Republic of Iran, Tunisia, and Egypt. The lesson that our analysis of these countries teaches is that, while the divergence-convergence cycle might be a useful generalization, every country's convergence experience has its own quality. A simple statement that development is converging or diverging across space does not capture the potential variety of trends required to inform policy makers' decisions. The Islamic Republic of Iran: Spatial Convergence in Living Stan- dards, but for the Wrong Reason. The Islamic Republic of Iran has enjoyed inter-ostan (province) spatial convergence in spending power since the Revolution. The population-weighted11 coefficient of variation 100 Poor Places, Thriving People of mean per capita household expenditure was 0.40 in 1976, about the same as in India. By 2005, it was 0.28. Underlying the inter-ostan con- vergence was an urban-rural convergence. In 1976, the urban-rural gap in per capita household expenditure was 53 percent. By 2005, the gap was down to 37 percent. However, the story of living standards convergence is not as happy as it seems. The leap forward in convergence, both inter-ostan and rural- urban, occurred from 1976 to 1996 (figure 3.17), when the economy was affected by conflict, urban private assets were expropriated (Behdad 1989), and GDP per capita averaged 34 percent lower than its 1976 peak. The urban-rural convergence came from urban per capita household expenditure falling by 21 percent from 1976 to 1985, while rural per capita household expenditure "only" fell by 7 percent. In fact, the reduc- tion in inequality came with an increase in overall poverty (Assadzadeh and Paul 2004). The rural sector also benefited from the revolutionary government's decision to adjust the terms of trade in favor of rural areas. Between 1976 and 1984, agricultural value added grew by 31 percent, twice the rate of the non-oil economy (Behdad 1989). One sure reason was that farmgate prices rose 55 percent in real terms from 1977 to 1985 (Mojtahed and Esfahani 1989). But once the overall economy began to resume steady growth, from the early 1990s onward, the spatial convergence trend changed. FIGURE 3.17 Inter-Ostan Disparities Shrink over Time in the Islamic Republic of Iran 10.0 population-weighted coefficient of 0.9 household consumption by ostan variation of mean per capita 0.8 0.7 0.6 0.5 0.4 0.3 0.2 0.1 0.0 on 1999 ra 998 Br 997 an 97 95 ic 1997 Isl Rep 6 5 Isl stan 6 Re 99 nia 5 96 7 7 98 9 0 99 9 19 Ira amic . 19 19 19 Ro . 20 19 Fe sia 1 n1 Sr il 1 .1 n1 ina ka Isl ndia Uz ep. Isl Rep p tio az ta e Ch R ma kis I ki iL ic ic de be Pa Ind am am am ian n, n, n, n, ss Ira Ira Ira Ru countries Sources: World Bank staff calculations using government of the Islamic Republic of Iran unpublished data, http://www.unescap.org/esid/psis/population/popin/profiles/iran/popin15.htm; Shankar and Shah (2003). Diagnose to Prescribe: Uncovering the Truth about Spatial Disparities 101 Inter-ostan convergence stagnated, and the urban-rural gap began to widen once again, from 29 percent in 1996 to 37 percent in 2005-- when the rural-divide was a major theme of political campaigning in the run-up to the presidential election. In the Islamic Republic of Iran, therefore, the story is that spatial divergence can be part of a positive growth process. Tunisia: A Case of Bipolar Convergence. Tunisia has seen significant interregional convergence. The population-weighted coefficient of vari- ance of median household consumption by region fell from 0.32 in 1980 to 0.24 in 2000. This fall brought Tunisia's interregional disparities down into the Western European range. There has been an identifiable trend for governorates with higher poverty rates to enjoy higher growth rates (Boulila, Gabsi, and Trabelsi 2009). Although the rural-urban gap also narrowed over the same period, the convergence trend is not clear. The ratio of "communal" to rural mean per capita household expenditures was 2.11 to 1.00 in 1980 and 1.85 to 1.00 in 2000. The rural-urban gap, however, fluctuates from year to year, probably according to rainfall; in 1975, it was as low as 1.81, even slightly lower than in 2000, so one cannot be totally certain of a converging trend. A critical feature of Tunisia's spatial convergence is the trajectory of the middle-ranking regions, namely the center-east, the northeast, and the south. In 1980, these three regions had more or less the same stan- dard of living as each other, sandwiched between the prosperous Greater Tunis region and the lagging center-west and northwest regions. Since 1980, however, each of the middle-ranking regions has taken its own course. On the one hand, living standards in the center-east have shot up to converge with Greater Tunis. On the other, the south and the north- east have stagnated, in relative terms, to converge with the center-west and the northwest (figure 3.18). Zooming in, we see that the center-east's success was focused on a breakout by Mahdia and Monastir governorates (figure 3.19). Monastir's mean per capita household consumption in 1980 was 31 percent lower than Tunis's and around the same as that of its neighbors. By 2000, however, it had split away from its neighbors, even those closer to the metropolis, and joined Tunis. There are important policy lessons in Tunisia's regional convergence experience. First, economic agglomeration is a more selective process than simple agglomeration theory suggests (that is, economic density leading to more economic density). Clearly, Tunisia's center-east region built on some kind of potential competitive advantage, despite being fur- ther from the Tunis metropolis than the equally rich northeast area. The implication is that distance from economic density is not the whole story 102 Poor Places, Thriving People FIGURE 3.18 Regions Split into Rich and Poor Clubs, While Converging Overall in Tunisia 160 150 regional median household expenditure as % of 140 130 national expenditure 120 110 100 90 80 70 60 1980 1990 2000 year Tunis north east northwest central-west central-east south Sources: INS 2000, 1990, 1980: Enquęte nationale sur le budget, la consommation, et le niveau de vie des ménages [National survey of household budgets, consumption, and living standards]. FIGURE 3.19 Proximity of Eastern Governorates to the Metropolis Does Not Explain Performance in Tunisia, 1980­2000 330 Mahdia (CE) 300 nominal mean per capita household Monastir (CE) expenditure growth 1980­90 (%) 290 270 250 230 Bizerte (NE) 210 Nabeul (NE) 190 Tunis Zaghouan (NE) Sousse (CE) 170 Sfax (CE) 150 130 0 50 100 150 200 250 300 air kilometers from Tunis Source: INS 2000, 1980, Enquęte nationale sur le budget, la consommation et le niveau de vie des ménages [National survey of household budgets, consumption, and living standards]. Note: CE = center-east, NE = northeast. Diagnose to Prescribe: Uncovering the Truth about Spatial Disparities 103 when it comes to predicting an area's success; effective spatial develop- ment strategies can tap other sources of development potential. This is a theme that we will discuss further in later sections. Second, spatial con- vergence at the national level may take the form of a "breakout" by middle-ranking areas. Therefore, policy makers should not focus their convergence concerns solely upon the most lagging areas. The Arab Republic of Egypt: Agglomeration of Poverty within the South. The rural-urban gap has been widening in Egypt (table 3.10). Between 2005 and 2008, the fastest per capita household consumption growth was in metropolitan, Upper Egypt urban, and Lower Egypt urban areas. The fastest growth, however, at 5.6 percent per annum, was in the urban areas of the poorer Upper Egypt governorates. Overall, it is not possible to identify clear convergence or divergence across governorates. The population-weighted coefficient of variation of the gross regional domes- tic product (GRDP) estimates published by the United Nations Develop- ment Programme's (UNDP's) national Human Development Reports was unchanged at 0.39 between 1998­2009 and 2003­04.12 However, there is a growing spatial concentration of poverty over time. Between 1995 and 2008, the number of Egyptians living below the poverty line rose by 3,070,000, with an increase of 3,796,000 in the gov- ernorates of Upper Egypt. What seems to be happening is that Egypt is facing a situation of widening spatial disparities at the micro level within Upper Egypt, even if there is no overall divergence at the national level. Not only is the gap between urban and rural Upper Egypt widening, but also some villages in rural Upper Egypt are being left behind while others "break out." As figure 3.20 shows, poverty reduction in Upper Egypt's villages was a two-speed affair. In 1996, most villages had poverty rates in the 50 to 75 percent band. Some villages broke away, reducing their poverty rates to 10 to 20 percent. But an equal number of villages remained stuck with TABLE 3.10 The Arab Republic of Egypt, 2005­08: Urban Convergence with Rural Divergence Annual growth in mean real household consumption, 2005­08 Urban (%) Rural (%) Metropolitan 3.3 n.a. Lower Egypt 4.2 3.1 Upper Egypt 5.6 1.6 Source: World Bank 2009b. Note: n.a. = not applicable. 104 Poor Places, Thriving People FIGURE 3.20 Two-Speed Development in Rural Upper Egypt, 1996­2006 a. Lower rural Egypt b. Upper rural Egypt 6 4 frequency 2 0 0.0 0.5 1.0 0.0 0.5 1.0 poverty rate (1 = 100%) kdensity P0_96 kdensity P0_06 Source: World Bank 2009b. their poverty rates in the 40 to 60 percent range. The Egyptian govern- ment's strategy of targeting assistance toward low-income villages in low-income governorates does indeed fit the evidence on the spatial dis- tribution of poverty. Convergence in Human Development Indicators: A Mirage? When it comes to social indicators such as health, education, and poverty outcomes, appearances of spatial convergence can be deceptive. As long as the national picture is improving, spatial convergence in basic social indicators will usually be observed. When the national picture is deteri- orating (as in the poverty case for Egypt 1995­2008 mentioned earlier) divergence may be noted; however, this change might be more to do with the choice of indicator than with spatial patterns in human development. For example, the Islamic Republic of Iran's human development indicators have improved and converged spatially during recent decades. Between 1976 and 1996, the female literacy rate rose 45 per- centage points from 17 percent to 62 percent, while for urban women it rose by only 26 percentage points from 56 percent to 82 percent (UNDP 1999). During 1994­2000, infant and under-five mortality rates fell fastest in the poorest provinces (Mehryar 2004). Rates of stunting, wasting, and underweight fell in all provinces, with some of Diagnose to Prescribe: Uncovering the Truth about Spatial Disparities 105 the biggest gains in lagging ostans. Inter-ostan inequalities in basic education have decreased over time; in the 10 years to 2006, the gap between the highest and the lowest literacy rates narrowed by 10 per- centage points (World Bank 2008c). However, such indicators of basic social welfare do not tell the full story. Poverty rates, malnutrition rates, illiteracy rates, and so on can hardly fail to converge if the overall human development situation is improv- ing, because it is impossible for the rates in leading areas to fall below zero. In the United Kingdom, for example, spatial disparities in mater- nal mortality are minuscule because they are close to zero percent everywhere. The United Kingdom still has significant spatial disparities in health, however. For example, a child from Wales is 77 percent more likely to be overweight than a child from the East of England (Hawkins and others 2007). The United Kingdom still has spatial disparities in health, but indicators of basic social welfare achievement no longer pick them up. So it may be with the Islamic Republic of Iran and other MENA economies. Convergence of basic social welfare indicators may be taking place, but this does not mean that other social indicators are not showing wide--and may be increasing--disparities. Understanding the potentially misleading convergence conclusions to be drawn from the standard basic social indicators is particularly important in MENA economies where the social service agenda is a rapidly advancing frontier: from primary school enrollment toward postprimary access and quality in education (World Bank 2008c, 83­114);13 from the communicable toward noncommunicable disease; from water connections toward water and sanitation service quality; and from extreme poverty toward borderline poverty and vulnerability (Iqbal 2006). Is Internal Migration Resolving MENA's Lagging Area Problem? Internal migration should reduce spatial disparities, as several processes come into play: · When people move from lagging areas to leading areas, they reduce the supply of labor in their place of origin, causing earnings per head to rise in lagging areas. · Migrants send remittances back home. For example, in Sub-Saharan Africa remittances account for 15 percent of rural income, and in China, migrant-sending households become richer in terms of con- sumption, production, and ownership of consumer durables (De Brauw and Giles 2006). 106 Poor Places, Thriving People · Prolonged migration leads to a reduction in the importance of the lagging area's development needs, for the simple reason that fewer people live in the lagging area. The 2009 World Development Report (World Bank 2009a, 146­69) classifies countries into two groups. On one side, there are countries in which the movement of labor from lagging to leading areas can bring about spatial convergence in living standards. The United States, with its northward wave of African American migration after the Civil War and its continuing tradition of long-distance relocation is one example. China, where migrants account for a third of urban employment, is another. On the other side are countries where the populations of lag- ging areas cannot or will not migrate to leading areas, perhaps because of social, ethnic, religious, or linguistic divisions. Although 40 percent of India's urban population are migrants, for example, only 4 percent of migrants cross state lines. This finding means that the migration solution is not going to resolve the lagging area problem for lagging states such as Bihar, Orissa, or Madhya Pradesh. The appropriate policy response to the lagging areas problem depends on which category a lagging area falls into. The more the pop- ulation is locked into its current spatial distribution, the more spatially targeted economic development is required to improve economic oppor- tunities in lagging areas. Note the difference between changes in the spatial distribution of pop- ulation and the propensity to migrate. A high propensity to migrate is a necessary but insufficient condition for a redistribution of the population from lagging to leading areas. Many migrants, both men and women, practice seasonal or temporary "circular" migration; they migrate, but their long-term domicile does not change. Spatial disparities in the fer- tility rate can also cancel out migration. And migration into and out of a lagging area can cancel itself out. In the Islamic Republic of Iran, for example, gross provincial outmigration numbers 1986­96 were seven times larger than net interprovincial migration figures. Where, then, does MENA lie? Is its spatial distribution of population fluid or fixed? Urbanization, but Not Metropolitanization All over MENA, the center of demographic gravity is moving from the countryside to the cities, with rapid increases in the urban sector's share of population. Overall population growth, however, is still fast enough to keep MENA's rural populations growing. Here the exception is Lebanon, whose rural population peaked pre-1960, reached a nadir around 1990, and has been creeping back up ever since. Diagnose to Prescribe: Uncovering the Truth about Spatial Disparities 107 Despite the popular view that there is a headlong rush of migrants to the capital cities and main metropolitan areas, official data show that in Algeria, Egypt, the Islamic Republic of Iran, Morocco, and Syria the main city's share of total population has stayed more or less constant. Only Tripoli in Libya and Sana'a in the Republic of Yemen have significantly increased their shares of the national population, while Amman and Beirut have lost population share. Most main cities have also seen their share of the urban population fall. Therefore, urbanization is spread across entire territories, and is not, as is popularly believed, taking the form of a unipolar concentration of pop- ulation in each country's biggest city or cities. MENA's observation is con- sistent with the worldwide observation (Henderson, Shalizi, and Venables 2001) that the relative size of cities changes little as economic growth occurs. In North Africa in 1960, 25 percent of the urban population was living in cities of more than 1 million inhabitants, a figure that had risen to only 33 percent by 2005. In the Islamic Republic of Iran, the share rose from 26 percent to 33 percent over the same period; and in Iraq, from 32 percent to 35 percent. Meanwhile, Amman's share of the urban pop- ulation fell from 48 percent to 24 percent, Beirut's from 70 percent to 51 percent, and Damascus's from 34 percent to 23 percent. Sana'a was the exception, with its share of the urban population doubling to 30 percent (United Nations Department of Economic and Social Affairs 2007). MENA's urbanization is on a par with that of other developing country economies and not, as is sometimes feared, a phenomenon of uniquely overgrown megacities. Henderson (2000) estimated the relationship between growth and the size of the "prime" (main) city for a sample of countries using 1965­95 panel data. Of the 70 prime cities for which results were available, 24 appeared so large as to be restricting economic growth. Only one of these was from the MENA region: Cairo. The other MENA cities in the study, Algiers, Tehran, Amman, Damascus, and Tunis, were among the 30 prime cities that came out as being appropriately sized. Table 3.11, therefore, summarizes the trend of the past three decades with respect to the urban-rural continuum: urbanization, but without an emptying of rural areas and without a concentration of the urban popu- lation within prime cities. Very Little Lateral Redistribution toward Leading Provinces The redistribution of MENA's population tends to be a gradual shift from rural to urban areas within provinces rather than a lateral agglomeration from lagging provinces to leading provinces. This shift is very different from what happened in the United States in the late nineteenth century or in China since the 1990s, where there were massive lateral redistribu- tions of the population across long distances. To visualize MENA's case, 108 Poor Places, Thriving People TABLE 3.11 Urbanization without Deruralization or Metropolitanization Increase in Increase in Increase in % Decrease biggest city's biggest city's urban % of in rural % of total % of urban 1980­2007 population population population population Algeria 22 ­13 1 ­5 Djibouti 15 ­15 -- -- Egypt, Arab Rep. ­1 ­77 ­1 ­2 Iraq -- -- -- -- Iran, Islamic Rep. 18 ­16 ­2 ­10 Jordan 18 ­41 ­5 ­18 Lebanon 13 26 ­13 ­27 Libya 7 ­52 10 9 Morocco 15 ­20 ­1 ­8 Syrian Arab Republic 7 ­92 2 0 Tunisia 15 ­10 -- -- West Bank and Gaza 10 -- -- -- Yemen, Rep. 14 ­124 6 12 Sources: Staff calculations; World Bank GDI database. Note: Gray = An indicator of agglomerating population redistribution. Brown = An indicator of nonag- glomerating population redistribution. -- = data not available. one can extrapolate from the different observed population growth rates of a country's provinces across consecutive censuses to project how its population might be allocated across the territory in future (table 3.12). The resulting scenarios are not a forecast in the strict sense because they do not reflect future changes in migration numbers or the fertility rate; however, they do give a clear sense of the current stasis in the inter- provincial distribution of MENA's population. Although many studies of international migration from MENA exist, analysis of internal migration flows is scarce and mostly focused on Wahba and Assaad's work in Egypt. What research is available tends to confirm its noncentripetal character. Data from the Egypt Labor Market Panel Survey (Wahba 2007) show that the overall propensity to migrate is low; only 1.6 percent of the population changed its governorate of resi- dence during the eight years between the 1998 and 2006 surveys. The life- time intergovernorate migration rate is significantly lower for men than for women--7 percent compared with 11 percent--suggesting that much intergovernorate migration is determined by marriage patterns rather than by opportunities of higher-paying employment. Moreover, migration flows between better- and worse-off areas can cancel each other out; during 1990­98, 0.2 percent of the population moved from rural to urban areas, while 1.1 percent of the population relocated in the opposite direction (Wahba 2007). Syria's Internal Migration Survey Project (Khawaja 2002) Diagnose to Prescribe: Uncovering the Truth about Spatial Disparities 109 TABLE 3.12 Little Lateral Redistribution of Population toward Leading Provinces, Save in Tunisia In . . . Egypt, Arab Rep. the poorest . . . 13 provinces' share of 58.0 in 1996 to . . . 54.7 in 2036 Iraq 7 the population 37.7 in 1997 (%) 39.6 in 2037 Iran, Islamic Rep. 15 is heading from 36.0 in 1996 33.4 in 2036 Jordan 6 . . . (%) 32.2 in 1994 32.6 in 2036 Morocco 27 48.4 in 1994 43.6 in 2034 Syrian Arab Republic 7 45.4 in 1981 44.9 in 2021 Tunisia 12 39.0 in 1986 15.0 in 2026 Sources: Arab Republic of Egypt, National Human Development Report and censuses; Iraq, http://cosit.gov.iq/english/2007/15-13.htm and http://cosit.gov.iq/english/2007/2-3.htm (accessed July 7, 2010); Islamic Republic of Iran, National Household Survey, http://www.unescap .org/esid/psis/population/popin/profiles/iran/popin15.htm and http://www.statoids.com/uir.html (accessed January 23, 2008); Jordan, http://www.statoids.com/ujo.html (accessed April 24, 2009) and National Household Survey; Morocco, http://www.statoids.com/yma.html (accessed April 23, 2009) and World Bank 2004a; Syrian Arab Republic, http://www.undp.org.sy/publications/national/Poverty/ Poverty_In_Syria_en.pdf, http://www.statoids.com/usy.html (accessed April 24, 2009) and http://www.populstat.info/Asia/syriap.htm; Tunisia, INS 2000: Enquęte nationale sur le budget, la consommation et le niveau de vie des ménages [National survey of household budgets, consumption, and living standards]. found a lifetime migration rate of 14 percent, which is low by middle- income country standards. Again, there was no sense of migrants being particularly drawn toward the larger cities; Damascus, Idlib, and Hama all had a negative lifetime net migration rates (2002, 26), and provincial capitals had a lower share of lifetime migrants in them than smaller urban centers (2002, 17). In the Islamic Republic of Iran during 1996­2006, 36 percent of migratory moves between 1996 and 2006 were within the same district (shahrestan), and another 26 percent between different districts of the same province (ostan).14 MENA's internal migration patterns deserve more research. At this stage, one can only hypothesize as to why the redistribution of popula- tion from lagging to leading provinces is slow. The role of social networks may be critical. In the absence of trans- parent market information for jobs and housing, the opportunity to migrate toward economic opportunity may depend upon having friends and family at one's destination. From the employer's point of view, it is important to hire workers one can trust, especially if labor laws and slow legal procedures make it difficult to fire staff. (In India, World Bank enter- prise surveys reveal that employers believe it harder to find an employee they can trust than an employee with the right skills [Amin 2009].) This finding would mean that patterns of internal migration are dictated by social connections across space; the poorest may have the greatest desire to migrate, but the best connected have the greatest opportunity. The primacy of social networks in determining internal migration flows is illustrated by studies of Egypt (box 3.3). Setting aside the employees who found work through public channels (that is, essentially public sector jobs), 55 percent of employed workers said they found their jobs through friends 110 Poor Places, Thriving People BOX 3.3 Caffeine-Based Networking among Cairo Construction Craftsmen "A craftsman has to be recognized by his peers as an insider to be able to get work. This is typi- cally done through socializing at the coffeehouse. After long discussions with craftsmen and sub- contractors, I discovered that besides having their own jargon, craftsmen also had secret signs that they would not reveal to me as an outsider. There were clear indicators that there was an insid- ers' culture [into] which one had to be initiated in order to belong. "A coffeehouse is frequented by a stable group of craftsmen who usually live nearby and who rely largely on relationships they forge in the coffeehouse to get work. Most craftsmen agree that there is no point extending their job search beyond the local coffeehouse." Source: Assaad 1993, 935. and relatives (Wahba and Zenou 2005). Moreover, where one lives is critical in determining one's chances of using a network. People already residing in cities, where social networks are densest, are on the inside track; 60 percent of Cairenes used friends and relatives to find a job, compared with 40 percent of residents of rural Upper Egypt, and increasing popula- tion density by 10 percent increases the probability of using social networks by 2.2 percent. Some regions of origin are more propitious for job seekers than others. For construction labor, employers rely on ties of kinship and place of origin to determine who gets hired and who remains unemployed (Assaad 1993). In fact, one of the reasons that it is difficult to use national household survey data to analyze migration patterns is that a small propor- tion of villages provides a high proportion of migrants, so cluster samples estimate migration rates with huge margins of error. Map 3.4 shows how the networking effect plays out in the Republic of Yemen. Areas sending migrants can be identified from receipts of remit- tance income, which are clustered around three poles, one in the arid center of the country and two in the southwest. None of these three areas is exceptional in terms of household incomes or poverty. What appears to be happening is that migrant networks are centered on certain send- ing localities, irrespective of their levels of income. Are gender relations in MENA a brake on interprovincial migration? It is known that there are social constraints upon female mobility (Assaad 2005) and labor market participation (Khawaja and others 2009). Gender inequality plays a major role in shaping the consequences of migration and spatial disparities (Chant 1992). However, a study of gender and spatial disparities in MENA commis- sioned for this report (Esber 2008) did not find clear evidence that Diagnose to Prescribe: Uncovering the Truth about Spatial Disparities 111 MAP 3.4 Remittance Income Clusters in the Republic of Yemen MAJOR INCOME SOURCE: REMITTANCE, SHARE IBRD 38172 NOVEMBER 2010 46% ­ 5% No obs 6% ­ 15% 16% ­ 29% 30% ­ 45% 46% ­ 83% This map was produced by the Map Design Unit of The World Bank. The boundaries, colors, denominations and any other information 0 50 100 150 200 250 shown on this map do not imply, on the part of The World Bank Group, any judgment on the legal status of any territory, or any endorsement or acceptance of such boundaries. KILOMETERS Source: World Bank 2009c. women's social status was limiting migration. The women interviewed tended to view internal migration as an attractive strategy to improve their livelihoods. An Egyptian woman spoke for many when she said, "It makes no difference to me whether we stay here or anywhere; the most important thing to me is to find a job with a good income for my husband so that we, as a family, live a decent life." In Qena, Egypt, there was no significant differ- ence between men and women in their average estimate of the minimum required increase in household income to justify the decision to migrate to Cairo: LE 1,155 versus LE 1,167 per month. In Tunisia, women inter- viewed were nearly as supportive as men of their sons migrating in search of a better life: 81 percent of women and 91 percent of men surveyed said that migration was beneficial for the young. An oft-repeated phrase was "Your country is where you earn your bread." Women even looked more favorably than men, by a rate of 100 percent to 39 percent, upon their daughters' migrating: 81 percent of women said that they would actively encourage their daughters to migrate. 112 Poor Places, Thriving People TABLE 3.13 Lebanon's Population Distribution Becomes Less Agglomerated Share of national population 1970 (%) 1996 (%) 2004 (%) Central Beirut 22 10 10 Greater Beirut 44 32 35 Central Lebanon (Greater Beirut plus Mount Lebanon) 62 48 50 All Lebanon 100 100 100 Source: Nahas 2009. In Lebanon, a disproportionate rate of external emigration may have been a factor in limiting the redistribution of population toward leading provinces. The interprovincial distribution of the population deagglom- erated during the civil war (Nahas 2009), with Greater Beirut's share of the country's population falling and then stabilizing during the postwar period (table 3.13). This change was despite the fact that migration from the periphery to Beirut and its environs had boosted the center's popula- tions. In 2004, only 61 percent of the Lebanese population of Greater Beirut and Mount Lebanon had been born there, compared with the 93 percent of the peripheral areas' population (the north, the south, and Bekaa) that was native to those regions. The overall conclusion, therefore, is that the redistribution of popula- tion in MENA economies is urbanizing but not metropolitanizing. The process of agglomeration is taking place more within subnational regions than among them. There are important implications for MENA's spatial development strategies. First, one cannot overemphasize the importance of confronting preconceptions with robust subnational data. Even though political concerns may focus on the specter of an uncontrolled human tide flooding the prime cities, this is not the reality of internal migration in MENA, or at least not the whole reality. Second, provincial cities are crit- ical poles for lagging areas' demographic transformation, and should, therefore, be central to spatial development strategies. Is the Lagging Area a Fringe, Belt, or Pocket? Lagging areas come in many varieties. The governorate of Aleppo in Syria is a second lagging area. Its mean per capita household expenditure is 56 percent of that of the capital. It is home to a great conurbation of 4.4 million people--possibly the oldest continuously inhabited city in the world--and a number of major employers, including the headquarters of the national railways. It is surrounded by fertile farmland. Diagnose to Prescribe: Uncovering the Truth about Spatial Disparities 113 The northeastern governorate of Al-Mafraq in Jordan is a a second lagging area. Its mean per capita household expenditure is 51 percent of that of the capital, but it could hardly be more different from Aleppo. With a population density of only nine people per km2 and a yearly aver- age of 150 millimeters of rainfall, it represents MENA's vast swathes of sparsely populated dryland. Yet its provincial capital is only 50 kilometers from the national capital--a 45-minute drive--and sits on the intersec- tion of two four-lane highways, a crossroads between three capital cities: Amman, Baghdad, and Damascus. A third lagging area is the wilaya of Djelfa in Algeria, where the high plateau meets the Sahara. More than half of Djelfa's population older than six years of age is reported as being without education. The private car ownership rate is only a tenth of that in Algiers. It lies on the Trans- Saharan Highway. Unlike Mafraq, though, it is far from the nearest big city; Algiers is 300 kilometers across the mountains. Aleppo, Al-Mafraq, and Djelfa all have lagging development indica- tors in common, but their geographical characteristics set them apart from each other. Aleppo has economic and demographic density and diverse resources; Al-Mafraq and Djelfa do not. Al-Mafraq is close enough to a major agglomeration to benefit from spatial spillovers. Aleppo and Djelfa are not. If one is apply the correct mix of policies as suggested in figure 2.6 and table 2.3, it is, therefore, necessary to diagnose which kind of lagging area we are dealing with. Aleppo, Al-Mafraq, and Djelfa provide good specimens of three types: the fringe, the belt, and the pocket. · Djelfa looks like a "fringe" area. It is sparsely populated, it is distant from economic density, and its economic potential is constrained by the lack of agglomeration or a resource base for growth. If this diagnosis is correct, then Djelfa is a prime candidate for Policy Package 1 as out- lined in table 2.3: a level playing field for development in terms of access to health, education, social protection, and the business environment, so that its citizens can make the most of a very difficult geographic context. · Al-Mafraq appears as a "belt" area. It could benefit from the level play- ing field for development, but it is also close enough to benefit from good connectivity with the capital nearby, Policy Package 2 in table 2.3. · Aleppo seems to be a "pocket" of unrealized economic potential. It already has population density and economic agglomeration, and just the brief description above should indicate its potential as a location for tourism and agriculture-related growth. Aleppo may, therefore, be a candidate for Policy Package 3 in table 2.3, facilitating private sector interest in investment in lagging areas. 114 Poor Places, Thriving People MAP 3.5 A Policy-Oriented Spatial Typology of the Middle East and North Africa Region lagging dense sparse nonlagging population extremely low population within 3 hours of a major city Sources: Elevation, SRTM30 (2000); land cover, GLC2000 (2000); population, Landscan (2005); rail, VMAP0 (1997); rivers, CIA World Data Bank II (1980s); roads, Euro-med (2000) and VMAP0 (1997); slope, in degrees, SRTM30 (2000); urban areas, GPW3-GRUMP (2000); water bodies, Global Lakes and Wetlands Database layer 1 (2004). Note: Lagging areas represent 40 percent of the administrative units by country with the lowest consumption, or 40 percent of the adminis- trative units by country that have the highest rates of poverty, and the nonlagging designation comprises the remaining of administrative units. Using Nelson methodology (2008) to calculate a travel-time model, "near" is defined as a travel time of less than or equal to four hours from a city of 500,000; likewise, distant is any travel time greater than four hours. Based on population density data from Landscan (ORNL 2005), "dense" is defined as greater than 50 people per square kilometer, and "sparse" is the remainder. Areas of population density less than 1 are displayed in grey. It is, of course, beyond the scope of this report to recommend a devel- opment strategy for any given lagging area. That would require location- specific studies; however, it has been possible, by means of Geographic Information System (GIS) technology, to summarize the characteristics of MENA's leading and lagging areas. Map 3.5 shows our policy-oriented typology of the MENA region. Locations are classified according to three dimensions: · Dense versus sparse population. The criterion for a dense area in map 3.5 is that it should have at least 50 inhabitants per km2. To put this in Diagnose to Prescribe: Uncovering the Truth about Spatial Disparities 115 perspective, this is approximately the average population density of Ireland, Mexico, and Lithuania, or one person for every two hectares of land. It is notable that the areas of high population density in MENA appear as small islands in a wide sea of sparsely inhabited areas. These high-density areas cluster in six strings: along the Maghreb coastline, in the Levantine hinterland, in the valleys of the Nile and Euphrates, among the western highlands of the Republic of Yemen, and along the central north-south axis of the Islamic Republic of Iran. Each of these strings is cut off from the others by arid desert or semi-desert areas. · Near versus distant. The criterion for a "near" area is that it should be no more than three hours' travel by road from a city of at least half a million people. These areas benefiting from proximity to agglomer- ation of course extend radially from areas of high density. 116 Poor Places, Thriving People · Lagging versus nonlagging. For the purposes of map 3.5, lagging areas were defined as the least developed 40 percent of administrative divi- sions. Mean per capita household consumption was used as a default indicator of development. If consumption data were not available, poverty data were used. Neither consumption nor poverty data were available for Algeria, so a composite index of development, which was based on the ratios of people without education, business registra- tion, and car ownership, to wilaya population, was applied. The mapping emphasizes MENA's fragmented geography. As much as 86 percent of the region's territory is more than three hours' travel from a city of half a million people. As little as 3 percent of MENA's surface area has a population density greater than 50 persons per km2. But the mapping exercise also reveals how MENA's human popula- tion is clustered into islands of density (table 3.14). Ninety-two percent of the population is squeezed into that 3 percent of the surface area with population density greater than 50 persons per km2. Eighty-four percent of the population lives in areas with population density greater than 200 persons per km2 (the average population density of Italy or Luxembourg), and 81 percent lives in areas with a density greater than 400 persons per km2 (the average population density of the Nether- lands). And 72 percent of MENA's population lives in the 14 percent of MENA's land area that is within three hours of a city of at least half a million people. TABLE 3.14 Indicators of Population Concentration Percentage of population living in areas of <50 <100 <200 Country persons per km2 (%) persons per km2 (%) persons per km2 (%) Algeria 12 17 22 Egypt, Arab Rep. 1 2 3 Iran, Islamic Rep. 11 14 18 Iraq 9 13 15 Jordan 5 8 11 Lebanon 4 6 10 Libya 8 10 11 Morocco 13 19 25 Syrian Arab Republic 11 15 20 Tunisia 15 22 28 West Bank and Gaza 2 5 9 Yemen, Rep. 12 19 28 Total 8 12 16 Source: See Appendix. Diagnose to Prescribe: Uncovering the Truth about Spatial Disparities 117 Lagging areas are, of course, more likely to have low population den- sities and to be far from cities. A resident of a lagging area in MENA is 79 percent more likely to live more than three hours from a major city than a resident of a nonlagging area. And a resident of a lagging area is 56 percent more likely to reside in an area of population density of less than 50 persons per km2 than someone from a nonlagging area. In fact, the average population density of a lagging administrative divi- sion in MENA is less than half that of a nonlagging administrative region, 22 persons per km2 compared with 55. However--and this is the single most critical message from the map- ping exercise--a significant proportion of the lagging area population lives in the small fraction of the total land area that is densely inhabited and close to agglomeration. Sixty-one percent of the lagging area popu- lation lives within three hours of a city with half a million people or more, and 89 percent of the lagging area population lives in places where the population density is 50 persons per km2 or more, 85 percent in areas of 100 persons per km2 or more, 81 percent in areas of 200 persons per km2 or more, and 77 percent in areas of 400 persons per km2 or more, which is the average population density of the Netherlands. In other words, most lagging areas may be sparsely inhabited, but most lagging area pop- ulations are reasonably dense. The proximity of populations to economic density seems to be a dis- tinctive characteristic of MENA when compared with other parts of the world. GIS analysis (Hon, Rojchaichanintorn, and Schmidt 2009) has estimated the average travel time for various countries from home to a city of 100,000 or more, which is obviously a less demanding criterion than the half-million used for MENA earlier. The results in table 3.15 suggest that MENA, with 72 percent of its population less than 3 hours from a city of half a million, enjoys a proximity-to-density advantage over major developing countries such as Brazil, Chile, Pakistan, and TABLE 3.15 Proximity Comparators Average travel time to a city Country of 0.1 million or more (hours) Brazil 3.0 Chile 2.9 China 2.1 India 2.2 Mexico 1.8 Pakistan 2.9 Thailand 3.4 Source: Hon, Rojchaichanintorn, and Schmidt 2009. 118 Poor Places, Thriving People Thailand. It is not surprising that this finding should be so, given MENA's relatively high agglomeration indexes. The character of MENA's lagging areas varies from country to coun- try (table 3.16 and figure 3.21). Basically, one can split MENA's economies into three groups according to the human geography of their lagging areas: · The Maghreb+ group (Tunisia, Morocco, Algeria, and Libya, plus the Republic of Yemen and the Islamic Republic of Iran). These coun- tries' lagging areas have a relatively high share of their lagging area populations in low-density areas and areas that are distant from major cities. Note, however, the use of the word "relatively"; as noted earlier MENA's lagging area populations tend to be denser and closer to agglomeration than those of other parts of the world. · The Mashreq group (Syria, Iraq, Jordan, Lebanon, and the West Bank and Gaza). Their lagging area populations are even more likely to live in densely populated areas or near to major cities. · Egypt. Egypt's lagging area populations are not in sparsely popu- lated areas, but a high proportion of them (basically Upper Egyp- tians) live far from major cities. TABLE 3.16 Geographic Character of Lagging-Area Populations in MENA % of country's lagging area population >3 hrs from city of >0.5 million <3 hrs from city of >0.5 million <50 people per km2 >50 people per km2 <50 people per km2 >50 people per km2 % of total population (%) (%) (%) (%) Algeria 15 48 4 33 Egypt, Arab Rep.a 0 43 0 56 Iraq 2 9 9 80 Iran, Islamic Rep. 8 36 6 50 Jordan 4 1 4 91 Lebanon 1 3 5 91 Libya 4 37 5 54 Morocco 19 34 11 36 Syrian Arab Republic 4 8 8 80 Tunisia 16 37 13 34 West Bank and Gaza 0 5 2 93 Yemen, Rep. 11 24 7 57 MENA12b 6 33 5 56 Source: See Appendix. Note: a. Leading and lagging categorization data unavailable for 1 percent of the Arab Republic of Egypt's population. b. Djibouti is not included in MENA12. Diagnose to Prescribe: Uncovering the Truth about Spatial Disparities 119 FIGURE 3.21 The Mashreq's Lagging Area Populations Are Denser and Closer to Density than the Maghreb's 35 30 Tunisia of lagging area population in area with population Morroco 25 density < 50/km2 20 Yemen, Rep. Algeria 15 Iran, Islamic Rep. Syrian Arab Republic Iraq MENA13 10 Libya Jordan Lebanon 5 West Bank and Gaza Egypt, Arab Rep. 0 0 10 20 30 40 50 60 70 of lagging area population > 3 hours from a city of 0.5 million people Source: Study Team. What does this mean for lagging area policies? Again, it must be empha- sized that any lagging area deserves its own made-to-measure diagnosis. But the general implication is that the Mashreq's lagging area popula- tions are more likely to have the necessary proximity to agglomeration to benefit from spillover connectivity (Policy Package 2). These are the "belt" areas, of which Al-Mafraq was an example. They and Egypt's lagging areas are also more likely--other things being equal--to get an economic competitiveness boost from their agglomeration of popula- tion, giving regional economic development approaches (Policy Pack- age 3) a better chance of success. These are the "pockets" of density, of which Aleppo was a specimen. In the Maghreb, however, a lagging area's population is relatively more likely to be living in a "fringe" area of low density and high distance from agglomeration, like the good citizens of Djelfa. In such areas, as in all lagging areas, it is critical to remove any spatial biases in access to services or the business environment (Policy Package 1). The purpose of the discussion earlier was to develop a general frame- work for analysis. Geographic generalizations, however, are hazardous when applied to specific places. The only way to identify the appropriate 120 Poor Places, Thriving People policy approach for a given lagging area is to recognize its specificities, by zooming in to the appropriate level of detail, which is the subject of the following section. To Understand a Lagging Area, Zoom In and Zoom Out Analyzing a lagging area at different levels of spatial disaggregation makes for better policies. Each time the analysis zooms in by one level, new policy-relevant information appears. Take the lagging area of Upper Egypt, for example. The analytical focus shifts depending on whether one is looking at Upper Egypt at a regional, governorate, district, or village level. The richest insights and policy options come from thinking at different spatial scales (box 3.4). As a region within a country. A standard breakdown of spatial dispari- ties in Egypt is between broad regions: metropolitan Egypt, Lower Egypt, Upper Egypt, and frontier governorates. At this level of aggrega- tion, Upper Egypt as a whole appears as the lagging area (table 3.17). The analysis, therefore, focuses on the broad differences between Upper Egypt and Lower Egypt: their distances from the Cairo metropolis, household composition, and land availability. When the World Bank analyzed Upper Egypt at the regional level (2006b), the policy conclu- sions therefore focused on trunk connectivity and agricultural productiv- ity and marketing. But the regional-level analysis is overshadowed by the observation that 80 percent of the differences in per capita consumption between Upper Egypt and Lower Egypt is due to demography: the higher ratio of children and old people to people of working age in BOX 3.4 The Arab Republic of Egypt's Human Development Reports: A Model for MENA? The Arab Republic of Egypt's annual Human Development Reports, a joint product of the Egyptian government and the United Nations Development Programme (UNDP), have helped the country to focus on governorate- and district-level approaches. Every year, this publication puts subnational indicators of economic, human, and social devel- opment into the public domain. At the same time, they have documented the continu- ing challenge of mobilizing governorate- and district-level administrations to deliver more responsive public services (see, for example, UNDP 2004). The reports have been posted online for every year since 1997, giving an overview of spatial development that is unmatched in MENA. As a public data resource on subnational development trends, Egypt's National Human Development Reports could serve as a model. Diagnose to Prescribe: Uncovering the Truth about Spatial Disparities 121 TABLE 3.17 Contrasting Broad Regions in the Arab Republic of Egypt Indicator Lower Egypt Upper Egypt UN Human Development Index (2008) 0.719 0.693 GRDP per capita (2005­06; LE) 6,911 5,867 Life expectancy at birth (2006) 71.1 69.8 Women as % of labor force (2005) 27 20 Maternal mortality per 100,000 live births (2005) 47 55 Percentage of unfit school buildings (2003­04) 20 26 Source: UNDP Egypt National Human Development Report 2008. Note: GRDP = gross regional domestic product; LE = Egyptian pounds. Upper Egypt (World Bank 2009b). At the regional level, therefore, the emphasis was on realizing Upper Egypt's agricultural potential in the face of an intractable demographic handicap. As a governorate of Upper Egypt. But when one zooms in to the gover- norate (province) level, new issues come into sharper focus. In 1998, the Egyptian government launched the Sohag Rural Development Project. Governorate-level analysis had shown that Sohag was the poorest gover- norate in Upper Egypt. In fact, Sohag was the core of a "Middle Upper Egypt" lagging subregion of the lagging region of Upper Egypt. And when the analysis concentrated on Sohag's development challenges, a new set of issues came to the forefront: the wide disparities among dif- ferent districts' access to public services such as sanitation and health care; and the difficulties experienced by the governorate and district authorities in allocating public investment budgets to priority projects in the priority district, with all investment decisions being taken in Cairo by sectoral ministries and the Ministry of Economic Development. Based on the governorate-level perspective, therefore, the government's policy response was a program that emphasized the bottom-up, participatory planning of social infrastructure investments, with financial allocations going to the poorest districts. As communities of Upper Egypt. Measuring spatial disparities at sub- provincial level used to be problematic. The options available were to use national household survey data, which are usually unreliable at subprovin- cial level because of the cluster sampling approach, or made-to-measure village surveys, which are either very localized or very expensive. The new technique of poverty mapping (Demombynes and others 2002), however, has given policy makers this option. Poverty mapping uses the household survey to identify poor households. It then looks up the census responses for these same households and thus finds out which 122 Poor Places, Thriving People census responses are associated with poverty. Finally, it looks in the cen- sus for other households that gave similar census responses. In this way, it can plot a map of where the poor people are, right down to village level. In MENA, Egypt and Morocco have completed poverty-mapping exercises.15 A poverty map instantly deepens one's understanding of spatial dis- parities. It brings into sharp focus the wide gap in development between communities separated by only a few kilometers. It helps identify the fac- tors that determine whether a village develops. Figure 3.20, above, illus- trates the importance of village- or community-level information. It shows how poverty reduction over time was a village-level process in Egypt, with the residents of entire villages breaking out as a group to join the "club" of more advanced villages, while the residents of other villages got left behind as a group. The conclusion is clear: if households in a vil- lage succeed together or fail together, it is critical for policy makers to know the correlates of village-level success. As the Morocco poverty-mapping exercise illustrated, the closer one zooms, the larger the proportion of differences in welfare that can be linked to location. Location represented only 18 percent of inequal- ity when the breakdown was at the regional level, but it accounted for 40 percent of inequality at the commune level (Lanjouw 2004). The finer the disaggregation, therefore, the easier it is to target the poor geographically, with public investments or social protection spending. Morocco's poverty-mapping exercise was, therefore, critical to the design of the National Human Development Initiative (INDH), which is budgeting US$0.26 billion per annum on spatially targeted public social and economic infrastructure investments. Lebanon's Economic and Social Fund for Development (ESFD) managed to improvise a useful fine-resolution poverty map in 2004 to underpin its social strategy, despite the lack of census data or household consumption data (Nahas 2009). The spatial divisions used were the cadastral limits, the smallest division available to estimate population on the basis of the buildings census. To estimate living standards, the map used the distribution of retail stores, restaurants and hotels, services, construction activity, electricity consumption, and the number of "minutes talked" per head on each telephone exchange. The resulting map was detailed enough to reflect poverty pockets within suburbs (Hay Al Sillum, Biddawi, and Palestinian camps) and the persistent influence of the wartime frontlines in seven distinct areas. A key conclusion of the ESFD analysis was that the geographically small pockets of dense poverty in the suburbs of the large towns, which would have been overlooked by a less detailed map, contained more poor Diagnose to Prescribe: Uncovering the Truth about Spatial Disparities 123 TABLE 3.18 Policy Implications of Multi-Level Spatial Analysis in Rural Upper Egypt Level of analysis Focus of analysis Focus of findings Emphasis of policy implications Leading versus Decomposing the Upper Egypt's high dependency Linking rural Upper Egypt Lagging region interregion gap in per ratio, dependence on farmers to urban markets capita productivity and low-value agriculture, and through supply chain consumption remoteness from Cairo's development, trunk road demand for farm products improvements, and high-value cropping Province and Improving service delivery Responsiveness of investment Deconcentration and district by governorate and decisions at governorate decentralization of public district administrations and district level investment planning Village and Poverty mapping and Disparities in the welfare of Village-level targeting of public community small-scale surveys neighboring locations investment projects Residing in a successful Targeting dense poverty village, which is an important pockets in richer urban part of escaping from poverty areas Source: Study Team. households than the extensive areas of rural deprivation. In other words, high-resolution analysis helps to remind policy makers that rich areas can contain many poor people. Table 3.18 summarizes the case for a multilevel analysis, taking the example of Egypt and zooming through the national, provincial and sub- provincial, scales. Although all MENA economies have major budgetary commitments to lagging areas, much analysis is required at every level to optimize their strategies. We can close this exploration of lagging area diagnostics in MENA with the conclusion that better information would make for better policies. We have seen that every lagging area is different and that spatial varia- tion is just as evident at the micro level as at the macro level. Indeed, the study of spatial disparities is reminiscent of fractal geometry, where each pattern contains an infinite series of smaller patterns nested within it. And yet the foundation of subnational data and analysis in MENA is cur- rently simply too weak to support policy makers' demands for an effec- tive and evidence-based response. This chapter has attempted to show what kinds of analysis can inform spatial development policy. It has shown how subnational data can give critical insights to policy makers, such as the importance of demography, education, dense populations, suspended rural transition, and within- province migration. The question is, therefore, whether MENA can afford to continue spending billions of dollars on spatial development programs without adequate subnational data. 124 Poor Places, Thriving People Endnotes 1. FAOSTAT, http://faostat.fao.org, accessed August 14, 2009. The data show increases in the arable land-labor ratio for Egypt, Iraq, Lebanon, and Libya. The figures showing a dramatic increase in arable land for Egypt from 1990 to 1995 are surprising. 2. Source: Palmer Drought Severity Index map from (IPCC 2007). 3. Personal communication: Director of national statistical service. 4. Another reason is that comparing rural-urban disparities with overall inequality downplays the well-being of minorities: if rural people are a small share of the overall population, they will never account for a large share of over- all inequality, no matter how poor they are. 5. Named after the surnames of its inventors: Elbers, Lanjouw, Mistiaen, and Özler. 6. The rise and fall of the household head's earning capacity with increasing age is captured by the coefficient on the square of the household head's age. 7. The household survey data did not include a rural-urban field, or data for two southern mohafazat (Nabatieh and S. Lebanon). 8. Excluding data from Guelmime-Es Semara. 9. Also with data from Guelmime-Es Semara and the areas of Oued Ed-Dahab Lagouira and Laayoun-Boujdour-Sakia El-Hamra. 10. The total fertility rate (TFR) is a synthetic indicator denoting births per woman per lifetime, and is based on the age profile of women in a given year and the number of births per woman in each age group in that year. 11. Weighting the coefficient of variation by population stops provinces with small populations from having a disproportionate impact on the statistic. 12. The UNDP GRDP estimates published in the national Human Develop- ment Reports are based upon incomes reported in the national household income and expenditure surveys (HIECS). The reliability of income survey responses is usually low. 13. "As more children were enrolled in school, the quality and efficiency of edu- cation came to the forefront. In response, MENA economies experimented with a variety of mechanisms, including decentralization, engaging the private sector in the provision of education, and the adoption of quality assurance pro- grams (World Bank 2008d, 4)." 14. Statistical Bureau of the Islamic Republic of Iran, census 1375. 15. Algeria's Carte de la Pauvreté (UNDP 2006) follows a different approach, using local government revenues as an indicator of the level of development of different administrative divisions. References Abbasi-Shavazi, M. J., and P. McDonald. 2005. "National and Provincial- Level Fertility Trends in Iran, 1972­2000." Working Paper in Demog- raphy No. 94. Australian National University, 10. Akin, M. S. 2005. "Education and Fertility: A Panel Data Analysis for Middle Eastern Countries." Journal of Developing Areas 39 (1): 55­69. Diagnose to Prescribe: Uncovering the Truth about Spatial Disparities 125 Ameen, J. 2008. "Kurdistan's Genocide: A New Insight." Xebat, April 16. Amin, M. 2009. How Tough Is It to Find an Employee You Can Trust? October 2, http://psdblog.worldbank.org/psdblog/2009/10/labor -issues-in-india-beyond-labor-laws.html#more (private sector devel- opment blog, accessed October 5, 2009). Assaad, R. 1993. "Formal and Informal Institutions in the Labor Market, with Applications to the Construction Sector in Egypt." World Devel- opment 21 (6): 925­39. Assadzadeh, A., and S. Paul. 2004. "Poverty, Growth, and Redistribution: A Study of Iran." Review of Development Economics 8 (4): 640­53. Ayadi, M., and M. Amara. 2009. Spatial Patterns and Geographic Determi- nants of Poverty in Tunisia. Cairo: Economic Research Forum. Behdad, S. 1989. "Winners and Losers of the Iranian Revolution: A Study in Income Distribution." International Journal of Middle East Studies, 327­58. Boulila, G., C. Gabsi, and M. Trabelsi. 2009. Regional Pro-Poor Growth and Convergence in Tunisia. Cairo: Economic Research Forum. Bourguignon, F. 1979. "Decomposable Income Inequality Measures." Econometrica 47 (4): 901­20. Centre d'Études et de Recherches Démographiques. 1989. Education et Changements Démographiques au Maroc. Rabat: Royaume du Maroc, Ministčre du Plan, Direction de la Statistique. Chant, S. 1992. Gender and Migration in Developing Countries. London: Belhaven. Chelala, C. 1996. "Egypt Faces High Population Growth." The Lancet (December 14): 1651. Cline, W. 2007. Global Warming and Agriculture: Impact Estimates by Country. Washington, DC: Center for Global Development; Peterson Institute for International Economics. De Brauw, A., and J. Giles. 2006. Migrant Opportunity and the Educational Opportunity of Youth in Rural China. IZA Discussion Paper 2326. Bonn: Institute for the Study of Labor. Demombynes, G., C. Elbers, J. Lanjouw, P. Lanjouw, J. Mistaien, and B. Ozler. 2002. Producing an Improved Geographic Profile of Poverty. United Nations University, WIDER. 126 Poor Places, Thriving People Elbers, C. L. 2007. "Re-interpreting Sub-Group Inequality Decomposi- tions." Policy Research Working Paper WPS 3687. World Bank, Washington, DC. Esber, R. 2008. "Gender and Spatial Disparities in the Middle East and North Africa. A Study of Saudi Arabia, Egypt, and Tunisia." Unpublished. FAO (Food and Agriculture Organization). 2009. Food Security and Agricultural Mitigation in Developing Countries: Options for capturing synergies. Rome: FAO. Hawkins, S., L. Griffiths, T. Cole, C. Dezateux, and C. Law. 2007. "Regional Differences in Overweight: An Effect of People or Place?" Archives of Disease in Childhood 93 (i): 407­13. Henderson, J. V. 1986. "Urbanization in a Developing Country: City Size and Population Composition." Journal of Development Economics 22 (2): 269­93. Henderson, V. 2000. "The Effect of Urban Concentration on Economic Growth." Working Paper 7503. National Bureau of Economic Research, Cambridge, MA. Henderson, J. V., Z. Shalizi, and A. Venables. 2001. "Geography and Development." Journal of Economic Geography 1 (1): 81­105. Hon, V., J. Rojchaichanintorn, and E. Schmidt. 2009. "A Framework for Bank Engagement in Lagging Areas." Unpublished. World Bank, Washington, DC. INS (Institut National de la Statistique). 1980, 1990, 2000. Enquęte Nationale sur le Budget. La Consommation et le Niveau de Vie des Ménages. Tunis: INS. IPCC (International Panel on Climate Change). 2007. Fourth Assessment Report. Geneva: IPCC. Iqbal, F. 2006. Sustaining Gains in Poverty Reduction and Human Develop- ment in the Middle East and North Africa. Washington, DC: World Bank. IRIN News. 2009. "Syria: Drought Driving Farmers to the Cities." September 2. http://www.irinnews.org/report.aspx?ReportID=85963. Khawaja, M. 2002. Internal Migration in Syria: Findings from a National Survey. Oslo: Fafo. Khawaja, M., R. Jurdi, S. Assaf, and J. Yeretzian. 2009. Unmet Need for the Utilization of Women's Labor: Findings from Three Impoverished Diagnose to Prescribe: Uncovering the Truth about Spatial Disparities 127 Communities in Outer Beirut, Lebanon. Cairo: Economic Research Forum. Lanjouw, P. 2004. "The Geography of Poverty in Morocco: Micro-Level Estimates of Poverty and Inequality from Combined Census and Household Survey Data." World Bank. http://siteresources.worldbank .org/INTMOROCCOINFRENCH/Resources/Lanjouw.Geography PovertyMorocco.pdf. Lowe, R. 2006. The Syrian Kurds: A People Discovered. London: Chatham House. McCormick, B., and J. Wahba. 2005. "Why Do the Young and Educated in LDCs Concentrate in Large Cities? Evidence from Migration Data." Economica 72 (285): 39­67. Mehryar, A. 2004. "Primary Health Care and the Rural Poor in the Islamic Republic of Iran." Paper presented at Scaling Up Poverty Reduction: A Global Learning Process and Conference, Shanghai, May 25­27. Mojtahed, A., and H. S. Esfahani. 1989. "Agricultural Policy and Per- formance in Iran: The Post-Revolutionary Experience." World Devel- opment 17 (6): 839­60. Nahas, C. 2009. "Spatial Inequalities in Lebanon: Development Trends and Public Policy." Unpublished background paper for this report. Nelson, A. 2008. Travel Time to Major Cities: A Global Map of Accessibility. Ispra, Italy: Global Environment Monitoring Unit, Joint Research Centre of the European Commission. http://gem.jrc.ec.europa.eu/. Nelson, G. C., M. W. Rosegrant, J. Koo, R. Robertson, T. Sulser, T. Zhu, et al. 2009. Climate Change: Impact on Agriculture and Costs of Adaptation. Washington, DC: International Food Policy Research Institute. Openshaw, S. 1984. The Modifiable Areal Unit Problem. Norwich: Geo Books. ORNL (Oak Ridge National Laboratory). 2005. Landscan Global Popula- tion Database. http://www.ornl.gov/landscan. Oweis, K. Y. 2009. "Water Crisis Uproots Syrian Farmers." Jordan Times, July 31. Ravallion, M., S. Chen, and P. Sangraula. 2007. "New Evidence on the Urbanization of Global Poverty." Policy Research Working Paper 4199. April. World Bank, Washington, DC. 128 Poor Places, Thriving People Rutstein, S. O., and K. Johnston. 2004. The DHS Wealth Index. DHS Comparative Report No. 6. Calverton, MD: ORC Macro. Shankar, R., and A. Shah. (2003). "Bridging the Economic Divide within Nations: A Scorecard on the Performance of Regional Development Policies in Reducing Regional Income Disparities." World Develop- ment 31 (8): 1421­41. Shenker, J. 2009. "We Are Going Underwater. The Sea Will Conquer Our Lands." The Guardian (Manchester), August 21. Shlash, A. 2009. "Iraq High Committee for Strategy Reduction. Draft Strategy Review Presentation." June. Amman. Shorrocks, A. 2005. "Spatial Decomposition of Inequality." Journal of Economic Geography 5 (1): 59­81. Sommeiller, E. 2008. "Spatial Analysis of Social Welfare in MENA Countries." Unpublished background paper for this report. Sutton, K. 2001. "Demographic Transition and Regional Population Change in Algeria, 1987­1998." Arab World Demography 4 (3): 206­16. Uchida, H., and A. Nelson. 2008. "Agglomeration Index: Towards a New Measure of Urban Concentration." Background paper. http:// siteresources.worldbank.org/INTWDR2009/Resources/4231006- 1204741572978/Hiro1.pdf. UN Habitat. 2006. State of the World's Cities 2006/7: The Millennium Development Goals and Urban Sustainability. Nairobi: United Nations Human Settlements Program. UNDP (United Nations Development Programme). 1999. Human Development Report of the Islamic Republic of Iran. Tehran: UNDP. ------. 2004. Egypt Human Development Report 2004--Choosing Decen- tralization for Good Governance.Washington, DC: UNDP. ------. 2006. Affinement de la Carte de la Pauvreté. Algiers: UNDP. ------. 2008. Poverty, Growth and Income Distribution in Lebanon. Beirut: UNDP. UNHCR (United Nations High Commissioner for Refugees). 2009. "Refworld: The Leader in Refugee Decision Support." Accessed June 16, 2009, from http://www.unhcr.org/refworld/topic,463af2212, 464dbe362,49ce361e28,0.html. Diagnose to Prescribe: Uncovering the Truth about Spatial Disparities 129 United Nations Department of Economic and Social Affairs. 2007. World Urbanization Prospects: The 2007 Revision Population Database. New York: United Nations. Wahba, J. 2007. "An Overview of Internal and International Migration in Egypt." Working Paper 0703. Cairo: Economic Research Forum. Wahba, J., and Y. Zenou. 2005. "Density, Social Networks, and Job Search Methods: Theory and Application to Egypt. Journal of Devel- opment Economics 78: 443­73. World Bank. 1999. Democratic and Popular Republic of Algeria: Growth Employment and Poverty Reduction. Washington, DC: World Bank. ------. 2006a. Tunisia: Agricultural Policy Review. Washington, DC: World Bank. ------. 2006b. Upper Egypt: Challenges and Priorities for Rural Development. Washington, DC: World Bank. ------. 2007a. Lebanon: Economic and Social Impact Assessment: From Recovery to Sustainable Growth. Washington DC: World Bank. ------. 2007b. Population Issues in the 21st Century. The Role of the World Bank. Washington, DC: World Bank. ------. 2008a. World Development Report 2008: Agriculture for Develop- ment. Washington, DC: World Bank. ------. 2008b. Syrian Agriculture: Towards the Social Market. Washington, DC: World Bank. ------. 2008c. The Road not Traveled. Education Reform in the Middle East and North Africa. Washington, DC: World Bank. ------. 2009a. World Development Report 2009: Reshaping Economic Geography. Washington, DC: World Bank. ------. 2009b. Arab Republic of Egypt. Upper Egypt: Pathways to Shared Growth. Washington, DC: World Bank. ------. 2009c. Yemen: Pathways Out of Rural Poverty. Washington, DC: World Bank. World Development Indicators Database. http://data.worldbank.org/ data-catalog. World Bank, Washington, DC. CHAPTER 4 Policy Package 1. Level the Playing Field and the Opportunity for Human Development in Lagging Areas Apply in All Circumstances: The Only Package for "Fringe" Areas Matching Policy Packages to Lagging Areas (from table 2.3) World Development Report 2009 Policy packages terminology Why? Where? Downside risk Package 1. "Institutions" Spatial disparities in well-being Applicable to all High cost of Level the playing field ("spatially can narrow even while lagging areas replicating and the blind production is agglomerating, (whether or metropolitan opportunity for policies") so long as policies are not Package services human development appropriate. MENA has a 2 or 3 applies) inappropriately in in lagging areas historical legacy of policies' lagging areas widening disparities. Source: World Bank 2009a. As we have seen in the previous chapter, lagging areas have many differ- ent qualities. Some lagging areas are close enough to agglomerations to benefit from growth spillovers, but some are not. Some lagging areas are sparsely populated, but many have high population densities. Some even believe in "lagging cultures" (box 4.1). But all lagging areas have one thing in common: they deserve fair policy treatment--a level playing field for development. As we saw in chapter 2, countries can reduce spatial disparities in living standards without having to spread economic growth across the territory. In fact, reducing spatial dis- parities in living standards, while letting the market determine the spatial 131 132 Poor Places, Thriving People BOX 4.1 Lagging Areas: Is Culture to Blame? Many people believe that poor places remain poor because of the culture of their inhab- itants. Every country has a deep stock of jokes and insults about the backwardness of the citizens of their lagging regions. We, therefore, wanted to see if we could find any evidence for a correlation between cultural values and geography in MENA. Data from the World Values Survey (WVS) were analyzed to assess whether people's views were correlated with the size of the settle- ment where they lived. The questions analyzed were those that tested social beliefs that might constrain economic development: · If jobs are scarce, should men have more of a right to a job than women? · Is an opportunity to use initiative important in a job? · Are thrift and saving money important qualities in a child? · Is education an important trait in a woman? · Should a person follow his individual beliefs or secular and religious authorities? The analyses did not reveal any trends. If cultural beliefs are a cause or an effect of spa- tial backwardness, we could not find any evidence. distribution of economic activity, is a win-win strategy; it brings both the economic benefits of agglomeration and the political benefits of spatial equity. So what is this "level playing field for development"? To answer this question, we need to delve into MENA's past. This chapter will show that the centralized state is the legacy of MENA's colo- nial and postindependence history. We will show how MENA's gover- nance characteristics have been inherited. We will then provide statisti- cal evidence that those with such governance characteristics tend to follow a more spatially concentrated development path. MENA's policy makers therefore face a dilemma. They have the task of promoting development in peripheral areas. But they are obliged to do so within the framework of administrative and political structures that are intrinsically centralizing. In short, the explicit policies are pushing in one direction, while the implicit structures and systems are pulling back in the opposite direction. Building a "level playing field for development" therefore means adjust- ing sector policies so that they are no longer biased in favor of metropoli- tan or central areas. The latter part of this chapter will thus move away Policy Package 1. Level the Playing Field and the Opportunity for Human Development in Lagging Areas 133 from political theory to the nitty-gritty of a range of government func- tions: public expenditure management, business environment regulation, education, health, the management of urbanization, and social protection. How Governance Affects Urbanization--Some Numbers and an Explanation Political organization has a spatial impact. The lower a country's level of political accountability, the more urbanized and agglomerated it is. Following the example of Ades and Glaeser (1995), but using a fixed effects panel data model with observations from 1975 to 2004 (tables 4.1, 4.2), we estimated the relationship between governance TABLE 4.1 Politics Affects the Size of a Country's Largest City Explaining (log) population of largest city logNonUrbanPop 0.41 0.42 0.4 0.41 (7.81)** (8.10)** (7.97)** (7.97)** logOtherUrbanPop 0.72 0.74 0.71 0.71 (25.85)** (26.13)** (26.55)** (26.51)** logPerCapGDP ­0.2 ­0.23 ­0.21 (8.73)** (9.97)** (8.60)** Poor: logPerCapGDP ­0.21 (8.09)** Rich: logPerCapGDP ­0.21 (8.48)** Trade (logOpenc) ­0.01 (0.24) Trade (logOpenk) 0.06 (2.81)** Trade (logTradeShare) ­0.01 ­0.01 (0.29) (0.29) DictatorDummy 0.06 0.06 0.06 0.06 (2.95)** (3.22)** (2.78)** (2.78)** FreeDummy ­0.04 ­0.04 ­0.04 ­0.04 (1.90) (1.99)* (2.02)* (2.03)* Ethnic conflict 0.03 0.03 0.04 0.04 (3.25)** (3.32)** (3.60)** (3.59)** Constant 3.15 2.79 3.27 3.26 (5.80)** (5.03)** (6.11)** (6.03)** Observations 563 563 588 588 Number of countryid 77 77 77 77 R-squared 0.91 0.91 0.92 0.92 Source: Study Team. Note: Fixed effects for countries and years included, but not reported. Absolute value of t statistics in parentheses. * significant at 5%; ** significant at 1%. 134 Poor Places, Thriving People TABLE 4.2 Political Rights Reduce Agglomeration Ordinary least squares Agglomeration index Political right ­16.657 (2.80)*** Political stability 2.367 (1.60) Log (GDP per capita) 32.255 (4.95)*** [Log (GDP per capita)] ­1.730 (3.69)*** Log (population size) 16.702 (7.02)*** Trade openness ­0.061 (2.64)*** Log (land area) ­4.038 (5.17)*** Share of paved roads 2.537 (0.82) Log (population in agricultural sector) ­10.358 (4.36)*** MENA 5.561 (1.22) Constant ­152.426 (6.74)*** Observations. 157 R-squared (adjusted) 0.75 Source: GDP per capita, trade openness, land area, population in agricultural sector: World Development Indicators. Political rights: http://www.freedomhouse.org/template. Political stability: Kaufman, Kraay, and Mastruzzi 2009. Note: ***, ** respectively denote significant coefficients at 1% and 5% level. Figures in parentheses are robust t-student. indicators and the size of a country's main city. The governance indi- cators were measures of political rights and freedom (Freedom House) and measures of political and social instability (Political Instability Task Force). We see that politics matter. All else being equal, popula- tion in the largest city rose when countries became dictatorships and fell when countries become "free" (as defined by Freedom House). There was also a tendency for ethnic conflict to cause rises in the pop- ulation in the largest city.1 Looking beyond the growth of the largest city to urbanization more generally, we conducted a similar analysis using the urban population in a country as the dependent variable. The patterns were similar. Being a "free" country (as defined by Freedom House) is associated with slower urbanization. Ethnic conflict increases urbanization. Interestingly, a Policy Package 1. Level the Playing Field and the Opportunity for Human Development in Lagging Areas 135 dictatorship does not increase urban population--which makes sense if an effect of dictatorship is to attract people from secondary cities to the center of power in the main city. Finally, the relationship between governance indicators and the World Bank's (2009a) agglomeration index was estimated for a world- wide sample of 182 countries. The governance indicators in question were the index of political rights from Freedom House (2009) and the index of political stability of Kaufman, Kraay, and Mastruzzi (2009). The exercise also took account of a number of other factors known or thought to be related to agglomeration: GDP per capita, population size, land area, trade openness, road quality, and the share of the population employed in agriculture. A dummy variable for MENA economies was also added. The variables explained 75 percent of the variation in the agglomera- tion index. The coefficients on GDP represent the strong link between agglomeration and growth. The MENA dummy variable was not statis- tically significant, suggesting that the other variables in the equation cap- tured all of MENA's special characteristics. The key finding was the extremely strong negative correlation between political rights and the agglomeration index, significant at the 1 percent level. In other words, more political rights meant less agglomeration. This correlation was significant for all the specifications tested. This matters for the MENA13, which has a lower average political rights score (0.18) than non-MENA lower-middle-income economies (0.26), other developing countries (0.31), and the rest of the world (0.40). The impact of MENA's low political rights on agglomeration, though not massive, is very clear. If MENA13 economies' political rights indicator increased to that of the average country in the world, this would be accompanied by a 4­percentage point reduction in the agglomeration index. What is happening here? One can think of political rights as spread- ing political accountability not only across the population but also across places. Democratic processes and public service accountability give citi- zens across the territory a voice in policy making. But in the absence of formal mechanisms for sharing political accountability across the terri- tory politicians and bureaucrats are disproportionately influenced by the people who are physically closest to them: their face-to-face contacts and the metropolitan population more generally. Contemporary drivers of spatial bias go back to MENA's colonial history. The institutions responsible for spatial policies are themselves the prod- uct of earlier spatial policies (North 1990; Kuran 2004). Yesterday's his- torical anti­poor area biases thus constrain the performance of today's explicit pro­poor place policies. In the MENA region, these patterns 136 Poor Places, Thriving People have their roots in colonial times, as far back as Ottoman rule.2 Four features of MENA's colonial legacy are important in this context: · A belief in the state's mission to regulate society centrally, supported by a highly centralized administrative machinery and fiscal system · A condition of "state autonomy," which manifested itself in a lack of representative institutions articulating local demands · An outward orientation of economic, trade, and fiscal policies that catered to the metropolis to the detriment of lagging areas · A concentration of the colonial elite in urban and coastal areas MENA's colonial legacy is interventionist government. At independence, most modern MENA economies were characterized by ideologies based on a belief in the state's capacity to regulate social and economic relations (Richards and Waterbury 1990; Owen 1992; Luciani 1989; Ayubi 2008; Ozbudun 1997). The statist tradition derived partly from the Ottoman rule. In addition to collecting taxes and waging wars, the Ottoman admin- istration provided welfare services, fixed prices for basic staples, adminis- tered public procurement schemes to keep certain prices low, and set up public enterprises to produce strategic commodities such as food staples and gunpowder. These features of Ottoman rule persisted--and in some cases were reinforced--during European colonization. They influenced MENA's state-driven development approaches and were further accentu- ated by Kemalist ideologies throughout MENA (Brown 1997). Colonial statehood was based on the development of strong central bureau- cracies. During the 1950s and 1960s, MENA experienced a regionwide process of bureaucratization: administrations grew, increased prestige attached to public office, and a strong belief in the developmental role of the bureaucracy prevailed (Ayubi 1988, 13). This feature was not born at independence either; MENA economies inherited key features of their government organization and administration structure from the Ottoman Empire and European powers. The Ottoman administration was a strong bureaucratic machine (Findlay 1980, Findlay 1997). Rather than deriving their status and power from the ownership of economic assets such as land, the local elites in the Ottoman provinces sought direct political power though positions in the state bureaucracy (Ozbudun 1997). Statism and bureaucratic rule in the provinces were based on the Ottoman fiscal system, which is widely considered to have shaped governance structures (Tosun and Yilmaz 2008). Colonial rule stifled the representation of local interests. Under the Ottoman rule, the governing apparatus was insulated from social pressures. The bureaucracy was rendered "autonomous," through a deliberate policy of Policy Package 1. Level the Playing Field and the Opportunity for Human Development in Lagging Areas 137 recruitment and compensation aimed at cutting identity and economic links with the dominant landed, commercial, and industrial classes, but also from regional entities (Ozbudun 1997). The separation between rulers and ruled was reinforced by the absence of representative systems. The Ottoman Empire did not evolve from feudal pluralism to constitu- tionalism and representation (via the estates and the corporate autonomy of bodies such as the church, cities, and the guilds) that characterized the rise of European states. Neither centrally nor in the provinces did Istan- bul develop representative mechanisms until the last quarter of the nine- teenth century (Ozbudun 1997). No substantial progress was achieved during European colonization either. So the economic grievances of mar- ginal areas were neither articulated nor addressed. Colonial tax systems encouraged a neglect of lagging areas. First, they dis- couraged economic development in marginal areas. Particularly during the classical period (1453 to the 1600s), the Ottoman Empire had very low provincial tax rates compared with Europe and other empires (Karaman 2009). Both taxation and internal security in the Ottoman provinces were delegated to local elites. This presented the central administration with a trade-off in setting tax rates. On the one hand, the center and their dele- gated authorities shared an interest in maximizing gross tax revenue. On the other, the center was concerned about the power balance between itself and its tax-collecting delegates. So the central administration sometimes capped tax rates in the provinces to limit the power of local elites. This cap also discouraged the development of the tax base because neither the cen- tral administration nor the local elites stood to gain from economic growth. The tax ceilings induced, at best, economic neglect of lagging areas and, at worst, inefficient or outright detrimental economic policies. Second, colonial tax systems diverted fiscal resources away from local investments. The Ottoman fiscal system shifted a large share of tax rev- enues away from the provinces and toward the capital region. The tax receipts that remained with provincial governors were only for security and administration expenses, not local investments. The bias toward police and security spending continued and was reinforced during French and British rule, during which some two-thirds of total expendi- ture was typically security related (Owen 1992). In catering essentially to the metropolis, colonial management avoided redress- ing regional inequalities. In addition to diverting taxes, colonial rule diverted actual output. Ottoman governors owed their ultimate allegiance to Istan- bul. Their performance was assessed on the basis of whether they catered to the security and consumption needs of the Porte, rather than on their development of the provinces. During European colonization, the colonial states continued to serve as providers of goods to the metropolis. Agricul- ture focused on export crops, such as wine, cereals, and tobacco. In the 138 Poor Places, Thriving People initial stage, industry was developed only for the processing in situ of certain goods. In favoring resource extraction over economic develop- ment, the colonial powers instated an economic geography of "useful ver- sus useless" regions. The agriculturally rich areas of Morocco--the coastal plains and plateaus as well as the north and west of the Atlas ranges--were known to the French and Spanish colonial powers as "Le Maroc Utile" (Pennell 2001, 169; Thomas 2005, 212; Eickleman 1976). Similarly, "l'Al- gérie utile" was where mineral, agricultural, and demographic resources could allow quick economic extraction. This spatial breakdown of colonies was reflected in "divide and rule" politics. Rather than seeking the consent of disadvantaged groups (and regions) through investments and economic benefits, European powers kept these groups divided along ethnic and reli- gious lines, to prevent their alliance against the government. Particularly in the colonial territories, which are today the Syrian Arab Republic, Lebanon, Iraq, Morocco, and Algeria, this maintained spatial disparities (Cleveland 1999; Owen 1992; Bengio and Ben-Dor 1999). The development of metropolis-oriented economies encouraged coastal agglom- eration. Export orientation resulted in an increase in the importance and size of coastal cities. New business opportunities and improving living con- ditions attracted populations from inland rural areas, as well as Europe. Between the early nineteenth century and the early twentieth century, the population in Tunis more than doubled. With the French colonial period after 1881, the influx of Europeans grew substantially, reaching close to 100,000 in 1936--that is, 45 percent of total city residents. Resources were directed toward urban settler communities. Investments were primarily oriented to the infrastructure necessary to support colonial eco- nomic activity and trade in coastal cities and allowing adequate trans- portation and connection links, but they also aimed to improve the living conditions in settler areas. A classic example is the Constantine Plan of 1955­63--one of the last development plans to be attempted under colo- nial administration. Conveniently located between Algiers and Annaba (called Bône at the time, and the second-largest city in Algeria), Constan- tine was a crossroads for the transport of agricultural products and pri- mary resources coming from the inland, as well as an important internal market. The plan provided for the increase of European settlements in the main coastal cities and their ancillary towns, the building of railway links between main port cities and internal hubs, the expansion of irriga- tion within settlement estates, and the expansion of settler agricultural development. Policies were explicitly aimed at favoring new settlers, such as the establishment of a housing authority in Tunis to facilitate the con- struction of inexpensive housing for French immigrants. Such policies left nonsettler regions behind and often occurred at the expense of the local national community. For example, in 1932 an infrastructure development project in education and health, financed with Moroccan taxpayer money Policy Package 1. Level the Playing Field and the Opportunity for Human Development in Lagging Areas 139 and profits from the local phosphate industry, benefited nearly exclusively the settler community that was just 4 percent of the total population. Euro- pean hospitals in the predominantly Muslim cities of Marrakech, Fčs, and Meknčs received up to 10 times more funding than hospitals of the major- ity population. The local population, representing more than 90 percent of total inhabitants of the three cities, received just over 10 percent of hospital funding (Thomas 2005). Sometimes, regional disparities were also a result of being subject to different colonial administrations. Morocco is a good example. Although facing the Mediterranean and strategically con- trolling the Gibraltar Strait, northern areas under Spanish rule were much neglected compared to areas under the French protectorate because of the lack of resources of the Spanish government. At independence, economic activity was limited and infrastructure scarce, with path dependency encouraging backwardness and an economic lag behind other areas. Strong spatial disparities therefore existed at the moment of independence. The pro-settler and coastal bias of colonial development processes-- already distinct under the Ottomans--resulted in a deepening of the spatial divide. At independence, many areas faced dire poverty. The pre- dominantly urban politics characterizing the colonial period, with the capital city as the center of a political arena dominated by a small elite, continued through independence. Administrative expansion proceeded in parallel with a centrally planned approach to economic management. The central government extended its interventions into rural areas, as large land estates were expropriated. As the central administrations expanded, power remained concentrated at the core. Whether the path to independence had been smooth or characterized by struggles, power remained centralized in the years following indepen- dence. As mentioned earlier, postindependence governments confronted a series of challenges, both internal and external (Korany 1987). Internally, weak bureaucratic capacity constrained the development of influence over the entire territory. Empirical metrics suggest that MENA economies remain less externally accountable than others at a similar level of development. Oil and gas revenues placed huge amounts of resources in the hands of some central administrations, permitting large bureaucracies, a cen- trally driven development policy, and state autonomy, as the state does not need to penetrate society and develop the economy to raise revenues (Luciani 1989; Beblawi and Luciani 1987; Henry 2004). Discontent with government policies is sometimes deeper in cities. (See table 4.3 and figure 4.1.) This is despite the fact that cities often fare better in terms of a variety of indicators. (The exception, shown in table 4.3, is the Islamic Republic of Iran, where inhabitants of Tehran expressed more confidence in the Khatami administration than did the rest of the country.) Given the increased opportunities for association 140 Poor Places, Thriving People TABLE 4.3 Nonmetropolitan People State More Confidence in Their Governments Percentage of interviewees who said they have "a great deal of confidence" in the government Outside the capital area Capital area Egypt, Arab Rep. (2000) 19.8 10.9 Iran, Islamic Rep. (2000) 24.5 33.4 Iraq (2004) 13.1 6.8 Jordan (2001) 54.8 48.0 Morocco (2001) 33.1 26.5 Source: World Values Survey Association, European and World Values Surveys Four-Wave Integrated Data File, 1981­2004. Note: Boldface identifies the larger number in the row. FIGURE 4.1 People in Smaller Settlements State More Confidence in their Government in the Arab Republic of Egypt % of population expressing 50 ``great confidence in government'' 40 30 20 10 0 s 0 00 0 0 00 e es ,00 ,00 ,00 or 0,0 0,0 dm dl ­5 20 50 ­1 10 an 0­ 0­ 00 an 0­ 00 00 2,0 ,00 ,00 00 5,0 ,00 2,0 10 20 0,0 50 50 population of towns and cities Source: World Values Survey Association, European and World Values Surveys Four-Wave Integrated Data File, 1981­2004. and protest in urban areas, it is understandable that governments might be more responsive to urban demands. But MENA's governments are now seeking to reverse the historical drivers of spatial disparities. The state-led model of economic development and its centralizing policies have been losing steam since the 1990s, for a num- ber of reasons: · The rapid growth of urban population has increased the cost of subsi- dizing metropolitan areas. Urban growth and unemployment have driven policy makers to reduce the incentives for rural-urban migration. · The performance of centrally planned economies worldwide has been disappointing, compared with that of market economies. Policy Package 1. Level the Playing Field and the Opportunity for Human Development in Lagging Areas 141 · In non­oil exporting countries, fiscal constraints have limited the state's ability to continue the subsidization of public employment and production. · Frustration with the poor performance of centralized bureaucracies has led policy makers to seek out alternative models of delivery--for example, quasi-autonomous agencies, public-private partnerships, and of course deconcentration and decentralization. MENA governments are, therefore, now seeking to reverse the historical legacy of disadvantage that peripheral areas' populations have experi- enced in their relationship with the state. The rest of this section will, therefore, examine what MENA policy makers can do and are already doing to create "a level playing field for development." How Differences in the Subnational Investment Climate Can Handicap Lagging Areas The literature on economic geography emphasizes agglomeration economies as an influence over where economic activities take place; entrepreneurs invest where their customers, suppliers, and potential labor force are already clustered. The implication is that low invest- ment in lagging areas is an efficient market response to economies of agglomeration. This implication is true--but it is not the whole story. The invest- ment climate is also an important determinant of the geographical spread of economic activity. The phrase "investment climate" can be defined as the "policy, insti- tutional, and behavioral environment, both present and expected, that influences the returns, and risks, associated with investment" (Stern 2002). It consists of the macroeconomic environment, the institutional context, and the infrastructural base. There is abundant evidence on the relationship between the investment climate and the level of economic activity at the country level. By logical extension, therefore, one would expect subnational disparities in the investment climate to affect the spatial distribution of investment and economic activity. Although all locations within a country share the same macroeconomic environment, spatial differences in the institutional and infrastructural situation are perfectly possible. Box 4.2 recounts a microscale anecdote of a business location decision based on spatial disparities in the investment climate. Within-country differences in the local institutional environment in Bangladesh, China, India, and Pakistan have a clear impact on firm pro- ductivity (Dollar, Hallward-Driemeier, and Mengistae 2005). The authors 142 Poor Places, Thriving People BOX 4.2 An Entrepreneur's Experience of Spatial Disparities in the Business Environment "I wanted to create a small center of Internet and video games. I could provide a little money as personal financing. So I presented an application for credit to banks of the town of Kasserine to finance my project so that I could settle in Thala or Kasserine. Unfortunately, all my requests were refused. I came to Tunis, and I settled in with cousins in Ettadhamoun [a low-income area in Tunis]. I presented the same request, and it was accepted." Source: A 35-year-old unemployed graduate from the lagging area of Kasserine, Tunisia. conclude, "Most of the existing work on the relationship between institu- tions and growth assumes that the important institutions are constant within a country. The empirical link that we establish between investment climate indicators and firm performance . . . reveals that there is signifi- cant variation in the investment climate across locations within countries. So, local governance is important." A study of India (Lall and Mengistae 2005) again finds that within-country differences in the institutional envi- ronment (represented by labor regulations and access to reliable electric power) explain a greater share of the variation in firm productivity than economic geography (agglomeration effects plus market access). Indeed, cities in locations with poor market access seem to receive the biggest pro- ductivity boost from a good institutional environment. Figure 4.2 presents a composite index of the subnational investment cli- mate for Algeria, the Arab Republic of Egypt, Jordan, Morocco, and the West Bank and Gaza. The index uses Investment Climate Survey (ICS) indicators that are common to all economies and are objectively quantifi- able. National-level conditions remain important; the Algerian subnational indicators cluster toward the bottom end of the range, the Jordanian and Palestinian indicators are grouped in midrange, and the Moroccan indica- tors are toward the top. However, the range of within-country investment climate variation is quite large in Algeria. And it is particularly striking for Egypt, where it stretches almost from the bottom to the top of the measured range. It is possible that the range of within-country variation would also have been significantly larger for Morocco had a greater number of locations been surveyed. The data for MENA economies are often insufficient to measure the impact of subnational investment climate disparities on enterprise per- formance. One of the reasons is that enterprise survey responses usually represent a narrow range of locations. Another is that multiple indepen- dent datasets are needed to separate different kinds of location-specific Policy Package 1. Level the Playing Field and the Opportunity for Human Development in Lagging Areas 143 FIGURE 4.2 A Composite Index of Objective Measures of the Subnational Investment Climate in Algeria, the Arab Republic of Egypt, Jordan, Morocco, and the West Bank and Gaza Région Ouargla Région B Boumerdes Qualyubia Région D Oran Région H Ghardaďa Kafr-el-sheikh Gharbiya Région C Blida Région A Alger Beheira Région G Sétif Région F Bordj BA Bani-suef Gaza Strip Port Said Amman Cairo Irbid Région E Annaba West Bank Suez Région Tlemcen Al-Zarqa Minya Alexandria Sharkiya Giza Damietta Menoufiya Rabat/Casablanca Tanger/Tétouan Dakahliya Ismailia Fčs Settat 0 10 20 30 40 50 60 70 80 90 100 composite index Algeria Arab Republic of Egypt Jordan Morocco West Bank and Gaza Source: World Bank staff calculations from World Bank Investment Climate Surveys. Note: The indicators used are the cost of corruption as a percentage of total annual sales, weeks of delay in filling a technical vacancy, percent- age of workforce with less than primary education, percentage of firms making informal payments to government officials, percentage of input value available on credit, weeks of waiting period for a construction permit, weeks of delay in obtaining a water connection, and percentage of firms using own generators. Each indicator is normalized from best (100) to worst (0), and a simple average of the normalized values is taken for each administrative division. Région Ouargla's score is zero. effects. Therefore, this is a potential area for further data collection and research; however, there is suggestive evidence that investment climate disparities are affecting the spatial distribution of economic activity. It is possible to link spatial disparities in the investment climate and spatial disparities in firm productivity. Mengistae and Thompson (2006) used enterprise survey and industrial census data from Morocco to ana- lyze the spatial dispersion of productivity among manufacturing plants. 144 Poor Places, Thriving People Their aim was to separate out three causes of disparities in productivity: natural geography, economies of agglomeration, and the institutional environment. The investment climate variables in the analysis were infrastructural (entrepreneurs' perceptions of electricity, telecommuni- cations, and transport), institutional (corruption, tax administration, the legal system, and access to credit), and educational (availability of skilled labor). Natural geography, agglomeration economies, and the invest- ment climate were all significant, but their relative importance was striking: disparities in the institutional environment explained more variation in firms' productivity than natural geography and agglomera- tion externalities combined. Although there was no clear correlation between leading-lagging and the investment climate effect, the effect on productivity appeared to put nonmetropolitan provinces at a disadvantage for the two sectors ana- lyzed. For food, tobacco, and beverages, the investment climate's effect on productivity was most unfavorable for Taourirt, Fčs, and Tanger-Assilah. For apparel, leather, and footwear the most unfavorable effect was for Berkane, Tanger-Assilah, and Fčs. MENA governments have been concerned to bring financial services into lagging areas. Although commercial banks in MENA13 economies have branch networks that match those in other middle-income regions, such as Latin America and East Asia, MENA has fewer accounts per branch than any other region (table 4.4), even Sub-Saharan Africa and South Asia. This may be because MENA economies have encouraged state-owned banks to open branches as a social mandate. TABLE 4.4 MENA's Commercial Banks Have Extensive Networks with Few Accounts Number of branches Accounts per 1,000 per 100,000 adults, Accounts per branch, adults, commercial banks commercial banks commercial banks Country group (unweighted average) (unweighted average) (unweighted average) Middle East and North Africa 574 14 4,200 Europe and Central Asia 1,282 23 5,600 High income: OECD 1,964 30 6,500 Sub-Saharan Africa 366 5 7,100 Latin America and the Caribbean 1,099 15 7,500 East Asia and the Pacific 1,123 12 9,400 South Asia 608 7 8,900 High income: non-OECD 2,842 21 13,700 Source: World Bank FinStats database. Note: MENA data are from Algeria, Jordan, Lebanon, Morocco, the Syrian Arab Republic, Tunisia, and the Republic of Yemen. Policy Package 1. Level the Playing Field and the Opportunity for Human Development in Lagging Areas 145 How to bring financial services to lagging areas is an important topic that cannot be covered in adequate detail in this report; however, the data presented in table 4.4 suggest that branch density is not the constraint on the geographical reach of the financial sector in MENA. Answers should be sought rather in the development of financial products and transactions technologies that meet the needs of people in poorer and more remote areas. What does the spatial dimension of the investment climate mean for lagging areas? One can break the investment climate­spatial develop- ment linkage into two components: · First, an overall improvement in a country's investment climate at the national level might reduce the spatial concentration of business activity. Although the degree of liberalization of MENA's economies is average for middle-income countries, regulations are applied unequally; MENA economies suffer from abnormally high levels of corruption for their level of national income, and the private sector is dominated by long- standing ties between business and government elites (World Bank 2009d). This situation is likely to make it critical for business owners and senior management to be located close to the capital, where they can maintain face-to-face contacts with senior government figures. Con- versely, greater transparency in the application of business legislation would make it safer for businesses to locate far from the seat of power. · Second, within-country disparities in the application of business reg- ulation might penalize lagging areas. Policy makers should, therefore, take heart that the evolution of their country's economic geography is not entirely predetermined by the forces of agglomeration. The spatial distribution of economic activity is at least partially amenable to policy intervention via the investment climate (see box 4.3). BOX 4.3 Making the Subnational Investment Climate Work for Lagging Areas: An Idea for Policy Makers Survey enterprises in leading and lagging areas to identify · The key factors determining choice of location · The key investment climate handicaps faced by lagging areas Source: Study Team. 146 Poor Places, Thriving People How to Manage the Spatial Dimension of Public Expenditure Public expenditure management (PEM) systems are the transmission mechanisms between the political will to reduce spatial disparities on the one hand and the allocation of resources to disadvantaged territories on the other. Historically, MENA's public expenditure management systems were often set up to centralize and sectoralize the public budget, so they do not automatically lend themselves to managing and monitoring expenditure by location. Now that MENA governments have an increas- ing interest in the spatial dimension of public policy, their spatial politi- cal objectives must be underpinned by accounting and budgeting systems that allow one to manage the spatial incidence of public expenditure. This is not to say that it is efficient to redirect all forms of public expenditure toward lagging areas. On the contrary, investment in eco- nomic infrastructure and productive activities is likely to offer a higher rate of return in leading areas benefiting from economies of agglomera- tion; however, robust spatial data on public funding for social services and public amenities will greatly strengthen any government's efforts to ensure a level playing field for development. This section, therefore, briefly outlines a number of possible approaches to the spatial management of public spending and where a sample of MENA economies stand. To manage the spatial incidence of public expenditure, the PEM sys- tem has to be capable of · Monitoring the spatial incidence of public expenditure, (measuring what is spent where), and measuring its impact · Controlling the spatial incidence of public expenditure (directing resources to political priority locations) Controlling the spatial incidence of spending depends upon monitoring it, which depends, in turn, upon expenditures being apportioned geograph- ically in the public accounts. This is where the challenges begin. The location of public expenditure at subnational administrative levels is, of course, automatically traceable to a specific location. In MENA, how- ever, the share of subnational expenditures is very low compared with the OECD average of 17 percent. This comparison then raises the question of how to estimate the apportionment of central government spending between different locations. · One option is to classify central government expenditure or revenue items or both with a location (zip) code (Jacobs, Hélis, and Bouley 2008). This classification can be done for budgeted expenditure, Policy Package 1. Level the Playing Field and the Opportunity for Human Development in Lagging Areas 147 actual expenditures, or both. Because of the high level of policy inter- est in spatial allocations of public spending, many countries ensure that some form of geographical classification is applied when they modernize their chart of accounts (e.g., Zambia's Activity-Based Budgeting [ABB] system, the Dominican Republic's 2002 manual of budget classifiers, and Mozambique's Budget Framework Law). Often, it is national parliaments that require expenditures to be clas- sified by location (Schiavo-Campo and Tommasi 1999), which is to be expected, given the spatial nature of parliamentary representation. · Another approach--used, for example, in the United Kingdom--is to ask central spending departments to state, after the end of the budget year, what proportion of actual spending benefited the popu- lation of specific regions, estimating the shares according to guide- lines specified by the finance ministry. This disaggregation of central expenditures is then added to subnational government expenditure to give the spatial breakdown of spending. The government of the United Kingdom estimates that about 83 percent of total managed expenditure (TME) can be apportioned geographically in this way, and that the other 17 percent benefits the population of the country as a whole (HM Treasury 2009). The advantage of this approach is that it does not require an overhaul of the chart of accounts. The dis- advantage is that it relies on the objectivity and neutrality of public servants and could be open to challenge, undermining its credibility. · A third option is to use administrative data on service outputs to appor- tion central government expenditures by location. In Madagascar, for example, investment expenditures were managed centrally, as in most MENA economies. So there was no statistical record of spending on school and health facility construction by administrative division; how- ever, it was possible to reconstruct an estimate of how education and health spending was apportioned between districts, by checking when new facilities appeared in the administrative records (World Bank 2004a). Tunisian policy makers have access to data on the allocation of central government spending by governorate. These data show a tendency for public investment per capita to be higher than average in some, but not all, poorer governorates (figure 4.3). However, most MENA economies' public financial management sys- tems currently do not support analysis of the spatial incidence of expen- diture. Not only are subnational expenditures a small share of total expenditures, but also data on the geographical apportionment of cen- tral government spending are often lacking, incomplete, or given lim- ited circulation. 148 Poor Places, Thriving People FIGURE 4.3 Public Investment and per capita Household Expenditure by Governorate in Tunisia 2,000 per capita by governorate (Tunisian dinars) 10th plan (2002­06) public investment 1,800 1,600 1,400 1,200 1,000 800 600 400 200 0 400 500 600 700 800 900 1,000 1,100 1,200 mean per capita household expenditure by governorate, 1990 (Tunisian dinars) Sources: Département de l'Aménagement du Territoire; INS (2000). Note: Each diamond represents a governorate. In countries where the government lacks information on the spatial inci- dence of public spending, there is a risk that spatial development policies explicitly targeting lagging areas are being undermined by an allocation of mainstream spending implicitly skewed toward leading areas. For example, one of the major findings of the Sapir Report (Sapir 2003) was that the European Union's regional policy was being nullified by the concentration of agricultural production subsidies in richer areas. The application of one or several of the methodologies previously discussed could remedy MENA's lack of information on the geography of public expenditure and thus could produce valuable insights for policy makers and citizens. For the moment, therefore, we can analyze only the spatial incidence of subnational public spending, and we must do so in the knowledge that it represents only a fragment of the whole picture. A useful indicator of whether subnational public spending is mitigating or sustaining spatial disparities is to be found in the allocation of transfers from the center to local governments. These so-called "equalization transfers" are usually intended to improve the spatial equity of public expenditure (Shah 2007), helping local administrations in poorer areas to compensate for their rel- atively weak tax base and greater human development needs. They finance about 60 percent of subnational expenditures in developing and transition economies, and about 30 percent in OECD countries. It is, therefore, interesting to ask whether intragovernmental fiscal transfers from the center to local governments are indeed serving this purpose in MENA. Policy Package 1. Level the Playing Field and the Opportunity for Human Development in Lagging Areas 149 Table 4.5 describes the systems for the allocation of transfers in six MENA economies. It was also possible to obtain quantitative data on transfers for three MENA countries: Egypt, the Islamic Republic of Iran, and Jordan. Taken together, they suggest that transfers could be playing more of an equalization role. In most of the six economies described in table 4.5, the allocation procedures tend to be either ad hoc (Egypt, the Islamic Republic of Iran, the West Bank and Gaza) or based upon an area's revenue-generating capacity (Jordan, Tunisia), neither of which would reliably serve an equalization function. The Republic of Yemen, TABLE 4.5 Transfer Characteristics and Allocation Rules in Six MENA Economies Transfer characteristics Source Allocation rules Egypt, Arab General development grants: National budget No rule-based allocation nor Rep. mutual plus earmarked (service formula; discretionary grants and development, land, housing, and donations cleanliness, road) Iran, Islamic Transfers to urban local govern- Oil revenues No formula; allocations decided Rep. ments appear to be negligible, annually by Parliament particularly in the larger cities. Development transfers: 60­70% earmarked, 30­40% discretionary. Jordan Transfers to all municipalities from Fuel taxes 6% of fuel tax generated within central government's fuel tax the municipality collections. Tunisia Earmarked: operating grants National budget Formula based--10% flat rate, through Local Government 45% size of local government, Common Fund; development 45% according to average grants managed by Fund for property tax Loan and Support for Local Government; sector-based transfers; Regional Development transfers. West Bank Current transfers and grants Transport fees and a A formula for the pool and and Gaza for transport fees are not separate account distribution of transport fees earmarked; discretionary and assigned for exists on paper (1997 Local emergency transfers that are discretionary and Government Law). (Yet, the channeled to specific projects emergency transfers formula is only partially applied. Discretionary and emergency transfers are ad hoc.) Yemen, 30% of grants from extra- Shared revenues from No formula. (Council of Ministers Rep. budgetary funds that are 28 taxes, mainly uses nonbinding guidelines-- earmarked; other Zakata population density, financing nonearmarked funds gap, degree of deprivation, per- formance in revenue collection.) Source: Tosun and Yilmaz 2009. Note: a. Religiously ordained charitable giving. 150 Poor Places, Thriving People however, does have nonbinding guidelines intended to steer resources toward poorer areas. The allocation of transfers to subnational administrations was ana- lyzed for three economies for which data were available: Egypt, the Islamic Republic of Iran, and Jordan. Arab Republic of Egypt. Most subnational expenditure in Egypt is exe- cuted by governorates. Governors are appointed by the president, and the governorates are deconcentrated administrations without policy- making power over sectoral issues. Although the governor of Alexandria was able to translate his city's robust economic base into a significant fis- cal resource (World Bank 2005c), governorates' own resources are usu- ally slight. Therefore, most governorates are dependent for investment resources upon their share of the block grant transfers administered by the Ministry of Local Development. These transfers have the potential to reduce disparities between governorates, by giving the highest trans- fers to governorates with the weakest economic base; however, the data from fiscal year 2002/03 (figure 4.4) suggest that resource transfers and governorate GDP per capita were not correlated. It should be noted, however, that the arrangements for block grants allocations to gover- norates are under review. Islamic Republic of Iran. Transfers from the central to local governments in the Islamic Republic of Iran show a contrasting picture. The sub- national administration is organized primarily at the provincial (ostan) FIGURE 4.4 Transfers from the Center to Governorates Are Not Reducing Spatial Disparities in the Arab Republic of Egypt 94 transfers as % of expenditure 92 90 February 2003 88 86 84 82 0 2,000 4,000 GDP per capita January 2002 (PPP US$) Source: World Bank 2006b. Note: PPP = purchasing power parity. Each diamond represents a governorate. Policy Package 1. Level the Playing Field and the Opportunity for Human Development in Lagging Areas 151 level. The apportionment of transfers from the central government to ostans does give priority to lagging areas. On a per capita basis, the trans- fers ranged in 2002 from 1.2 million rials for Ilam to under 0.1 million rials for Qazvin and Tehran (figure 4.5). An ostan's level of development does not explain all of this wide variation, but it does explain an important part of it. The correlation coefficient between per capita transfers and ostans' Human Development Index was ­0.42. With a Life Expectancy Index, it was ­0.53; with an Income Index, it was ­0.46. Jordan. Municipalities in Jordan, like Egypt's governorates, have limited financial independence: local government expenditures are only 6 percent of total government spending (FEMISE 2005); how- ever, there are significant variations in public expenditures per capita among municipalities, both between and within regions (figure 4.6). These differences are seemingly not associated with spatial disparities. It is clear from the previous discussion that most MENA economies can make public financial management a more powerful tool for their spatial development policy: first, by putting in place systems to moni- tor the spatial incidence of public expenditure; and second, by moving toward a more rules- and needs-based system for intragovernmental fiscal transfers. FIGURE 4.5 Much Variation, but a Tendency for the Intergovernmental Transfers to Favor Lagging Ostans in the Islamic Republic of Iran 1,400,000 per capita transfer amount, 2002 1,200,000 1,000,000 (Iranian rials) 800,000 600,000 400,000 200,000 0 Ko istan an es am n an rs st Za m Es ilan Te in ar n Ke Ahm il ak n an Kh Ila i M tan an za an Se san Bo Ar tan all K zgan Go azd Lo hah Bu ran G r rm an uz m Qa n z Kh hari eh an d W H sta Az nja b d B ma an ka ha az Fa Qo zv aij rm a ye de mn hr t A ad Ho baij les es Y da sh a s ar ist fa re h an er o rb or luc rd Ba r M nd Ea na nd ah sta ha M Si ye ar ilu ah hg Ch Ko lowest Income Index (2002) highest Source: World Bank 2005d. 152 Poor Places, Thriving People FIGURE 4.6 Variation in per Capita Expenditure of Local Government by Governorate in Jordan 250 expenditure by governorate 2007 Karak per capita local government 200 (Jordanian dinars) 150 Mafraq 100 50 0 150 170 190 210 230 250 270 per capita household consumption by governorate 2002 (Jordanian dinars) Sources: http://www.moma.gov.jo/municipalDB/landuse.aspx, accessed August 24, 2009; 2002 national household survey. Note: Amman is not shown in the figure. The data for Amman governorate excluded Amman municipal- ity. Separate data on Amman municipality were not available. Each diamond represents a governorate. How to Correct Spatial Disparities in Public Education Systems "In spite of the relative improvement of transportation, I knew that my children could not do better than me if we remained in the same area; therefore it was necessary to go down [from the mountain] to the city for the children's education." --A parent from Kasserine, a lagging area town in Tunisia Education, as we have seen, is critical to many pathways for the reduc- tion of spatial disparities. First, convergence in educational attainment leads to convergence in human capital and in earnings. Second, educated people are more mobile and can participate in the redistribution of the population from lagging to leading areas. Third, the education of women unlocks the economic potential of 50 percent of the population and helps close the spatial disparities in fertility that have contributed to the cur- rent spatial disparities in well-being. Fourth, today's children in lagging areas are tomorrow's parents, and they will pass their learning on to the next generation. Closing the spatial gap in education is, therefore, a win- win-win-win strategy. This section puts the spotlight on education by asking three questions: · What are the spatial disparities in educational access and attainment in MENA? Policy Package 1. Level the Playing Field and the Opportunity for Human Development in Lagging Areas 153 · To what extent are spatial disparities in education attributable to dis- parities in service provision? · What are MENA governments doing to narrow spatial disparities in education, and what is the way forward? The analysis of spatial disparities in education draws on country case studies of Egypt, the Islamic Republic of Iran, Jordan, Morocco, and the Republic of Yemen. There Are Significant Spatial Disparities in Enrollment, Attainment, and Achievement Rural-urban disparities in primary and secondary enrollment rates are an obvious feature of MENA's education indicators (table 4.6). The nonpoor and students living in urban areas tend to have higher access to education than the poor and those who live in rural areas (World Bank 2008a). There are also notable differences in educational attainment between rural and urban in Egypt, Jordan, and Morocco. Regarding the completion of primary education, the disparity between rural and urban has been nar- rowing in Egypt and Jordan, while a discernible spatial difference remains in Morocco (figure 4.7). In Morocco, boys' grade repetition rates in rural primary schools are much higher (17 percent) than those of urban girls (10 percent) (World Bank 2007b). Regarding completion rates, spatial dif- ferences are higher for secondary and tertiary education than for primary education in Egypt and Morocco (figures 4.7 and 4.8). The Trends in International Mathematics and Science Study (TIMSS) allows us to gauge spatial disparities in achievement. TABLE 4.6 Lower Enrollment Rates for Rural Children, Especially the Poor (Percent) Urban Rural Poor Nonpoor Poor Nonpoor Algeria 1995 Primary 96.0 95.0 89.0 89.0 Secondary 77.0 82.0 59.0 66.0 Egypt, Arab Rep. 1999 6­15 95.8 98.5 93.5 96.7 15­19 72.4 84.9 64.7 72.9 Iran, Islamic Rep. 2001 6­10 99.0 100.0 98.0 98.0 11­13 92.0 97.0 76.0 84.0 Morocco 1998 7­15 69.4 87.2 36.4 49.8 Tunisia 2000 6­18 79.4 82.2 67.0 70.7 Yemen, Rep. 1998 10­14 83.0 92.1 59.6 62.0 Source: World Bank 2008a. 154 Poor Places, Thriving People FIGURE 4.7 Percentage of Adults Older Than Age 25 Who Completed Primary Education 80 60 percent 40 20 0 95 00 03 97 02 92 4 ­0 19 20 20 19 20 19 03 p. p. p. an an co 20 Re Re Re oc rd rd co Jo Jo or ab ab ab oc M Ar Ar Ar or t, t, t, M yp yp yp Eg Eg Eg economy rural urban Source: EdStats, World Bank. FIGURE 4.8 Percentage of Adults Older Than Age 25 Who Completed Secondary Education 60 40 percent 20 0 03 04 5 0 2 9 00 99 20 3­ 19 .2 o1 p. 0 . ep ep 20 Re c oc bR bR co ab or oc ra ra M Ar ,A ,A or t, M pt pt yp y y Eg Eg Eg economy rural urban Source: EdStats, World Bank. Policy Package 1. Level the Playing Field and the Opportunity for Human Development in Lagging Areas 155 MENA economies show rural-urban disparities in mathematics scores, with statistically a significant gap in Egypt, the Islamic Republic of Iran, Jordan, Lebanon, and Tunisia. The Islamic Republic of Iran's rural- urban disparity in the mathematics score is particularly striking. More- over, the findings of the Program for International Student Assessment (PISA) suggest that being at a rural school is more problematic for low- achieving students than for high-achieving students in Jordan (World Bank 2008b). Gender gaps in literacy are often widest in lagging areas. In Morocco, the urban literacy rate is 75 percent for women, while it is 14.2 percent for rural women. Being male in a nonrural setting increases the proba- bility of completing primary education by 7 percent, while being male in rural areas increases it by 16 percent. The primary completion urban and rural gender gaps are 5 percent and 9 percent, respectively, in Jordan. Most Spatial Disparities in Educational Attainment (Completion of School) Are Attributable to the Household Characteristics of the Rural Population3 If children in rural areas receive less schooling (figure 4.9) than do their urban counterparts, the next question to ask is to what extent is this lack due to spatial disparities in children's socioeconomic status (demand-side FIGURE 4.9 Percentage of Adults Older Than Age 25 Who Completed Tertiary Education 30 20 percent 10 0 00 3 4 92 95 00 ­0 20 19 19 .2 3 00 p. co ep p. Re o2 Re oc bR or ab c ab ra oc M Ar ,A Ar or t, M pt t, yp yp y Eg Eg Eg economy rural urban Source: EdStats, World Bank. 156 Poor Places, Thriving People factors) and how much is due to spatial disparities in education services (supply-side factors). The appropriate policy response will depend on the answer to this question. Where some or most of the rural-urban dis- parity can be attributed to inequalities in access to quality educational services, then the case for improving education provision in lagging areas is strongest. This study examined the determinants of educational attainment in one country with high spatial disparities in attainment (Morocco) and another with low disparities (Jordan). Morocco. For Morocco, our study (Katayama 2009) used data from the Population and Family Health Survey to examine what deter- mines who completes primary education. The rural-urban difference was 33 percentage points, but how much of this gap was because rural children tend to be poorer and would have had worse educational chances wherever they lived? The analysis revealed that a child's household's economic status was the most important factor; being in the lowest wealth quintile reduces the probability of having com- pleted primary school by 42 percentage points compared with some- one in the highest wealth quintile. The next most important factor was gender: being female reduces the probability of having completed primary school by 16 percentage points compared with males. After the respondent's age and the household head's level of education were taken into account, the rural-urban factor explained a gap of only 5 percentage points in the probability of having completed primary education. The analysis was repeated for secondary education completion, for which there is an urban-rural gap of 13 percentage points. This dispar- ity was found to be explained entirely by household characteristics. Finally, regarding higher education completion, there was a 10­ percentage point rural-urban disparity, and the gap was also signifi- cantly attributed to household characteristics. Jordan. In Jordan, urban-rural gaps in educational attainment rates were found to be relatively small: 5 percentage points for primary edu- cation and 11 percentage points for secondary. As in Morocco, house- hold economic status was the main factor behind a person's probability of having completed an educational level. Someone from the poorest wealth quintile was 17 percentage points less likely to have completed pri- mary education than someone from the richest quintile. The gender gap effect was next most important, with a 3­percentage point lower proba- bility for girls than for boys, and was bigger in rural areas. After house- hold characteristics were taken into account, the rural-urban gap was insignificant for the probability of having completed both primary and secondary school. Policy Package 1. Level the Playing Field and the Opportunity for Human Development in Lagging Areas 157 Spatial Disparities in Learning Achievement Are Caused by Inequality in Education Provision Spatial disparities in learning achievement are largely due to geographic differences in schooling (figure 4.10). Jordan and the Islamic Republic of Iran. In both Jordan and the Islamic Republic of Iran, 2007 TIMSS test scores are significantly lower in rural areas for both science and mathematics. This gap remains significant even after taking account of rural-urban differences in the student or family characteristics described in the TIMSS data.4 Therefore, it is probable that students in rural schools are studying in a less favorable learning environment than those in urban schools. Teachers in rural schools, for example, tend to be younger and less experienced. Jordanian mathematics and science teachers have 1.1 years' and 3.1 years' less experience (respectively) in rural areas than in urban areas. In the Islamic Republic of Iran, the experience gap is 4.5 years and 3.7 years. The World Bank (World Bank 2008a) noted MENA's spectacular achievements in increasing enrollment: "For good reasons, the region initially focused on establishing mass education systems by building schools, recruiting teachers, producing textbooks and setting the FIGURE 4.10 Rural-Urban Disparities in Mathematics Learning, 2007 TIMSS Mathematics Scores 500 2007 TIMSS mathematics scores 450 400 350 . . ria an n o lic za a p p no cc isi Re Re Ga ub rd ge o n ba Jo or ab ic ep Tu Al d Le lam M an Ar bR k t, Is ra an yp n, nA tB Eg Ira ria es W Sy MENA region rural non-rural Source: World Bank staff calculation, based on 2007 TIMSS results. Note: Urban is defined as a location with a population greater than 15,000. 158 Poor Places, Thriving People curriculum." Nevertheless, MENA has been less effective than com- parator and competitor countries in converting these investments and enrollments into retention and learning (World Bank 2008c, 3): "The edu- cational achievements are compromised in part by high dropout rates, and relatively low scores on international tests. Literacy rates remain low." The idea that spatial disparities in learning achievement might be largely due to geographical differences in the quality of education reflects the same diagnosis. Policy Implications: Analysis, Demand-Side Interventions, Quality, and a Focus on Girls These analyses have important implications for MENA policy makers seeking to reduce spatial disparities in education: Analysis of education spatially. First, on the supply side, there is scope across the region to improve the spatial planning of education infra- structure and to use information technology to bridge physical dis- tances. Spatial analysis of the determinants of educational performance is also critical. Although spatial disparities in educational indicators are glaringly obvious, the causes of these disparities may be subtler. A reac- tive policy, for example, of school construction in rural areas may be treating the symptoms rather than the cause of the problem if the real bottleneck is on the demand side or related to the quality of education. Demand-side interventions. Second, spatial disparities in educational attainment--the highest degree of education that a person completes-- may have more to do with spatial disparities in households than with the spatial disparities in the availability of educational services per se. Pol- icy interventions to motivate poor households to keep their children in school may, therefore, be a more appropriate response than a focus on educational infrastructure. In Morocco, for example, the government's Conditional Cash Transfer programs will address the costs for poor par- ents of keeping their children in school (World Bank 2008c). Quality. Rural children seem to be learning less at school than their urban counterparts, even when they come from similar socioeconomic backgrounds. It is, therefore, critical to address quality issues such as learning environment, teacher training, motivation, and career man- agement in rural areas. School development initiatives and human resource management reforms have the potential to reduce spatial dif- ferentials in quality. A focus on girls. Reducing gender disparities in education, particularly in completion rates, may contribute to reducing spatial disparities in development. This is because girls' education contributes to MENA's Policy Package 1. Level the Playing Field and the Opportunity for Human Development in Lagging Areas 159 demographic transition and because gender disparities in education are wider in lagging areas. Attacking Spatial Disparities: What MENA Governments Are Doing, and Priority Strategies On the basis of the considerations discussed earlier, four priority axes can be proposed to combat spatial disparities in education: · Reducing spatial disparities in quality through human resource man- agement innovations and school development initiatives · Stimulating the demand for education in lagging areas · Reducing gender disparities in lagging areas · Using spatial planning techniques to deploy appropriate infrastructure These priority axes will be illustrated with examples from the country case studies on Egypt, Jordan, Morocco, and the Republic of Yemen. Reducing Spatial Disparities in Quality through Human Resource Management Innovations and School Development Initiatives The spatial deployment of teachers and administrators creates inequities in the quality of education. For instance, the TIMSS results of the Islamic Republic of Iran have showed that more experienced teachers deliver better educational outcomes, but the tendency is for rural areas to be allocated the least experienced. In Morocco, teachers generally move from rural to urban areas as they acquire more experience. A 2007 survey of teachers in Morocco found that difficult living conditions in rural areas led young teachers to seek a change of posting at the first opportunity. This lack of motivation to work in rural schools is reflected in the high level of unexplained absenteeism in rural areas, which in 2007 was the equivalent of 417 teachers being absent all year. The same survey found that underutiliza- tion of teachers was particularly acute for small rural lower secondary schools, with teachers working as few as 16 of the legal requirement of 24 hours--already low when compared with international levels (e.g., 37.5 hours in Spain). It is probable that rural principals are less experi- enced than their urban counterparts, compounding the motivational problems of young teachers. In Jordan, teaching posts are allocated on the principle of queuing; those who have waited longest get the next job in line. In Egypt, as in Morocco, teachers migrate from rural to urban areas as they gain expe- rience. In the Republic of Yemen, teaching placements are somewhat haphazard; a comparison of Annual Education Survey data with payroll 160 Poor Places, Thriving People found discrepancies for 30,000 teachers. Teachers are hired on the basis of local preference. The relocation subsidy is poorly regulated and often received long after redeployment. In this rather confused environment, an uninfluential school in a lagging area is unlikely to obtain experienced teachers. Improved human resource management in public education services is, therefore, critically important for the reduction of spatial disparities in education quality. First, improved incentives are required for teachers to locate in lag- ging areas. In the Republic of Yemen, for example, living conditions are reported to be extremely harsh and restrictive for female teachers posted away from their place of origin. Incentives may involve improved living conditions and housing, monetary payments, or improved career prospects. Box 4.4 describes how Algeria has used salary incentives to meet skills shortages in lagging areas. Motivating first-rate principals to accept postings in lagging areas is a key step toward improving education quality, and targeting incentives on such individuals could have a dispro- portionately positive impact. Second, more formal, systematic, and needs-based postings proce- dures are required, to match teachers to vacancies. Where informality and influence characterize posting arrangements, the default outcome will be for disadvantaged areas to be further disadvantaged in teacher allocations. Arrangements for recruiting and training local teachers to BOX 4.4 Algeria Introduces Spatial Salary Incentives, with Mixed Results The Algerian government provides salary incentives for skilled public sector workers to take up employment in certain lagging areas of the High Plateaus and the South: Djelfa, Laghouat, Naama, El Bayadh, Béchar, Ouargla, Ghardaia, El Oued, and Biskra. A salary supplement is calculated as a percentage of the employee's basic salary: 10­20 percent for primary school principals and secondary school teachers, 20­50 per- cent for senior civil service administrators, and 100­120 percent for specialist doctors and university teachers. The initiative has been most successful for primary school principals, secondary school teachers, and administrators. The targeted regions are now able to fill these posts with relatively little difficulty. Their medical specialist and university teacher vacancies, however, are proving harder to fill, despite the high salary supplementation on offer. Quality of life considerations seem to be the main constraint. Source: Khelladi 2009. Policy Package 1. Level the Playing Field and the Opportunity for Human Development in Lagging Areas 161 work in their home communities could help overcome spatial biases in teacher careers, and would have the added benefit of increasing the num- bers of female teachers and students (Golnar 1995). School Development Initiatives (SDIs) are development plans man- aged at the school level by school boards, teachers, and parents. They have demonstrated great potential as an instrument for improving edu- cation quality in targeted geographic areas. SDIs owe their origin to the Anglo-Saxon and Nordic tradition of school boards or committees, representative of parents and the commu- nity, taking an active role in the management of schools. In the United States, for example, many school boards play a part in selecting school principals, teachers, and even books. In developing countries, concerns over the ability of central ministries to manage local expenditures have led to the delegation of spending oversight to parents, teachers, and communities. This change created opportunities for spatially targeted grants to schools with unmet needs. Of course, the introduction of the SDI approach requires a sound institutional platform. It relies on the capacity and motivation of the local school governance committee. School grant programs require fraud- resistant operational and financial management procedures. SDIs also presuppose that school principals have the skills and motivation to act as leaders in the community, not just as the most senior teacher. The SDI model, therefore, requires pragmatic awareness of the risks involved in countries where school governance remains centralized (for example, Egypt) or subject to the influence of local elites (for example, the Repub- lic of Yemen). In Jordan, a National School Development Plan (NDSP) has been launched, in line with the SDI approach. Teachers and parents will eval- uate their schools, identifying strengths and weaknesses, and draw up development plans incorporating specific measures--new teaching pro- files, better equipment, more books, or changes in school activities. They will receive and expend funds, with oversight by parent-teacher associa- tions and community representatives. The goal is also to improve reten- tion. (Currently, 30 percent of students drop out before grade 12.) In Morocco, the School Quality Initiative, which is based on school manage- ment committees ("conseils de gestion"), is another exciting development. Reducing Spatial Disparities through Demand-Side Interventions As we saw earlier, the low economic status of children in lagging areas may be more of an obstacle to lagging areas' educational attainment than disparities in educational service supply. Poorer households are least able to afford the direct costs of education (fees, uniforms, books, transport, 162 Poor Places, Thriving People and so on) and the opportunity cost of the child's labor time. One strat- egy for reducing spatial disparities in education is, therefore, to intervene on the demand side, making it economically more attractive for families to keep their children, and especially their girls, in school. In conditional cash transfer (CCT) programs, the transfer of a welfare benefit depends upon the receiver's actions. Requirements may involve enrolling children in public education. Unlike many welfare programs, CCTs' impact has been rigorously evaluated through experimental or quasi-experimental methods. There is considerable evidence that CCTs have been an effective way of redistributing income to the poor (Fiszbein and Schady 2009). CCT programs have potential as an instrument of spatial education policy. Geographic targeting is a common feature (Rawlings and Rubio 2005). Mexico constructs a "marginality index" from census data to iden- tify which communities will participate. In Honduras, beneficiary com- munities are identified from a census of the height-for-age of first-grade schoolchildren. In Nicaragua, departments and municipalities with the highest poverty incidence are included. In Colombia, the geographic tar- geting focuses on city status, with cities under 100,000 being selected and department capitals being excluded. The education components of CCT programs address demand-side obstacles to education by contributing toward the costs of sending a child to school. In Mexico, Honduras, and Turkey, governments have allowed for the opportunity cost of the child's labor as well as the direct cost of fees, school supplies, and travel. In Colombia and Mexico, the grants are higher for secondary-school children than for primary-school children, reflecting the fact that older children are more productive as workers. CCT programs' positive impacts upon education enrollment have been clear. In Mexico, the program led to an increase in primary-school enrollment ranging from 0.7 to 1.1 percentage points for boys and from 1.0 to 1.5 percentage points for girls. The impacts at secondary level were much bigger, ranging from 3.5 to 5.8 percentage points for boys and from 7.2 to 9.3 percentage points for girls. Colombia's program raised secondary-school enrollment rates by 5.5 percentage points in rural areas and 14 percentage points in urban areas. The Nicaraguan scheme led to improved attendance, with a 30 percent increase in the share of children with fewer than six unexcused absences in a two-month period. The government of Morocco has launched a pilot CCT to encourage school attendance. It involves 266 rural schools with 88,000 children, serving 49,000 households. The high level of participation is due to the government's strong communications effort. Children's attendance at school is verified through fingerprint recorders. Households receive a check either through their local post office or by postal delivery to the Policy Package 1. Level the Playing Field and the Opportunity for Human Development in Lagging Areas 163 primary school. A mobile team of monitors ("vérificateurs") double- checks attendance. There have been start-up problems: incorrect house- hold registration data, delays in implementation of the Management Information System, and delays in providing the monitors with training and transport. However, these problems are being resolved, and there is already anecdotal evidence that the program is having a positive impact upon the retention of rural girls. Reducing Gender Disparities in Education in Lagging Areas "The fact that we don't have a high school in Tanougha has hampered the education of girls, because their parents are not willing to send them, via bus, to the neighbouring village. The girls would indeed have to walk back home after 6 p.m., since the bus is very old and often breaks down." --Mothers of Ait Hammi (quoted in World Bank 2007a) As we have seen, the effects of gender disparities and spatial disparities combine to create especially wide gender gaps in lagging areas. In Morocco, for example, the gender gap for literacy is 5 percentage points in urban areas and a staggering 35 percentage points in rural areas. Many causes of the gender gap in education are linked to social constraints on girls' mobility. In low-density rural areas, long distances from home to school and the lack of transportation are known to be a particular imped- iment to the girls of Algeria, Egypt, the Islamic Republic of Iran, Iraq, and Tunisia. In Upper Egypt, at least half the secondary schools are more than two kilometers from their catchment villages, with a negative impact on girls' attendance. (Upper Egypt accounts for 65 percent of girls dropping out of education.) In the Islamic Republic of Iran, the lack of boarding facilities discourages the enrollment of girls from nomadic families. The absence of upper elementary classes in remote villages has a specific impact on girls, if they are not allowed to continue schooling away from home. Living conditions in lagging areas may be particularly unaccept- able to female teachers. The cultural isolation of rural areas and the short- age of educated mothers and other female role models may lead parents to be especially dismissive of the value of girls' education. Children from ethnic minorities in peripheral areas may not be fully familiar with the language of instruction. Finally, the school calendar in many MENA economies matches the rhythm of life in the cities, without reflecting the agricultural cycle and other demands for female labor (Golnar 1995). Reducing spatial disparities in girls' education will remove one of the causes of spatial disparities in fertility rates. Reducing spatial disparities in fertility rates will, in turn, reduce spatial disparities in dependency 164 Poor Places, Thriving People ratios, which, as we have seen, are one of the prime causes of spatial dis- parities in well-being. Figure 4.11 shows how education and location affect women's fertility rates in Egypt. In the poorer governorates of Upper Egypt, such as Assiut, Sohag, Fayoum, and Menia, the difference between illiteracy and secondary-school completion is associated with a reduction in the total fertility rate of 1.5 children per woman. It is also worth noting that the returns to this level of education in terms of fertility reduction (the gap between the top line and the fourth line) is around twice as high in these lagging governorates of Upper Egypt as in the leading governorates of the country. One can compare the total fertility rate (TFR) of lagging Fayoum in Upper Egypt with better-off Ismailia in Lower Egypt, on the west bank of the Suez Canal. The TFR in Fayoum is 0.96 higher than in Ismailia, meaning that Fayoum's women are expected to have, on average, approx- imately one more child during their lifetime than the women of Ismailia. Using the data in figure 4.11, however, one can estimate that the spatial fertility gap between Ismailia and Fayoum could have been reduced to FIGURE 4.11 Women's Education Reduces Fertility Rates in the Arab Republic of Egypt 5.0 4.5 total fertility rate 4.0 3.5 3.0 2.5 2.0 So t ni g Fa ef M m Qu ia Sh na en a Da ufia Be lia Ka As ira El- n Gh ikh lyo a Ism bia ia Da za ex ta ia ez Po airo id siu Be ha M ki fr wa Ka bi u en ail dr Sa Su Al iet Gi Su e h he ar e yo ar u As ka o an C sh m rt city lagging leading illiterate reading and writing basic education complete secondary education tertiary education Sources: Goujon, Alkitkat, Lutz, and Prommer 2007; UNDP 2005. Policy Package 1. Level the Playing Field and the Opportunity for Human Development in Lagging Areas 165 0.65 if Fayoum's women of childbearing age had the same level of edu- cation as Ismailia's. The key policy principle for tackling spatial-cum-gender disparities in education is that any intervention aimed at spatial disparities should be made with girls' distinct needs in mind: · In the Republic of Yemen, the human resource management agenda is gender-specific: the Girls' Secondary Education Project will subsi- dize transport costs for woman teachers. · CCT programs can be designed to provide families with special incentives to keep girls in school. In Mexico, grants at the secondary level are higher for girls, to take account of women's role in passing on human capital to the next generation (Rawlings and Rubio 2005). In the Republic of Yemen, the Girls' Secondary Education Project plans to increase girls' retention rates by providing CCTs to families with girls in grades 4 to 12, inclusive. · The spatial planning of public investment in education can reflect girls' educational needs. Standard distances between schools and res- idences should take account of social limitations on girls' mobility. Schools should be provided with adequate water and sanitation, which are a major factor in determining girls' attendance. Information tech- nology can be used to mitigate limitations on the mobility of female students and teachers. Although the evidence suggests that boys tend to monopolize computers in schools at the expense of girls, there are successful examples of computer-based learning programs specifically targeted at girls and women (Huyer, Hafkin, Ertl, and Dryburgh 2005). The Republic of Yemen's Girls Secondary Education Project allows the best women teachers to reach out to selected project schools through videoconferencing and e-mail to provide personal instruction. Bringing dropped-out girls back into education may require nontraditional approaches. Field studies report the sense of "sadness, anxiety, boredom, and uselessness" felt by girls who have dropped out of education (Golnar 1995). Complementary programs can be designed to help girls who have dropped out of formal education to return to basic education or skills train- ing. In Bangladesh, the nongovernmental organization (NGO) BRAC pio- neered this approach, renting houses, courtyards, and storefronts as improvised schools. It has been suggested that women's training annexes could be added to primary schools in the Republic of Yemen, which was the practice in France in the nineteenth and early twentieth centuries. The "Bridge School" approach, which has proved successful in India, creates a pathway for girls to transition back into mainstream public education. 166 Poor Places, Thriving People Using Spatial Planning Techniques to Improve the Spatial Efficiency of Infrastructure The race to achieve the Millennium Development Goal of full primary- level completion has given strong emphasis to school construction across the MENA region; however, insufficient attention has been given to spa- tial planning. · As the goal of universal primary enrollment and completion comes close to being achieved, the focus of attention is shifting from pri- mary to secondary and from lower secondary to upper secondary. Secondary schools are significantly more expensive than primary schools, and they must be in full use for 30 years to amortize their cost. Moreover, secondary schools are fewer in number than primary schools, so distance from home to class has an even greater impact upon enrollments and retention. Therefore, deciding where to put secondary schools is an increasingly important spatial policy issue. · Population distributions are changing rapidly with urbanization and the spread of the demographic transition from leading to lagging areas. The Republic of Yemen, for example, has the second-highest total fertility rate in the world, the fastest growing city in the world, Sana'a (Economist 2005), and a shift in the demographic center of gravity from the highlands toward the coastal cities. The spatial allo- cation of education assets therefore has to reflect not only the current needs but also the demography of future decades. · There are powerful and user-friendly Geographical Information Sys- tems (GIS), which enable the optimization of school catchment areas with respect to population trends. For example, the Republic of Yemen faces a number of spatial imbal- ances in education. Twenty-five percent of rural schools had fewer teachers than the prescribed minimum, while about 205 urban schools had overcrowded classes in the "unmanageable" range of 70­150 chil- dren per class. Only 21 percent of secondary schools met the minimum school size of 180 students specified by the Ministry of Education. Con- sequently, there were not enough classes for the full range of subjects to be taught; only half the schools had a science stream. Therefore, the government is planning for future needs using GIS. An assessment of all existing first-level schools (grades 1­6) identified their physical condi- tion and the needs for maintenance, renovation, or reconstruction. The survey also identified the potential for building additional classrooms on the current site. The assessment of building requirements took account of projected growth of the population ages 6 to 14, by village. Geo-ref- erenced GIS maps were developed for potential new school sites and for Policy Package 1. Level the Playing Field and the Opportunity for Human Development in Lagging Areas 167 catchment areas, based on the estimated number of children within a 2-kilometer radius of each potential location, and adjusting distances for the difficulty of terrain as measured by slope. Despite Egypt's high population densities, there are significant spa- tial inequalities in access to educational infrastructure. A study of mother villages and satellite villages in the governorate of Sohag revealed that while one satellite village had a complete basic education school (grades 1­9), children in another village had to travel 2 kilome- ters. In the case of one satellite village, the nearest secondary school (grades 10­12) was 5 kilometers away, while it was 10 kilometers away in another. Fortunately, the General Authority for Education Buildings (GAEB) now has the tools to analyze the situation in the form of a geo- referenced GIS database of all schools in Egypt. It has the technical capacity to rationalize the catchment areas of upper basic and secondary schools, taking into account population trends, before any major new construction program is undertaken. Jordan is another illustration of the potential benefits of a spatially based planning approach for education. The country suffers from a seri- ous spatial disequilibrium in the distribution of basic and secondary schools. The current school map results in high maintenance costs and low utilization rates. There is an excess capacity of 172,950 places affect- ing 59 percent of schools. Many recently constructed schools are located in catchment areas with excess capacity, sometimes within 1 kilometer of another school. At the same time, many schools lack the specialized classrooms and workshops required to deliver instruction in subjects such as science, technology, computing, and the arts. Despite the con- struction of new facilities, there is a US$36.4 million backlog of expen- diture on existing schools, which has led to 42 schools being declared "unsafe." A GIS mapping exercise, along the lines conducted for basic education in the Republic of Yemen, could develop maps of current catchment areas, their "overlap" and likely demographic trends. This exercise would make it possible to identify which schools should be con- solidated, which should be expanded, which should be split, and which should be provided with temporary shelter. Morocco provides a similar example of spatial imbalance in which a GIS map of primary education could provide efficiency savings. While enroll- ments are mainly concentrated in the 12 northern provinces, many schools in mountainous areas and the south do not have their full complement of students, leaving them with less than the full six primary grades. This lack of students creates problems for families in poorer areas, who may have to split their children between two schools or face long journeys to school. GIS analysis of catchment area populations and travel times could inform decisions on whether to expand, relocate, or consolidate such schools. GIS 168 Poor Places, Thriving People approaches could also be used to assist decisions on the construction or consolidation of lower secondary schools and "collčges de proximité." GIS could provide a rational basis for deciding whether new education facilities or providing transport to school is the most cost-efficient solution for improving access to secondary education. Information and communications technology (ICT) has the potential to be a "spatial equalizer," abolishing the handicaps of distance experienced by students in low-density, lagging areas (box 4.5). School principals and teachers can obtain quality, standardized curricula, model lessons, advice, and in-service training. ICT can also deliver exams and standardized tests with integrity and efficiency. BOX 4.5 Leveling Spatial Disparities in Education: Nine Ideas for Policy Makers · Closing the spatial gap in education is a paramount strategy for reducing spatial dis- parities in development. · Analyzing subnational education indicators to diagnose the causes of spatial dispari- ties in education. Building more schools in lagging areas is not always the answer. Many existing facilities in low-density areas are underused. The quality of teaching and demand-side issues might be more of a problem for lagging areas than the num- ber of schools. · Using spatially targeted conditional cash transfers to encourage lagging-area parents to keep their children in school. · Systematizing the arrangements for teacher postings so that more good, experienced teachers go to lagging-area schools. · Identifying why experienced teachers do not want postings in lagging areas and put- ting incentives in place--for example, salary supplementation, career advantages, housing, and transport. · Recruiting teachers locally for lagging areas. · Making girls and women teachers a priority in all strategies to address spatial dispar- ities in education. · Using GIS technology and population projections to improve the spatial efficiency of school infrastructure. · Using ICT to improve teaching and curriculum quality in lagging area schools. Source: Study Team. Policy Package 1. Level the Playing Field and the Opportunity for Human Development in Lagging Areas 169 · In Egypt, the Ministries of Education (MoE) and Communications and Information Technology (MCIT) have launched the Egyptian Education Initiative (EEI). The EEI is a public-private partnership between government, the World Economic Forum's ICT member community, and various multinational corporations. It has two "tracks." The pre-university track aims to build capacity of administra- tors and teachers and improve the delivery of teaching materials. The MCIT plans to use fiber optics and high-speed satellite connections to address locational disadvantages by bringing these services into mother and satellite villages. The higher education track goes one step further, creating electronic content for students in English, information technology, the sciences, and "survival skills" (personal health, basic finance, and statistics). The higher education track also encompasses the development of new management information sys- tems (accreditation and evaluation, personnel management, program budgeting, and infrastructure management). · The Jordan Education Initiative (JEI) is a similar public-private part- nership, involving Cisco and Intel and the networking of schools and universities. The JEI aims to deploy a new outcomes-based grade 1­12 curriculum, supported by online capacity building for teachers and principals. · In March 2005, the Moroccan Ministry of Education adopted a strat- egy of generalizing ICT to all public schools by 2008, with the aim of improving the quality of teaching. ICT is to be used to deliver teacher training and improved curricula. The program includes the creation of a national laboratory for curriculum development and a national education Internet portal. The portal will offer teachers access to educational resources, a virtual library, discussion forums, a search engine, and e-mail. The overall program comes at a total cost of around $11 million. Health: Correcting Spatial Disparities in Access to Care Geographical differences in health status are both direct and indirect con- cerns for policy makers: MENA13 governments are spending an average of 3.3 percent of GDP on health services, and the impact of this spending in terms of health status should not depend on where a person lives. But spatial disparities in health also have important indirect policy signifi- cance. They can encourage migration from peripheral areas to towns in search of better health and social services (Lall, Timmins, and Yu 2009). 170 Poor Places, Thriving People In addition, poor access to health services in lagging areas can contribute to high fertility rates, and thus to high dependency ratios and poverty. Moreover, the performance of health systems in the MENA region will increasingly be measured in terms of their ability to reduce dispari- ties. As the coverage levels for basic health services approach 100 percent, it will be possible to raise national health indicators only by targeting the worst off. Health Services and Health Outcomes in Poor Places There are large spatial disparities within MENA economies in terms of (1) their distribution of resources for health care, (2) their populations' access to and use of health services, and (3) health outcomes (see table 4.7). There is a concentration of health care personnel in urban areas, pre- dominately in capital cities. Metropolitan areas are likely to offer higher wages, better living conditions, and greater opportunities for profes- sional development than rural areas. In Libya, there are 340 physicians per 100,000 people in the capital city of Tripoli, compared to 96 per 100,000 in the rest of the country as a whole, or a mere 54 per 100,000 in the district of Joufara.5 In Lebanon, the density of physicians in Beirut is three times higher than that of the Beqaa province.6 Of Djibouti the World Health Organization reports, "[M]ost doctors and state-accredited or senior health technicians reside in the capital . . . which exacerbates the disparities in health services between the city and inland districts."7 In the Republic of Yemen, 16 percent of health personnel service Aden City, and 15 percent service the capital Sana'a, while these governorates represent only 3 percent and 8 percent of the total population, respec- tively.8 The concentration of physicians in Iraq ranges from 93 per 100,000 people in the capital of Baghdad to 67 in the large city of Basra to 31 in the smaller city of Nassiriya.9 The distribution of health care facilities, however, tends to follow the opposite pattern: there are more facilities per capita in low-density areas. This pattern may be due to policy makers' desire to make facilities geo- graphically accessible to all citizens. Data from 2001 indicate that in Jor- dan, the number of hospital beds per capita ranged from 5 per 10,000 persons in the capital of Amman to 12 in the sparsely populated rural governorate of Ma'an.10 The distribution of primary health centers in Libya follows a similar pattern, with the lowest numbers per capita in the biggest cities of Tripoli and Benghazi.11 Putting these two trends together, one would be right to suppose that rural facilities have fewer staff members and are less used than those in cities. For instance, Tunisia has achieved a fairly even geographic distri- bution of primary health centers (PHCs), but the individual facilities demonstrate striking differences in productivity. Facilities in the southeast Policy Package 1. Level the Playing Field and the Opportunity for Human Development in Lagging Areas 171 region of the country record an annual average of about two-thirds fewer visits than those in the capital of Tunis.12 In other cases, rural facilities may not be able to offer certain services at all. In Morocco, most of the primary health facilities (known there as établissements de soins de santé de base--ESSBs) are not staffed by physicians, and more than 10 percent of those that were on the books in 2004 closed because of a lack of any trained personnel at all.13 Data from 2002 in Morocco indicate that public health centers in the urban regions achieved 0.6 consultations per capita per annum, compared to just 0.3 in rural areas.14 In the Republic of Yemen, while geographic access to health facilities is reported to have improved slightly since 1990, access to a functioning facility is thought to be much lower because of chronic shortages of resources and a lack of trained personnel in rural facili- ties.15 There, the 8 percent of national health expenditure contributed by donors is used almost exclusively for capital projects, which may fur- ther aggravate the imbalance between structures and the personnel and equipment needed to make them function.16 A number of demand-side constraints reduce use of health services in poor areas--notably, limited insurance coverage. MENA's health insur- ance systems tend to focus on wealthier urban residents. In Egypt, where health insurance coverage exhibits a strong association with income (80 percent of insured women are in the two wealthiest quintiles, and the ratio of the number of insured in the wealthiest quintile to the poorest is greater than 100 to 1), the prevalence of coverage is about two times as high in urban areas as it is in rural ones.17 The more urbanized mohafazas (administrative units) in Lebanon--Beirut and Mount Lebanon--report insurance coverage of 60 percent and 54 percent, respectively, while the other mohafazas hover at around 34 percent coverage.18 Of course, indicators of utilization do not take into account possible differences in the underlying burden of disease and, therefore, cannot serve alone as a measure of satisfied need. For this, we must look at spatial disparities on services for which there is a consistent need in dif- ferent places, such as the availability of clean drinking water and effective sanitation facilities, and the proportion of births preceded by appropri- ate antenatal care and delivered by skilled attendants. These variables have been documented extensively in national surveys across the major- ity of the countries in the MENA region. · Rural-urban disparities in access to improved potable water supplies usually fall with higher levels of national income. MENA's rural- urban water supply gaps tend to be low by world standards, but Morocco and Iraq have anomalously large gaps for their level of development (figure 4.12). Yemen has a very high sanitation gap (figure 4.13). 172 Poor Places, Thriving People FIGURE 4.12 Unusually High Urban-Rural Gaps in Access to Potable Water in Morocco and Iraq 70 y = ­6.471ln(x) + 66.578 60 urban-rural difference in % of population with access to safe drinking water 50 40 Morocco 30 Iraq 20 Syrian Arab Republic Tunisia 10 Jordan West Bank and Gaza Algeria Yemen, Rep. 0 Lebanon Egypt, Arab Rep. 10 20 100 1,000 10,000 100,000 GNI per capita (US$), 2007 logarithmic scale country MENA13 economies Source: WDI; UNICEF 2009. · In terms of utilization of key services to protect child health, such as antenatal care, skilled attendance during delivery, and immunization against infectious disease, there are significant disparities between urban and rural populations across the MENA economies. There is a general pattern of diminishing inequality at higher levels of national income, although Morocco and Egypt both exhibit dispro- portionate inequalities compared with other countries of their level of GDP. The Islamic Republic of Iran is an outlier in the other direc- tion, having reduced the number of women who give birth without skilled attendance 3.5-fold between the late 1990s and the 2000s, with a particularly profound effect on the rural areas (Movahedi and others 2009). Disparities in the proportion of children who are fully vaccinated are typically smaller than those for antenatal care or skilled birth attendance, which may be a result of strong international support for immunization activities and a particular focus on rural areas. In the Republic of Yemen, where such disparities do exist, three-quarters of parents with unvaccinated children reported that distance prevented them from reaching the necessary services.19 · MENA economies' rural-urban gap in infant mortality rates (IMRs) fall in the same range as those exhibited by rural and urban populations Policy Package 1. Level the Playing Field and the Opportunity for Human Development in Lagging Areas 173 FIGURE 4.13 Rural-Urban Disparities in Sanitation Coverage 100 80 urban-rural differencial in sanitation coverage y = ­3.742ln (x) + 45.36 60 Yemen, Rep. (percentage points) 40 Egypt, Arab Rep. Tunisia Morocco 20 Jordan West Bank and Gaza Algeria Syrian Arab Republic Iraq 0 Libya ­20 ­40 100 1,000 10,000 100,000 GNI per capita ($), 2007 (log scale) Source: WDI. Note: GNI = gross national income. in other countries of similar size and economic status, such as Colom- bia, Peru, Ukraine, and the Philippines. Most striking, however, is the large absolute difference in infant and maternal mortality rates between the urban and rural populations of Morocco. Judging from the limited amount of longitudinal data available, urban-rural dispari- ties have not been reduced significantly during the past 20 years. In Morocco, although IMR has fallen in both rural and urban popula- tions, the absolute disparity between them has remained largely the same. In Jordan, the disparity widened in the early 1990s and narrowed again after 2000. In the Republic of Yemen, the disparity has grown, doubling in size between 1990 and 2006. Egypt is the exception, hav- ing narrowed the gap between IMR in urban and rural populations from more than 50 deaths per 1,000 live births to less than 15. · In some countries, pockets of bad health indicators are very localized. In Lebanon, 80 percent of infants in Beirut were immunized with the measles, mumps, and rubella (MMR) vaccine, but only 27 percent in 174 Poor Places, Thriving People the Bekaa. In Tunisia, Kasserine and Sidi Bouzid have not shared in the rest of the country's increased rate of skilled deliveries. In Jordan, under-five mortality rates are 10 percent in Zarqa and 39 percent in Karak. In Morocco, the reduction of the rural-urban gap in maternal mortality between 2000 and 2006 was mostly due to gains in rural areas bordering the Atlantic, while many rural and mountainous regions in west-central Morocco continue to lag behind (Ministčre de la Santé 2003, 2007). Bringing Better Health to Lagging Areas: Not More of the Same What can MENA governments do to reduce the health status divide between leading and lagging areas? It is important to recognize the limits of a "copy and paste" strategy based upon the replication of health facility infrastructure. Health facilities in low-density areas often already have low utilization rates and suffer from a lack of qualified personnel and nonstaff recurrent funding. Therefore, the key is to improve the quality of health service delivery. To achieve this means finding service-delivery mechanisms that fit the specific needs of poor or low-density areas (see box 4.6). When it comes to strengthening public health delivery facilities in lagging areas, successful interventions have focused on staff incentives and nonstaff recurrent funding. It is critical to ensure that health personnel are motivated to provide quality services in lagging areas despite the better professional opportunities and quality of life in richer places. Morocco's Integrated Management of Childhood Illnesses (IMCI) program, for example, explicitly focused on the question of staff motivation, with training and free housing for personnel in lagging areas. An evaluation of IMCI found that it had doubled the number of correct antibiotic pre- scriptions, with similar results found in Egypt, Tanzania, Uganda, and Brazil (WHO and DFID 2003). The Social Priorities Program Basic Health Project in Morocco, which targeted facility, equipment, and medication supply rehabilitation and the establishment of lodging facilities for health personnel in the 14 most disadvantaged and least urbanized provinces, was associated with a 150 percent higher rate of medically assisted deliveries over approximately five years (18.5 to 47 percent) (World Bank 2004b). Spatially adapted delivery systems have proved their cost-effectiveness for low-density areas. The Islamic Republic of Iran has trained local behvarz (health workers) to provide family planning and reproductive health care (Sadrizadeh 2004). As a result, service coverage in rural areas has increased from nearly 40 to 90 percent in 15 years (1985­2001). In Policy Package 1. Level the Playing Field and the Opportunity for Human Development in Lagging Areas 175 the Republic of Yemen, community empowerment strategies in early childhood development have proved particularly cost-effective in tack- ling nutrition in children under five. The Child Development Project (2000­05) involved the organization and mobilization of community- level committees to oversee health training, water distribution, and edu- cational access, and was associated with an improved enrollment rate for primary school girls by 20 percentage points and an increase in access to safe drinking water by 17 percent over five years (World Bank 2006c). With spatially adapted delivery systems for potable water and sanitation, serving low-density areas does not have to be expensive and institu- tionally complex. The Egyptian government's Integrated Sanitation and Sewerage and Infrastructure Project (ISSIP) recognized the limita- tions of extending the urban network architecture into rural areas. Client villages were, therefore, clustered around hydraulic subbasins instead of administrative boundaries, natural dilution was substituted for secondary treatment and decentralized or on-site systems replaced centralized systems, allowing the project to provide connections at a low average capital cost of $140 per inhabitant (Soliman and others 2009). In rural areas of the Republic of Yemen, community ownership and management of piped water services has led to an increase in con- nection rates and increased consumer satisfaction (Dasgupta, Meisner, Makokha, and Pollard 2009). Understanding and addressing the demand-side constraints on health serv- ice use in poor areas can assist the design of interventions. As for educa- tion, a multitude of nonhealth demand-side factors may be limiting the expansion of health services in lagging areas. In the Islamic Republic of Iran, household economic status and mothers' education together explain more than half the variation in the infant mortality rate. Strate- gies that aim to expand education, transportation, nutrition, water, sani- tation, and incomes in lagging areas may have a bigger payoff in terms of spatial disparities in health outcomes than a simple emphasis on the stock of health facilities (CSDH 2008). International evidence sheds light on the important role that improved educational level, social status, and women's welfare have played in countries as disparate as Singapore (World Bank 2009a), Sri Lanka, Costa Rica, and the United Kingdom.20 Demand-side interventions, therefore, have helped improve the health status of lagging areas' populations. Successful demand-side interventions have focused on the two pillars of demand: awareness and affordability. On the awareness side, Egypt's World Bank­funded Population Proj- ect managed to expand the provision of family planning services through demand-creation; Social Change Agents raised women's confidence and awareness; and income-generation activities empowered women with 176 Poor Places, Thriving People their own resources to attend family planning services. The project tar- geted the lagging area of Upper Egypt, and its demand-side approach succeeded in raising the contraceptive prevalence rate from 18 percent to 37 percent in five years in the target zone, quadrupling attendance at family planning clinics (World Bank 2005h). This was no small-scale pilot project: the Social Change Agents made 3.5 million visits to 230,000 women during the life of the project (Georgetown University 2007). The women saw the Social Change Agents as a "picture of success and respect," which encouraged them to attend health education, public meetings, and adult education. Similarly, the Islamic Republic of Iran's behvarz community midwives have helped close rural-urban health dis- parities by encouraging women to seek formal health services (Aghajan- ian, Mehryar, Ahmadnia, and Kazemipour 2008). On the affordability side, expanding insurance schemes into poor areas encourages public and private providers to deliver appropriate forms of care by stimulating demand. In the Islamic Republic of Iran during the past decade, the government has extended insurance to both the urban and rural uninsured. The result of this policy has been to raise rural insurance coverage from 33 percent to more than 76 percent between 2001 and 2006, eliminating what had been a large disparity between the prevalence of coverage in urban and rural areas.21 Improved management information systems help health departments to ensure an efficient spatial targeting of resources. The Islamic Republic of Iran's "master plan" approach to health care planning, involving infrastructure development by assessing the number and locations of all levels of health facilities at the district level, has improved the distribu- tion of resources as well as the monitoring of health status progress geo- graphically. In the West Bank and Gaza, the unification of information systems from the Governmental Health Insurance bureaus resulted in a more accessible and transparent beneficiary registry. Together with quality improvement standards and guideline development, these improved information technology systems were associated with an increased procurement, delivery, and availability of pharmaceuticals in lagging areas; an increase of 30 percent in the utilization rates of reha- bilitated facilities; and a nearly sevenfold improvement in diabetes man- agement indicators (World Bank 2005g). The Better Social Safety Nets Are Targeted, the Better Spatial Disparities Are Mitigated Social Safety Net programs can have a spatial significance. Where spatial pockets of poverty exist, these programs, if well targeted, can be a weapon Policy Package 1. Level the Playing Field and the Opportunity for Human Development in Lagging Areas 177 BOX 4.6 Leveling Spatial Disparities in Health: Four Ideas for Policy Makers · Analyzing the causes of spatial disparities in health out- comes, to find out whether the root cause lies in socioeco- nomic inequalities, the quality of service, or physical access to health facilities. · Implementing information systems for the monitoring of health care activities and the efficient spatial allocation of resources. · Designing packages of staff incentives and nonstaff recur- rent funding for health facilities in lagging areas. · Investigating nonconventional models for the delivery of health care, potable water, and sanitation services in low- density lagging areas. Source: Study Team. in government armories to reduce spatial disparities (see box 4.7). If they are poorly targeted, however, they can reinforce the rift between rich and poor places. Social Safety Nets Are an Instrument of Spatial Policy-- For Better or for Worse Many MENA economies have some kind of Social Safety Net (SSN) program: an arrangement for resource transfers intended to improve social welfare (table 4.8). SSNs can play an important role in reducing poverty and vulnerability to shocks. SSN approaches range from direct cash and in-kind transfers, food distribution programs, public works, and conditional cash transfers (CCTs) to commodity subsidies, which are usually applied to energy and food. SSNs can be categorized into direct and indirect transfers. Direct transfers include cash transfers (conditional and nonconditional), in-kind transfers, as well as social (noncontributory) pensions, and public works programs. Indirect transfers include commodity subsidies as well as assis- tance and waiver programs. Social safety nets have multiple goals that differ by program and political motivation. In times of crisis, they aim 178 TABLE 4.7 Indicators of Rural-Urban Disparities in Health Iran, West Syrian Egypt, Islamic Bank and Arab Yemen, Algeria Djibouti Arab Rep. Rep. Iraq Jordan Lebanon Libya Morocco Gaza Republic Tunisia Rep. Total health expenditure (% of GDP) 3.60% 6.70% 6.30% 7.80% 3.80% 9.90% 8.90% 2.90% 5.10% -- 3.90% 5.30% 4.60% Government expenditure on health (% of GDP 2004) 2.60% 4.40% 2.20% 3.20% -- 4.70% 3.20% 2.80% 1.70% 7.80% 2.20% 2.80% 2.20% Private expenditure on health (% of total) 23% 25% 59% 44% 28% 58% 53% 30% 64% -- 52% 56% 54% Contraceptive prevalence modern methods (% urban) 52% 18% 60% 55% 36% 43% -- 54% 56% -- 48% 53% 34% Contraceptive prevalence modern methods (% rural) 52% 4% 54% 57% 27% 36% -- 46% 53% -- 35% 54% 13% Total fertility rate (urban) 2.19 4.1 2.7 1.8 4 3.6 -- 3.8 2.1 4.4 3.4 2.46 4.5 Total fertility rate (rural) 2.38 4.9 3.4 2.4 5.1 3.7 -- 4.9 3 4.4 4.4 3.29 6.7 Access to improved drinking water (% urban) 88% 98% 99% 97% 92% 98% 100% 98% 98% 91% 94% 99% 74% Access to improved drinking water (% rural) 82% 54% 97% 86% 57% 94% 100% 99% 57% 90% 80% 84% 52% Access to improved sanitation (% urban) 98% 70% 85% 86% 98% 92% 100% 99% 98% 100% 100% 96% 92% Access to improved sanitation (% rural) 87% 19% 52% 78% 82% 76% 87% 98% 58% 93% 94% 64% 34% Utilization of antenatal care (% urban) 94% 82% 82% 96% 90% 99% -- -- 68% 99% 90% 75% 69% Utilization of antenatal care (% rural) 85% 41% 62% 88% 75% 98% -- -- 48% 99% 78% 55% 38% Utilization of skilled birth attendant (% urban) 98% 95% 89% 96% 86% 99% -- 96% 85% 99% 98% 98% 62% Utilization of skilled birth attendant (% rural) 92% 40% 66% 90% 71% 97% -- 99% 40% 98% 88% 89% 26% Children fully vaccinated by two years (% urban) 85% -- 89% -- 62% 94% -- -- 94% 95% 89% 96% 58% Children fully vaccinated by two years (% rural) 89% -- 89% -- 41% 91% -- -- 84% 97% 87% 94% 29% Infant mortality rate (urban) 33 68 31.7 27.7 35 20 -- 21.4 33 26.8 16 16 55.3 Infant mortality rate (rural) 44 54 45.2 30.2 35 23 -- 31.4 55 26.6 20 30 73.3 Stunting (% urban) 10% 32% 16% 11% 19% 12% -- 4% 13% 11% 22% 4% 44% Stunting (% rural) 12% 44% 18% 22% 24% 13% -- 6% 24% 9% 23% 10% 56% Under-five mortality rate (urban) 40 95 39.1 36.8 41 22 -- 27.5 38 29.7 19 -- 56.7 Under-five mortality rate (rural) 56 73 56.1 34.6 41 27 -- 36.3 69 29.9 23 -- 86.1 Sources: See Appendix. Note: -- = not available. 179 180 Poor Places, Thriving People TABLE 4.8 Social Resource Transfers in MENA Direct transfers Indirect transfers Cash and in- Public works Food Conditional Food Energy kind transfers programs distribution cash transfers subsidies subsidies Djibouti x Egypt, Arab Rep. x x x x x Iran, Islamic Rep. x x x Iraq x x x Jordan x x x Lebanon x x x Morocco x x x x Syrian Arab Republic x x x Tunisia x x x x x West Bank and Gaza x x x x Yemen, Rep. x x x Source: Study Team, based on World Bank 2005a, 2005b, 2007c, 2007d. Countries included the Arab Republic of Egypt, Morocco, the Republic of Yemen, Jordan, the Syrian Arab Republic, the West Bank and Gaza, Iraq, the Islamic Republic of Iran, Djibouti, Tunisia, and Lebanon. mainly at protecting poor households from economic shocks and, in par- ticular, from decisions that may cause irreversible loss of human capital, such as malnutrition and dropping out of school. SSNs help reduce inequality and extreme poverty, enable households to make better deci- sions for the future, and help governments manage the social impacts of beneficial reforms (World Bank 2008d). Commodity subsidies are by far the major SSN in MENA (figure 4.14). In the Republic of Yemen, energy and food subsidies account for almost 8 percent of GDP and are as high as the country's spending on health and education. Energy subsidies are the major subsidy expenditure item. Across the MENA region overall (including GCC economies) energy subsidies were averaging 7.1 percent of GDP (World Bank 2009b), even before the fuel price increases of 2008. For the MENA13, they ranged from Tunisia's 1.5 percent of GDP or 6 percent of government spending, to 12­18 percent of GDP and 30­50 percent of government spending in Algeria, Egypt, the Islamic Republic of Iran, Iraq, and the Syrian Arab Republic (World Bank 2009b). There are, of course, several problems with SSNs that consist mainly of energy subsidies. They encourage the overuse and ineffi- cient use of energy. They also amplify the impact of worldmarket price fluctuations on the balance of payments and the budget. For policy makers concerned with spatial disparities, however, a major drawback of energy subsidies is that they are "regressive." This finding Policy Package 1. Level the Playing Field and the Opportunity for Human Development in Lagging Areas 181 FIGURE 4.14 Commodity Subsidies: A Big Slice of MENA's GDP 16 14 12 10 % of GDP 8 6 4 2 0 Syrian Egypt, Yemen, Morocco Arab Republic Arab Rep. Rep. country direct transfers commodity subsidies Source: World Bank 2009b. means that the people who consume the most, the nonpoor, receive the lion's share of the subsidy. · In Jordan, for example, only 7 percent of fuel subsidies reached the poorest quintile, while 44 percent went to the richest quintile (World Bank 2007d). · Map 4.1 shows how the regressivity of fuel subsidies helps to widen spatial disparities, taking the example of Iraq. The residents of the low-poverty governorates of Baghdad and the north receive high per capita transfers through the subsidy mechanism, while the worse-off citizens of Kirkuk, Diala, Wasit, Missan, Thi-Qar, Basra, Karbala, An-Najaf, Al-Qadisiya, Babil, and Salahuddin receive a fraction of these amounts. · Similarly in Egypt, where more than 6 percent of GDP is spent on fuel subsidies, the leading area city of Cairo alone has 58 percent of the country's private cars and, therefore, captures a disproportionate share of the benefits. The lagging governorates of Siwa, Assuit, and Menya, in contrast, have 16 percent of the population but only 3 percent of Egypt's cars between them (IDSC 2007). Social security systems mostly consisting of mandatory pension and health insurance systems have a low coverage and tend to be focused on 182 Poor Places, Thriving People MAP 4.1 Petroleum Product Subsidies Exacerbating Spatial Disparities in Iraq (thousands of Iraqi dinars per person per month) POVERTY RATES (%) IBRD 38173 NOVEMBER 2010 < 10 10 ­ 20 20 ­ 30 30 ­ 40 > 40 Duhouk PC FUEL SUBSIDY Erbil 20 (ID '000/month) Ninevah Sulaimaniya Kirkuk Salahuddin Diala Baghdad Al-Anbar Babil Wasit Kerbela Al-Qadisiya Missan Thi Qar An-Najaf Basrah 0 50 100 150 200 250 KILOMETERS Al-Muthanna This map was produced by the Map Design Unit of The World Bank. The boundaries, colors, denominations and any other information shown on this map do not imply, on the part of The World Bank Group, any judgment on the legal status of any territory, or any endorsement or acceptance of such boundaries. Source: Authors' calculations based on Iraq Household Socioeconomic Survey 2006­07 and World Bank 2009b. Note: Calculations assume that the percentage of per capita fuel expenditure subsidized is constant across governorates. high-income and public sector workers in metropolitan and urban areas. For MENA as a whole, they cover a modest 33 percent of the workforce, including a large share of public sector workers. They have very little redistributive effect and leave out vulnerable groups, in particular big shares of the rural population, the informal economy (which predomi- nates in rural and lagging areas), and the elderly (who represent a higher share of the population in rural and lagging areas). In addition, no econ- omy in the region offers genuine unemployment insurance. Some transfer programs have, on the contrary, a proven positive effect on spatial disparities. Evidence indicates that social funds have the ability to target lagging areas effectively (table 4.9). These funds enable com- munities to identify and implement local investments, which could be in Policy Package 1. Level the Playing Field and the Opportunity for Human Development in Lagging Areas 183 TABLE 4.9 Most Social Funds Target Lagging Areas Well Geographic Distribution of Social Fund Resources (%), 1991­99 Decile Armenia Bolivia Honduras Nicaragua Peru Zambia 1 (poorest) 10.7 10.8 14.5 14.7 25.3 9.8 2 13.2 13.6 10.7 14.4 21.2 10.6 3 12.4 15.2 11.4 12.8 19.2 14.9 4 12.0 13.0 9.0 11.4 12.3 8.6 5 9.4 11.0 10.0 10.6 8.3 8.6 6 5.7 11.9 10.4 10.7 4.9 13.5 7 5.7 8.0 8.0 8.3 3.2 8.4 8 6.9 7.6 5.0 6.9 3.0 10.6 9 10.1 6.0 15.1 5.5 1.9 6.4 10 (richest) 13.9 2.8 6.0 4.7 0.7 8.6 Source: Rawlings, Sherburne-Benz, and van Domelen 2004. TABLE 4.10 Conditional Cash Transfers that Are Proven to Reduce Poverty Colombia Honduras Mexico Nicaragua Poverty measure 2002 2006 2000 2002 1998 Jun-99 Oct-99 2000 2001 2002 Headcount Control 0.95 0.9 0.88 0.91 0.89 0.93 0.94 0.84 0.91 0.9 Index Impact A ­0.03* A B 0.02** ­0.01** 0.00 A ­0.07** ­0.05** Poverty Gap Control 0.58 0.54 0.49 0.54 0.47 0.55 0.56 0.43 0.50 0.50 Impact A ­0.07** A ­0.02* 0.01* ­0.03** ­0.02* A ­0.03** ­0.09** Squared Poverty Control 0.53 0.43 0.30 0.36 0.28 0.35 0.36 0.26 0.32 0.32 Gap Impact A ­0.02** A ­0.02* B ­0.09** ­0.03* A ­0.12 ­0.09** Source: Fiszbein and Schady 2009. Note: A = Baseline, before households in CCT treatment groups received transfers; B = No significant impact on poverty measure; * = Significant at the 10 percent level; ** = Significant at the 5 percent level. the field of health, education, infrastructure, or income-generating activ- ities. Spatially targeted social funds in MENA include Morocco's Initia- tive Nationale de Développement Humain (INDH) and Egypt's 1,000 Villages program. Conditional cash transfers (CCTs) are acquiring a successful track record (see table 4.10), confirming their potential as an instrument of spatial policy. They provide cash assistance to poor families, dependent on their behavior, such as keeping children in school or having them vaccinated. The cash reduces poverty directly, and indirectly improves human devel- opment by compensating beneficiaries for investing in their education and health. In several countries, CCTs were accompanied by detailed 184 Poor Places, Thriving People poverty maps, which facilitated high-resolution geographic targeting (Fiszbein and Schady 2009). But in MENA, these positive examples of transfer programs are by no means prevalent. Effective SSNs fall short of the needs. CCT pilots have been put in place only in Morocco and the West Bank and Gaza (plans exist in the Syrian Arab Republic). Today, most transfer programs in MENA use categorical targeting approaches. This finding means that a household or individual is entitled to benefits if it falls into an eligible cat- egory, such as single mothers, widows, the unemployed, the elderly, the disabled, and so forth. Often, these categories are not synonymous with being poor and vulnerable. So many existing programs could contribute more to reducing spatial disparities if better targeted. In the Republic of Yemen in 2005, for example, the Social Welfare Fund cash transfer program was reaching only 13 percent of the poor population. Of those who did receive cash transfers, 70 percent were not in the target group (Blomquist 2008). Figure 4.15 illustrates the spatial incidence of the recently introduced cash transfer program in Iraq. It is categorically targeted at certain groups of society, including the unemployed, the disabled, orphaned children, divorced or widowed women, and married male university students. Under this plan, house- hold heads must initiate their own registration and provide their own documentation proving that they are poor enough to qualify. When one observes the spatial impact of the program, the correlation between governorates' poverty rates and SSN assistance rates is rather weak. For example, in Najaf, where the 23 percent poverty rate is slightly below the national average, 3.6 percent of the population is benefiting from the subsidy, while in Muthanna, where nearly half the inhabitants FIGURE 4.15 Spatial Incidence of Cash Transfer Program, Iraq 4.0 Najaf Wasit Karbala population receiving safety net (%) 3.5 Qadisea Babil Ninewa 3.0 Maysan Thi-Qar Dyalaa Muthanna 2.5 Anbar Basra Salah Aldeen 2.0 Kirkuk 1.5 Baghdad 1.0 0.5 0.0 0 10 20 30 40 50 60 poverty rate (%) Source: Shlash 2009. Policy Package 1. Level the Playing Field and the Opportunity for Human Development in Lagging Areas 185 are below the poverty line, just 2.7 percent receive the cash transfer. The lower enrollment rates in Kirkuk and Baghdad, however, do reflect those governorates' lower poverty incidence. Jordan's National Aid Fund (NAF) enforces its categorical targeting through home visits by NAF officers to verify eligibility, as well as reviewing the statements and documents provided (World Bank 2007d). Under this design, the geographic distribution of social transfers maps more closely onto the spatial distribution of poverty. Per capita income under the category "other government transfers" (i.e., not social security benefits or pensions) was JD 402 in 2006 in rural areas of Jordan com- pared with JD 248 in urban areas. Residents of the Amman governorate received JD 194, significantly less than the urban average.22 Tunisia's Fund of National Solidarity (FSN23) shows how needs-based targeting serves as spatial targeting. The FSN was created in 1992 to finance projects too small to fit into sector ministries' normal service packages. A national committee used objective criteria to select 1,811 project zones, and local administrations identified projects. The result is that most FSN expenditure takes place in the country's lagging western and southern areas. Someone's chance of living in an FSN zone is 13 to 17 times higher than the national average in the lagging regions of Le Kef and Gabčs. Targeting Is Key Better targeting and coverage will therefore make MENA's SSN pro- grams more powerful instruments for the reduction of spatial disparities. A number of approaches could improve coverage and targeting and enhance programs' contribution to spatial convergence in living stan- dards. If measuring people's incomes directly ("means testing") is diffi- cult and costly, there are proven alternative targeting approaches. · Geographical targeting distributes benefits to people who are identi- fied as likely to be poor or vulnerable on the basis of where they live. This targeting mechanism is cheap and administratively simple, and especially efficient when poverty is highly concentrated and corre- lated spatially with observable social indicators. The smaller the geo- graphical unit of measurement, the lower the risk of including the nonpoor or excluding the poor. In the Syrian Arab Republic, for example, targeting a transfer by governorate might reduce the national poverty rate by 1.8 percentage points, whereas targeting the same budget by region would only achieve a reduction of 0.6 percentage points. Poverty mapping is, therefore, frequently used to identify tar- get areas. A simulation of geographical targeting using Morocco's 186 Poor Places, Thriving People poverty map illustrated the higher efficiency of fine geographic tar- geting. Targeting at district level could achieve the same poverty- reduction impact as targeting at regional level for 36 percent of the cost (Douidich, Ezzrari, and Lanjouw 2008). Refining spatially disaggre- gated indicators therefore goes hand in hand with improving the exact- ness of spatial targeting. · Proxy means testing (PMT) is a targeting approach that uses easily observed household indicators, such as ownership of assets and household size, to predict household income with a fair degree of accuracy. It is gaining popularity in the developing world. Chile was the first country to introduce PMT in 1980, followed by Colombia and Mexico, Armenia since 1994, and Turkey since 2002. By 2004, Argentina, Costa Rica, Ecuador, Jamaica, Honduras, and Nicaragua were putting PMT arrangements in place, and the Russian Federa- tion, Egypt, Zimbabwe, and Sri Lanka were running pilot programs (Coady, Grosh, and Hoddinott 2004). Simulations of proxy means testing in the Syrian Arab Republic suggest that it could reduce the poverty rate by close to 2 percentage points more than a universal lump sum transfer of equal budget size. In Egypt, a shift from the current SSN program to proxy means testing plus geographic target- ing could produce a 3 percentage point drop in the poverty rate with an unchanged budget (World Bank 2005a). Similar results apply to low-income countries such as the Republic of Yemen, with the differ- ence that the improvements are seen as people becoming less poor (i.e., less depth and severity of poverty) rather than rising above the poverty line altogether. Below are two simulations of the spatial impact of SSNs. The first experiment simulates a program in the Syrian Arab Repub- lic, with a budget of US$260 million, or about 1.2 percent of GDP. Map 4.2a shows the spatial distribution of poverty in the Syrian Arab Repub- lic before the program begins, with the highest poverty rates in the far north and the far south. Under the first variant of a SSN program, the budget available is sim- ply allocated to each of the Syrian Arab Republic's 14 governorates in pro- portion to the number of poor. Aleppo, therefore, receives more than a third of the program budget (US$96 million), while Damascus capital governorate receives only 5 percent (US$13 million). The governorate allocations are then shared out equally within each governorate. This is a very crude, low-resolution, form of geographical targeting. Map 4.2b shows the impact: a drop in the poverty rate of up to 4.5 percent across the country, with the poorest areas tending to have the greatest decrease in the poverty rate. For example, in the poor governorate of Aleppo, the Policy Package 1. Level the Playing Field and the Opportunity for Human Development in Lagging Areas 187 MAP 4.2 Poverty Incidence and Poverty Gaps in the Syrian Arab Republic (a) Poverty Incidence by Governorate, Syrian Arab Republic; (b) Change in Poverty Incidence, with Geographic Targeting; (c) Change in Poverty Incidence, with Geotargeting and PMT; (d) Change in Poverty Gaps, with Geographic Targeting; (e) Change in Poverty Gaps, with Geotargeting and PMT (a) This map was produced by the Map Design Unit of The World Bank. The boundaries, colors, denominations and any other information IBRD 38174 NOVEMBER 2010 (b) This map was produced by the Map Design Unit of The World Bank. The boundaries, colors, denominations and any other information IBRD 38175 NOVEMBER 2010 shown on this map do not imply, on the part of The World Bank shown on this map do not imply, on the part of The World Bank Group, any judgment on the legal status of any territory, or any Group, any judgment on the legal status of any territory, or any endorsement or acceptance of such boundaries. endorsement or acceptance of such boundaries. El Haska El Haska Halab Halab El Requa El Requa Adelb Adelb El Lazeqia El Lazeqia Hamaa Deir El Zour Hamaa Deir El Zour Tartous Tartous Hemes Hemes CHANGE IN POVERTY POVERTY INCIDENCE % INCIDENCE (PPT) Rural Damascus <8 Rural Damascus > -7.5 8 ­ 10 -7.5, -6 El Quanitera El Quanitera 10 ­ 12 -6, -4.5 Deraa 12 ­ 14 Deraa -4.5, -3 El Swayedaa El Swayedaa 0 25 50 75 100 14 ­ 16 0 25 50 75 100 -3, -1.5 > 16 < -1.5 KILOMETERS KILOMETERS (c) This map was produced by the Map Design Unit of The World Bank. The boundaries, colors, denominations and any other information IBRD 38176 NOVEMBER 2010 (d) This map was produced by the Map Design Unit of The World Bank. The boundaries, colors, denominations and any other information IBRD 38177 NOVEMBER 2010 shown on this map do not imply, on the part of The World Bank shown on this map do not imply, on the part of The World Bank Group, any judgment on the legal status of any territory, or any Group, any judgment on the legal status of any territory, or any endorsement or acceptance of such boundaries. endorsement or acceptance of such boundaries. El Haska El Haska Halab Halab El Requa El Requa El Lazeqia Adelb Adelb El Lazeqia Hamaa Deir El Zour Hamaa Deir El Zour Tartous Tartous Hemes Hemes CHANGE IN POVERTY CHANGE IN POVERTY Rural Damascus INCIDENCE (PPT) GAP (PPT) > -7.5 Rural Damascus > -2 -7.5, -6 -2, -1.6 El Quanitera El Quanitera -1.6, -1.2 -6, -4.5 Deraa -4.5, -3 Deraa El Swayedaa -1.2, -0.8 El Swayedaa -0.8, -0.4 0 25 50 75 100 -3, -1.5 0 25 50 75 100 < -1.5 < -0.4 KILOMETERS KILOMETERS (e) This map was produced by the Map Design Unit of The World Bank. The boundaries, colors, denominations and any other information IBRD 38178 NOVEMBER 2010 shown on this map do not imply, on the part of The World Bank Group, any judgment on the legal status of any territory, or any endorsement or acceptance of such boundaries. El Haska Halab El Requa Adelb El Lazeqia Hamaa Deir El Zour Tartous Hemes CHANGE IN POVERTY GAP (PPT) Rural Damascus > -2 -2, -1.6 El Quanitera -1.6, -1.2 Deraa -1.2, -0.8 El Swayedaa 0 25 50 75 100 -0.8, -0.4 < -0.4 KILOMETERS Sources: Authors' simulations using 2003­04 Household Income and Expenditure Database; work by John Blomquist and Nader Kabbani (2007). Note: Geographic targeting allocates budget to governorates on the basis of the proportion of poor individuals residing in that governorate and then gives each person within that governorate an equal share of the governorate's resources. Administrative cost for targeting assumed to be SL 0.95 million. Geographical + proxy means first allocates budget to governorates proportionally to poverty and then provides the governorate per capita transfer level to each eligible individual under the proxy-means test. Proxy-means targeting provides the national per capita transfer level to each eligible individual under the proxy means test. Administrative cost for targeting is assumed to be SL 1.24 million. 188 Poor Places, Thriving People poverty rate drops from close to 20 percent down to 15.5 percent. Map 4.2c shows what happens if the geographical targeting is complemented by proxy means testing: the leveling of spatial disparities is even more pro- nounced. Poverty rates in Aleppo and El Swayedaa fall by nearly 8 and 4 percentage points, respectively. The enhanced poverty impact of the com- bination of PMT and spatial targeting is reflected at the national level: from a starting point of a 11.5 percent poverty rate, a universal lump-sum transfer would reduce the rate to 9.9 percent, the low-resolution geo- graphic targeting to 9.6 percent, and the geotargeting with proxy means testing to 8.2 percent--assuming equal budgets for each option. A second simulation shows that the spatial impacts of targeted social safety net plans can be even clearer when looking at the depth of poverty. The depth of poverty ("poverty gap") measures how far below the poverty line the poor lie, on average. A transfer might not actually bring a poor household above the poverty line, and thus will not necessarily reduce the poverty rate, but it will always bring it closer to the poverty line and thus will reduce the poverty gap. As the targeting of the social safety net becomes progressively more needs-based (map 4.2d and e), the spatial dis- tribution of the impact on the poverty gap maps more closely onto the spatial distribution of poverty, thus narrowing spatial disparities more effectively. A simulation for Egypt (World Bank 2005a) compared the spatial impact of the country's existing social safety net plan with that of low- resolution (governorate-level) geographical targeting. The gap between the poverty rates of Upper Rural and Lower Urban Egypt was 31 per- centage points in 2005. Distributing LE 3 billion per year (0.6 percent of 2005 GDP) under geographical targeting would reduce the gap to 28 per- centage points, and geographical targeting plus a proxy means test would further reduce the gap to 22 percentage points (World Bank 2005a). In the Republic of Yemen, poverty rates are high, averaging 35 per- cent nationally and exceeding 50 percent in the poorest governorates. A simulated social safety net with a total budget equivalent to that in the Syrian example (US$194 million, or 1.24 percent of GDP) did not have a significant impact on the poverty rate, because many of the poor fell too far below the poverty line to be lifted above it. However, as the targeting of the simulated social safety net becomes more precise with the use of geographic and PMT methodologies, the spa- tial incidence of its impact on the poverty gap (map 4.3b and c) maps more and more closely onto the poverty map (map 4.3a). The governorates with the greatest poverty incidence--Amran, Shabwah, Mareb, Hajja, and Abyan--are those with the greatest decrease in the depth of poverty. These simulations confirm the potential role of targeted SSN pro- grams as components of strategies to reduce spatial disparities. It is part Policy Package 1. Level the Playing Field and the Opportunity for Human Development in Lagging Areas 189 MAP 4.3 Poverty Gaps in the Republic of Yemen (a) Poverty Gaps by Governorate, Republic of Yemen; (b) Change in Poverty Gap, with Geographic Targeting; (c) Change in Poverty Gap, with Geotargeting and PMT (a) POVERTY GAP IBRD 38179 NOVEMBER 2010 (b) CHANGE IN POVERTY GAP (PPT) IBRD 38180 NOVEMBER 2010 < 4% > -0.30 4% ­ 8% -0.30, -0.24 8% ­ 12% -0.24, -0.18 12% ­ 16% -0.18, -0.12 16% ­ 20% -0.12, -0.06 > 20% < -0.06 Sa'adah Al-Maharh Sa'adah Hadramout Hadramout Al-Maharh Al-Jawf Al-Jawf Amran Hajja Amran Hajja Al-Mahweet Al-Mahweet Mareb Mareb Sana'a Region Sana'a Region Al-Hodeida Shabwah Al-Hodeida Dhamar Shabwah Remah Dhamar Remah Al-Baida Al-Baida Ibb Ibb Al-Dhale Abyan Al-DhaleAbyan Taiz Taiz Laheg Laheg Aden Aden This map was produced by the Map Design Unit of The World Bank. This map was produced by the Map Design Unit of The World Bank. The boundaries, colors, denominations and any other information 0 50 100 150 200 250 The boundaries, colors, denominations and any other information 0 50 100 150 200 250 shown on this map do not imply, on the part of The World Bank shown on this map do not imply, on the part of The World Bank Group, any judgment on the legal status of any territory, or any Group, any judgment on the legal status of any territory, or any endorsement or acceptance of such boundaries. KILOMETERS endorsement or acceptance of such boundaries. KILOMETERS (c) CHANGE IN POVERTY GAP (PPT) IBRD 38181 NOVEMBER 2010 > -0.30 -0.30, -0.24 -0.24, -0.18 -0.18, -0.12 -0.12, -0.06 < -0.06 Sa'adah Hadramout Al-Maharh Al-Jawf Amran Hajja Al-Mahweet Mareb Sana'a Region Al-Hodeida Shabwah Remah Dhamar Al-Baida Ibb Al-Dhale Abyan Taiz Laheg Aden This map was produced by the Map Design Unit of The World Bank. The boundaries, colors, denominations and any other information 0 50 100 150 200 250 shown on this map do not imply, on the part of The World Bank Group, any judgment on the legal status of any territory, or any endorsement or acceptance of such boundaries. KILOMETERS Sources: Authors' simulations using 2005­06 Household Budget Survey; work by John Blomquist and Nader Kabbani (2007). Note: Geographic targeting allocates budget to governorates on the basis of proportion of poor individuals residing in that governorate and then gives each person within the governorate an equal share of resources. Administrative cost for targeting is assumed to be YR 1.65 million. Geographical with proxy means first allocates budget to governorates proportionally to poverty and then provides the governorate per capita transfer level to each eligible individual under the proxy means test. Proxy-means targeting provides the national per capita transfer level to each eligible individual under the proxy means test. Administrative cost is assumed to be YR 2.36 million. of the broader theme that the best way to support disadvantaged places is often to support the disadvantaged people who live in them (see box 4.7). Smoothing the Path for City Growth Urbanization helps to reduce spatial disparities. Smoothing the path for urbanization is a recipe for spatial convergence in living standards. First, as we have seen in our discussion of lagging-area diagnostics, poor 190 Poor Places, Thriving People BOX 4.7 Harnessing Social Safety Nets to Lagging-Areas Policy: Three Ideas for Policy Makers · Working out whether the spatial incidence of existing resource transfers and subsidies is coherent with national spatial devel- opment policies · Involvement of ministries and institutions responsible for social programs in the formulation and implementation of lagging-areas policies · Piloting/rolling out needs-based targeting of Social Safety Net programs Source: Study Team. places are typically held back by large households and low land-labor ratios. Allowing leading urban areas to grow and absorb migrants from lagging areas is, therefore, critical to improving lagging areas' living standards. City growth and rural-urban migration mean more compe- tition for jobs in cities and less in lagging areas, which will help close the wage gap between rich and poor places. Second, artificial limits on city expansion reveal themselves as high rents for urban land and buildings. If people need factories, offices, or accommodation in the city to be close to their work or customers, and the supply of land is limited, rents and real estate prices will rise. In this situation, access to urban land is like a monopoly. The urban population captures an income premium thanks to the choking-off of land supply, thus entrenching the urban-rural incomes gap. Third, the redistribution of population out of lagging areas reduces the spatial component of total inequality. But fear of urbanization and of spatial disparities has led governments around the world to adopt anti-urbanization policies. These policies come in two varieties: · Regulatory blocks on urbanization and city growth. Administrations have set limits on land utilization and conversion, rents, migration, private titling, the provision of services to in-migrants, and density (World Bank 2009a, 204). · Rural economic development programs, designed to (among other things) reduce the economic incentive for people to migrate to the cities. Urbanization is proceeding apace in MENA13; the average rate of growth of the urban population is 3.3 percent per year, compared with Policy Package 1. Level the Playing Field and the Opportunity for Human Development in Lagging Areas 191 2.6 percent for national populations (table 4.11). Both urban growth and slum growth in MENA are essentially driven by national population growth and the economics of agglomeration, as figure 4.16 illustrates. (A reasonably precise rule of thumb is that a MENA economies' annual urban population growth rate is the national population growth rate plus 0.7.) The question, therefore, is not how to slow down city growth, but rather how policy makers can ensure that MENA's city growth is as smooth as possible, involving the minimum of sociopolitical friction and economic distortion. The choice is between the urbanization of planned construction, provision of services to new arrivals, and social integration on the one hand, and the urbanization characterized by slums, haphazard zoning, poor human development, and social dislocation on the other. A discussion of "smooth city growth" must address three topics: land markets, urban public services, and institutional frameworks. Whether city growth is smooth or painful will depend on how these three issues are resolved. We will look briefly at each in turn, fully recognizing that a few examples cannot do justice to the urban development agenda. Land Markets that Support Smooth Urban Growth Well-functioning land markets are critical for smooth urban develop- ment. They need to provide urban housing and commercial and industrial TABLE 4.11 Urban and Slum Growth in MENA, 1990­2001 thousands of people Total Urban Slum Total Urban Slum population population population population population population Country (1990) (1990) (1990) (2001) (2001) (2001) Algeria 24,855 12,776 1,508 30,841 17,801 2,101 Djibouti 504 408 n.a. 644 542 n.a. Egypt, Arab Rep. 56,223 24,999 14,087 69,080 29,475 11,762 Iran, Islamic Rep. 58,435 32,917 17,094 71,369 46,204 20,406 Iraq 17,271 12,027 6,825 23,584 15,907 9,026 Jordan 3,254 2,350 388 5,051 3,979 623 Lebanon 2,713 2,284 1,142 3,556 3,203 1,602 Libya 4,311 3,528 1,242 5,408 4,757 1,674 Morocco 24,624 11,917 4,457 30,430 17,082 5,579 Syrian Arab Republic 12,386 6,061 629 16,610 8,596 892 Tunisia 8,156 4,726 425 9,562 6,329 234 West Bank and Gaza 2,154 1,379 n.a. 3,311 2,222 1,333 Yemen, Rep. 11,590 2,648 1,787 19,114 4,778 3,110 Source: UN Habitat 2006. Note: n.a. = not applicable. 192 Poor Places, Thriving People FIGURE 4.16 National Population Growth Underlies Urban and Slum Population Growth in the Arab Republic of Egypt and Tunisia 8 y = 0.9201x + 0.0087 R 2 = 0.7394 6 annual rate of growth, 1990­2001 (%) 4 2 0 Egypt, Arab Rep. ­2 ­4 Tunisia ­6 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.0 annual rate of total national population growth, 1990­2001 (%) slum population growth, 1990­2001 urban population growth, 1990­2001 linear (urban population growth, 1990­2001) Source: UN Habitat 2006. sites, and to facilitate the consolidation of agricultural holdings. In MENA, however, there are signs that land markets could be performing better. Land for housing. The supply of housing in MENA economies does compare well with that of other economies at a similar level of develop- ment (World Bank 2005e). The per capita availability of floor space serviced with both water and sanitation is similar to that of comparator economies, with Jordan and the Islamic Republic of Iran faring slightly better than the trend and the Republic of Yemen, Morocco, Algeria, and Tunisia slightly below it. Note that these figures exclude any housing, including slums, without water or sanitation. However, there are problems of affordability and utilization of the housing stock. MENA's housing is expensive. In a typical OECD economy, the ratio of house prices to income is around 5 or below. In MENA, however, only Egypt and Tunisia come close to this ratio; Morocco, Lebanon, Jordan, Algeria, Lebanon, and the Islamic Republic of Iran are significantly Policy Package 1. Level the Playing Field and the Opportunity for Human Development in Lagging Areas 193 above it, in the range 6.5­17 (table 4.12). The affordability problem is most acute for low-income households (World Bank 2005e), and anec- dotal evidence suggests that it has been worsening over the past decade. Housing subsidies are sometimes massive but not necessarily reaching those who need them. Interest-rate subsidies, tax benefits, and land subsi- dies for house purchase and construction are found in Algeria, Egypt, the Islamic Republic of Iran, Jordan, Morocco, Tunisia, and the Republic of Yemen. The cumulative cost of these subsidies can be considerable, and is estimated at 5 percent of GDP for Morocco in 2002 and 7.5 percent of GDP in the Islamic Republic of Iran. Low-income households may have difficulty making the investments that give them access to these subsidies. In Algeria, for example, the poorest 25 percent of urban house- holds were estimated as only receiving 14 percent of all public housing assistance (World Bank 2002). This is not to say that governments cannot act effectively to provide affordable housing for low-income households. The Iranian and Tunisian governments have a record of success in this area. The key fea- ture was the existence of a dedicated institutional setup with a clear focus on the low-income segment rather than a broad-brush subsidization of the market (box 4.8). Underutilization is a characteristic of MENA's housing stock. In Morocco, according to the 2001 Housing Survey, 500,000 urban units were vacant. In Egypt, about 1.8 million housing units were vacant in 1996, of which 800,000 to 1 million units were in Cairo, where the vacancy rate was 14.5 percent. Lebanon also suffers from a vacancy rate TABLE 4.12 Housing Supply Indicators for Eight MENA Economies Floor area serviced with water and sanitation per person Country (square meters) House price to income ratio Algeria 8.5 (1993) 8.1 (2002 est.) Egypt, Arab Rep. 12 (1993) 4.9 (1998) Iran, Islamic Rep. 21 (2002) 9 (2002) Jordan 10 (1993) 6.5 (1998) Lebanon 10­12 (2003) 9 (2002) Morocco 6 (1993) 9.2 (1998) Tunisia 6.5 (1993) 5.2 (1998) Yemen, Rep. 4 10­17 (2000) Source: World Bank 2005e. Note: Floor area per person = median usable living space per person. House price to income ratio = ratio of the free market price of a dwelling unit to the median annual household income. 194 Poor Places, Thriving People BOX 4.8 Stories of Low-Income Housing Provision: Tunisia and the Islamic Republic of Iran One of the main objectives of the post-1979 revolutionary government in the Islamic Republic of Iran was to improve the supply of low-income housing. In the 10 years to 1989, 856 square kilometers of urban land were acquired by the Urban Land Organiza- tion (ULO), roughly half from the government and half by confiscation from the private sector according to the interpretation of Islamic Law that private ownership could not exist for mavat (undeveloped) land. The ULO engaged private contractors to provide this land with utilities and thus provided land to some 423,000 families, approximately 27 percent of housing units produced in the country. A criticism of the program is that much of the land acquired by the ULO was not allocated for development. This was in order to regulate migration flows to the larger cities, especially Tehran. The Tunisian government created the Urban Upgrading and Renovation Agency in 1981. By means of partnerships with municipalities and private developers, the agency managed to reduce the stock of slum housing from 24 percent of the total housing stock to 2 percent from 1974 to 1993. From 1975 to 1994, the percentage of household expen- diture devoted to housing went from 22 percent to 28 percent, despite the country's urbanization. The slum population halved between 1990 and 2001. These improvements have had a particularly beneficial impact on the quality of life and productivity of women. Key factors in Tunisia's success were: · The shift in the 1970s from slum eradication with relocation to slum regularization and upgrading · The creation in 1981 of the Agence de Réhabilitation et de Rénovation Urbaine as an execution agency for municipalities · The division of labor between the state (providing infrastructure and connections) and the residents (upgrading housing at their own pace), as exemplified by the upgrading of Kram West in the outskirts of Tunis Sources: Keivani, Mattingly, and Majedi 2008; Cities Alliance 2003; UN Habitat, 2006; Kraiem 2009. of 17 percent. These low utilization rates may be partly due to rent con- trols in the three countries, which discourage owners from renting out and maintaining their properties (World Bank 2005e). The high cost of urban housing and underutilization of the housing stock act as a brake on rural-urban migration. They, therefore, deny peo- ple in migrant-sending areas a key developmental pathway. Land for business. Surveys of MENA entrepreneurs indicate that they find difficulties in accessing land to be one of their main obstacles to Policy Package 1. Level the Playing Field and the Opportunity for Human Development in Lagging Areas 195 competitiveness. The World Bank's investment climate assessments found that the percentage of firms citing access to land as a "major" or "very severe" constraint was around 40 percent in Morocco, 35 percent in the Republic of Yemen, 35 percent in Algeria, 35 percent in Syria and 20 percent in Egypt (World Bank 2005f). Even when titled land has been identified, the purchase transaction itself is often burdensome. Table 4.13 shows the time and expense of purchasing a property that already has a registered and undisputed title. The MENA13 economies perform well on the time criterion, but worse than any region of the world except Sub-Saharan Africa in terms of the cost, with transactions costing more than 5 percent of the value of the property in Algeria, the Islamic Repub- lic of Iran, Jordan, Syria, Lebanon, Djibouti, and Tunisia. Algeria's 2002 Investment Climate Assessment highlighted how weak land markets impede competitiveness. Land access ranked as firms' third-biggest complaint. Indeed, 37 percent of firms surveyed said that they were currently searching for land, and they had been doing so for an average of five years. The key problems were inadequate titling and property rights; less than half of industrial land was titled, and business- people stated that corruption was affecting the allocation of government TABLE 4.13 The Time and Cost Required to Purchase Property: How the MENA13 Match Up against Other Regions Days Cost (% of property value) Algeria 51.0 7.5 Egypt, Arab Rep. 72.0 0.9 Iran, Islamic Rep. 36.0 10.0 Iraq 8.0 6.5 Jordan 22.0 10.0 Libya n.a. n.a. Morocco 47.0 4.9 Syrian Arab Republic 19.0 28.0 West Bank and Gaza 63.0 0.9 Yemen, Rep. 19.0 3.8 Lebanon 25.0 5.9 Djibouti 40.0 13.2 Tunisia 39.0 6.1 Average MENA13 36.8 8.1 East Asia and the Pacific 99.0 4.1 Eastern Europe and Central Asia 72.1 1.9 Latin America and the Caribbean 71.4 6.0 OECD 30.3 4.7 South Asia 106.0 5.9 Sub-Saharan Africa 95.6 10.5 Source: http://www.doingbusiness.org/ExploreTopics/RegisteringProperty/, accessed August 26, 2009. 196 Poor Places, Thriving People land. Yet at the same time, half of serviced industrial land was lying idle because of administrative controls and the lack of a secondary market (World Bank and IFC 2002). Land for the rural transition. We have seen that MENA economies' transition of the workforce out of agriculture and into services and man- ufacturing has been slower than one would expect for their level of devel- opment. Efficient agricultural land markets will be important to facilitate the shift of labor out of agriculture in the future, and to increase the earnings of those who remain in farming. Secure and transferable title over farmland enables farmers to consolidate landholdings, to mobilize agricultural finance, and to minimize the inefficiency of absentee owner- ship when landowners migrate. A key obstacle to the smooth functioning of MENA's agricultural land markets is the gap between the theory and the practice of land registra- tion. Legislation in economies as diverse as Tunisia and the Republic of Yemen stipulates that only registered land confers ownership rights under the formal judicial system. However, it is often difficult in practice for landowners to obtain and maintain registration; procedures are onerous, the capacity of land registries is limited, and Islamic inheritance rules require subdivision. The result is that the actual pattern of recognized ownership is not reflected in the official record. In the Republic of Yemen, only 13 percent of land held actually has a registered title (World Bank 2009c). Even with a concerted effort, the Tunisian government managed to raise the share of land with a registered title from 25 percent in the 1960s to 50 percent by 2001 (World Bank 2006a). The Egyptian govern- ment mandated a land registration system in 1923, but had achieved only 54 percent registration by 1994 (Hanigan and Ibrahim 1994). The widespread existence of unregistered land is not considered an impediment to day-to-day farming activities. The right to use the land can be confirmed by registration with customary authorities, as in the Republic of Yemen, or by the recognition of the community; however, the gap between the reality of ownership on the ground and the official register can create difficulties in accessing credit or consolidating frag- mented landholdings. Some MENA governments are concerned that agricultural land resources are being consumed by conversion for urban use. The situation varies considerably from country to country. While the ratio of land within urban extents to agricultural land is relatively high in Egypt (68 percent), the West Bank and Gaza (57 percent), Lebanon (55 percent), and Jordan (30 percent), it is 10 percent or less for all other economies of the MENA13.25 In most MENA13 economies, therefore, urban growth is unlikely severely to compromise the agricultural land base. Moreover, efforts to regulate against the development of farmland can harm the efficiency of land use. Figure 4.17 illustrates this point by the Policy Package 1. Level the Playing Field and the Opportunity for Human Development in Lagging Areas 197 FIGURE 4.17 Some of Cairo's Most Valuable Land Is Reserved for Agriculture a. Variation of developed land price with distance from the city center 1,100 1,000 900 value range of agricultural land in the study area 800 price of developed land per m2 700 600 500 400 300 200 100 0 0 5 10 15 20 25 30 35 40 45 50 55 60 65 distance from city center (km) b. Use of land by distance from city center 160 140 120 100 km2 80 60 40 20 0 5 10 15 20 25 30 35 40 45 50 55 60 65 distance from city center (Midan al Tahrir) desert agricultural Sheikh Zayed Badr al Shorkuk al Obour 1st of May 6th of October 10th of Ramadan new Cairo Cairo built-up Source: Presentation to the government of Arab Republic of Egypt by Alain Bertaud, December 2007, personal communication. 198 Poor Places, Thriving People distribution of land use in the city of Cairo. A sizeable area of land between 5 and 35 kilometers distant from the city center has been reserved for farming. Land in this zone potentially has a high market value, reflecting the fact that locations near the center of a major agglomeration exhibit high productivity for residential and commercial uses. Its reservation for agriculture is, therefore, a significant loss of value to the economy. A more efficient and effective approach to land conservation would be to relax regulatory limits on density that force many cities to sprawl outward. Rather than preventing peri-urban agricultural land from achieving higher value by being developed, governments can allow city- center land to achieve higher value by being densified. Services for Smooth Urban Growth The integration of new arrivals into the fabric of the city depends critically upon their ability to access public and private services: water and sanita- tion, power, solid waste collection and disposal, and urban transportation. If public authorities do not provide services to new arrivals, this absence does not slow down the pace of urbanization. On the contrary, it makes urbanization unnecessarily painful, with the emergence of slums, social exclusion, and a potential for political disaffection. It is not possible or appropriate to set out MENA's entire agenda for urban service development here, but it is useful to illustrate a few key principles with examples from the region. Planning ahead, rather than reacting to in-migration. There is sometimes a tendency for public authorities to delay the provision of services to new informal settlements. The delay might be because informal settlements are seen as temporary and unwelcome additions to the urban fabric. Or it might be because administrative procedures (e.g., planning, obtaining rights of way) are not adapted to informal settlements. However, the earlier that services for emerging settlements are planned, the more efficient investments in road, power, water, and san- itation networks will be. But once an ad hoc pattern of settlement has been established, it can be difficult for utilities to provide services cost- effectively or to use their existing administrative procedures. The expe- rience of providing water and sanitation services to illegal peri-urban settlements in Morocco is a case in point (Chauvot de Beauchęne and Mantovani 2009). Many of difficulties encountered in bringing services to these slums were because the utilities were coming in at the end of the settlement process. Peri-urban areas were not included in the serv- ice areas of water and sanitation operators, lack of formal title made it difficult for them to obtain the rights of way needed to install services, Policy Package 1. Level the Playing Field and the Opportunity for Human Development in Lagging Areas 199 and unplanned construction made access difficult for their workers. This finding contrasts with the experience of new informal settlements in the Islamic Republic of Iran, where the government seems to have made significant investments in services at an earlier stage (Alaedini, Deichmann, and Shahriari 2007). With global warming, MENA cities' vulnerability to sea-level rise gives the region a special reason to care that urban growth should be planned. Of all world regions, MENA (including GCC economies) will have the highest share of its population (53 percent) living in coastal areas that are vulnerable to storm surges (Dasgupta, Laplante, Murray, and Wheeler 2009). Urban planning can both minimize construction in vulnerable areas and ensure investment in protective infrastructure, thus reducing future costs of catastrophic flooding. The risk of pre-servicing urban residential land, however, is that it becomes overpriced and captured by elite groups at the expense of low- income people--who will occupy unplanned areas elsewhere. It is, there- fore, important for the authorities to mandate the levels of density required for housing to be truly low-cost. Responding to the "spatial flow" of market-led city growth. It is also un- realistic to hope that government can reshape the geography of urban growth. As elsewhere in the world, MENA governments have sometimes attempted to limit the development of major cities by creating new ser- viced settlements in less agglomerated areas. The general lesson is that service-provision and new town creation should follow people's settle- ment preferences, rather than attempting to lead them. Informal settle- ments should, therefore, be seen as "part of the housing solution rather than a problem" (World Bank 2005e). The Islamic Republic of Iran has created 18 new towns since the Rev- olution, with the aim of protecting farmland and preventing informal settlements around major cities (Ziari 2006). The new towns were con- structed at considerable distance from their parent cities, an average of 40 kilometers for the four new towns under construction around Tehran and an average of 35 kilometers for the three around Isfahan. These new towns, however, became dormitory communities for salaried employees of the mother city, without any economic base of their own. The new towns' total population has only reached 5 percent of the target. Sig- nificantly, the only exception was Baharestan; its closeness to Isfahan (15 kilometers) made it the most successful new town, and it accounts for 35 percent of the residential units under construction in the entire program. Meanwhile, the main growth in Tehran is in informal peri- urban settlements, where 60 percent of new inhabitants are long-term urban residents unable to find a place to live in the formal areas (Zebar- dast 2006) The authorities have, therefore, refocused their attention 200 Poor Places, Thriving People away from new towns and toward increasing the supply of formal hous- ing within the metropolis itself (box 4.8). The Egyptian government's 1969 Greater Cairo Region Master Scheme called for satellite cities in the desert around Cairo. President Sadat in 1974 called for a "new population map of Egypt." The new ring road around Cairo was intended to limit the city's growth. The 10th of Ramadan City, Sadat City, and 6th of October City were constructed in 1977, 1978, and 1981, respectively. As in the Iranian case, they were built distant (an average of 60 kilometers) from the parent city. Unlike the new Iranian towns, however, these satellite cities in the desert attracted sig- nificant levels of business thanks to incentives, including a 10-year tax holiday and free zone status. The problem, however, was that people chose not to live in them because of the difficulties in finding affordable housing, transportation, water, social facilities, schools, shops, and other services, as well as communities and social networks (Stewart 1996). A pattern of reverse commuting, therefore, evolved, with firms busing in their workforces from Cairo. By 2005, the building completion rate in residential subdivisions of 6th of October City ranged from 2 percent to 20 percent (World Bank 2008e). Meanwhile, however, the forces of agglomeration were accelerating the growth of Cairo's metropolis; established areas such as Mohandes- sine, Zamalek, Maadi, and Madinat Nasr were growing vertically and densifying, and adjacent informal settlements were expanding laterally. About 60 percent of Cairo's population was probably in informal settle- ments, and 81 percent of those were on agricultural land (World Bank 2008e). By the early 1990s, therefore, the government was refocusing its efforts on using services to improve the quality of growth in the metrop- olis rather than to create a "new population map" of alternative agglom- erations. The emphasis was now on slum upgrading, the integration of Cairo's adjacent suburbs into the metropolis, and the creation of smaller industrial sites within existing agglomerations such as Basatin, Helwan, and Shubra El Kheima (Sutton and Fahmi 2001). El-Batran and Arandel make the distinction between Egypt's satellite new towns of the 1970s and 1980s and the "new settlements" approach of the 1990s. The new settlements' objective is to learn from informal settlements and emulate them on public desert land. The new settlements rely more heavily on the concept of sites and services than the new towns and incorporate core housing schemes. The target population is the people currently occupying spontaneous settlements, namely low- and middle-income dwellers who are excluded from the formal housing market. New settlements are located in close proximity to Cairo and allow easy access to the city. (El-Batran and Arandel 1998) Policy Package 1. Level the Playing Field and the Opportunity for Human Development in Lagging Areas 201 In both the Iranian and Egyptian cases, therefore, the authorities seem to have shifted their approach. Both recognized the futility of re- directing metropolitan growth into less agglomerated areas. The efforts of both are now instead concentrated on using service provision to improve the quality of metropolitan agglomeration. Institutional Arrangements for Smooth Urban Growth If the quality of planning, regulation, and service provision determine whether urbanization is smooth or traumatic, it is the institutional frame- work for municipalities that largely determines the quality of planning, regulation, and service-provision. MENA's high administrative and fiscal centralization makes it hard for municipalities to coordinate and finance urban growth. Decentralization, organizational consolidation, financial management, and capacity building are key issues for strengthening MENA municipalities' ability to respond to the challenge of urban growth (box 4.9). Examples of initiatives in the MENA region are listed in table 4.14. Can It Work? How long will the stresses of urbanization last? Today's wealthy cities, such as Paris, Melbourne, London, New York, Shanghai, Stockholm, and Helsinki, were not always so elegant. During the rapid urbanization of the nineteenth and early twentieth centuries, they were infested with slums, filth, and infection (World Bank 2009a, 69). Many decades elapsed between the initial burst of slum growth and the integration of slum populations into the fabric of the city. Do MENA's cities have to wait so long? Fortunately, we now have examples of quicker transitions. Cities such as Seoul; Singapore; Hong Kong SAR, China; and San José (Costa Rica) are proving that good policies really can smooth the path for urban growth and can shorten the period of painful transition. · Institutional arrangements for smooth urban growth. Costa Rica and the Republic of Korea's regional government boundaries and institu- tional mandates were reconfigured at an early stage of city growth to serve the need for integrated city management. · Land transactions. In Costa Rica, 80 percent of property owners possess registered titles. In Hong Kong SAR, China, flexible land markets were crucial in balancing the demands of residential and commercial real estate. · Investing in people. Costa Rica invests 20 percent of its budget in edu- cation, enabling the social mobility that discourages the emergence 202 Poor Places, Thriving People BOX 4.9 Facilitating Smoother Urbanization: Six Ideas for Policy Makers 1. Simplifying land development regulations and basing zoning regulations on consid- erations of environmental impacts, connectivity, and amenity values rather than as an attempt to slow down urban development. Inappropriate standards, sometimes bor- rowed from more developed countries, prohibit the construction of affordable hous- ing. For example, minimum plot sizes in Morocco and Jordan are beyond the budgets of most people there (World Bank 2005e). 2. Shifting public finance away from subsidies for construction and purchase, toward public-private partnerships for affordable housing. 3. Simplifying titling and registration systems and integrating informal arrangements into formal legislation. Tunisia's Certificat de Possession and the Arab Republic of Egypt's Hekr System and Hand Claim with Property Taxes, for example, are practi- cal ways of granting defined rights to rural and urban land occupants without formal written title.25 The Islamic Republic of Iran's informal "Promissory Notes" are now the main form of tenure in Tehran's fringe settlements (Zebardast 2006). 4. Releasing government-owned urban land into the market, where it represents a pool of underused real estate and is doing so transparently. This approach could apply to Egypt, Algeria, and the Islamic Republic of Iran. 5. Cutting the red tape and bureaucracy associated with land sales. 6. Planning the provision of urban services in a way that supports the spatial "flow" of city growth: in other words, slum upgrading, incorporation of informal areas into service networks, and regulations facilitating densification of established areas. Source: Study Team. of an urban underclass. Seoul's period of rapid urbanization, 1960­90, coincided with a 120 percent increase in the Korean work- force's average years of education. · Connectivity makes it unnecessary for migrants to crowd near places of employment and allows slum dwellers to integrate into the urban economy. San José, Costa Rica, has good affordable bus services within the city, to sub- urbs, and to neighboring cities. Seoul's radial transport links, including a subway and a beltway, kept pace with growth in the urban extent. · Wisely targeted housing expenditure. Costa Rica's housing subsidies were means tested, so they were not received by the better off. Eighty-six percent of Singapore's population lives in privately owned but publicly built units. During 1960­80, the government in Hong Kong SAR, Policy Package 1. Level the Playing Field and the Opportunity for Human Development in Lagging Areas 203 Table 4.14 How MENA Is Upgrading the Institutional Environment for Urban Growth Approach What MENA governments are doing Establishing spatially integrated Cairo's boundaries as a spatial entity now extend well beyond Cairo planning institutions governorate into adjoining jurisdictions: since 2008 Greater Cairo has included Cairo, Giza, Helwan, and 6th of October governorates. Law no. 119/2008 aims to address this fragmentation, calling for a National Higher Council for Urban Planning and Development. Matching resources and The problem of unfunded municipal mandates occurs frequently across mandates the region. Tunisia's 10th Development Plan estimated that around half the country's municipalities were in a "weak" or "vulnerable" financial condition.a The government therefore set the target of doubling the share of municipalities collecting the TCL, a tax on industrial, commercial, and professional entities. Improving access to finance for If central governments are wary of the financial risk of municipal borrowing on capital investment the open market, other mechanisms may be developed. Morocco's Fonds d'Equipement Communal acts as the central government's banker toward municipalities. Its loans for municipal investments were around MAD 7.5 billion at the end of 2008.b Jordan's Cities and Villages Develop- ment Bank fulfills a similar role. Tunisia has the Caisse des Pręts et de Soutien des Collectivités Locales. Another option, currently pursued under the Arab Republic of Egypt's Affordable Mortgage Finance program, is to ease low-income households' access to housing loans. Strengthening managerial In 2001, the Jordanian government responded to a crisis in municipal finances capacity at municipal level and services by overhauling the structure of municipal government. Under the so-called Amalgamation, the number of councils was reduced from 190 to 99, 762 technical and senior administrative staff were recruited, a management information system was set up, intensive training was provided to staff, and a permanent municipal training center was created.c Source: Study Team, from World Bank project documents and http://www.fec.ma. Note: a. Project Appraisal Document for World Bank Municipal Development III Project. b. See http://www.fec.ma. c. Project Appraisal Document for World Bank Regional and Municipal Development Project. China contracted out urban development to a specialist organization dominated by private sector interests. No successful city has avoided the growing pains of urbanization alto- gether. But international experience does show that the transition can be smoother and quicker if institutions have adequate mandates, if land markets work, if the workforce is educated and mobile, and if public housing expenditure targets affordability for the poor. Endnotes 1. The sign of the coefficient on trade openness depended on the choice of independent variable: growth of the main city, the urbanization rate, or the agglomeration index. 204 Poor Places, Thriving People 2. At the beginning of the twentieth century, large swathes of the Middle East were still under Ottoman rule, as part of a global empire that had existed for roughly 400 years. The Ottoman Empire included extensive territories known today as the Syrian Arab Republic, Iraq, Lebanon, Jordan, the Republic of Yemen, Israel, and Palestine, in addition to important footholds in Libya, although its other North African dominions had already been lost to the French (Tunisia and Algeria) and to the British (the Arab Republic of Egypt) (Owen 1992, 8). 3. This analysis uses linear and probit econometric models. Although it is dif- ficult to separate the impacts of different variables, the robustness of the analysis and the low degree of correlation among independent variables are adequate to identify the determinants of spatial disparities. See Katayama (2009) for details. 4. In the analysis of TIMSS 2007, "rural" is defined as a school location in which 15,000 people or fewer reside. 5. World Bank. 2009. "Improving Health Services for All in Libya." In Libya Public Expenditure Review. Washington, DC: World Bank Group, p. 142. 6. Regional Health Systems Observatory, World Health Organization (WHO). 2006. Lebanon: Health Systems Profile. Cairo: WHO Regional Office for the Eastern Mediterranean, p. 38. 7. Regional Health Systems Observatory, WHO. 2006. Djibouti: Health Sys- tems Profile. Cairo: WHO Regional Office for the Eastern Mediterranean, p. 87. 8. Regional Health Systems Observatory, World Health Organization (WHO). 2006. Yemen: Health Systems Profile. Cairo: WHO Regional Office for the Eastern Mediterranean, p. 51. 9. Regional Health Systems Observatory, WHO. 2006. Iraq: Health Systems Profile. Cairo: WHO Regional Office for the Eastern Mediterranean, p. 35. 10. Department of Statistics [Jordan] and Macro International Inc. "Jordan Population and Family Health Survey 2007." Calverton, MD: Macro Interna- tional Inc., p. 33. 11. World Bank. 2009. "Improving Health Services for All in Libya." In Libya Public Expenditure Review. Washington, DC: World Bank Group. 12. El-Saharty, Sameh and others. 2006. Tunisia Health Sector Study. Washing- ton, DC: MNSHD, World Bank Group, p. 44. 13. World Bank. 2009. Kingdom of Morocco--Health Policy Note. Washington, DC: World Bank, p. 15. 14. Regional Health Systems Observatory, WHO. 2006. Morocco: Health Systems Profile. Cairo: WHO Regional Office for the Eastern Mediterranean, p. 34. 15. Regional Health Systems Observatory, WHO. 2006. Yemen: Health Systems Profile. Cairo: WHO Regional Office for the Eastern Mediterranean, p. 3. 16. National Health Accounts Team, Republic of Yemen, and Partners for Health Reform Plus. 2006. Yemen National Health Accounts: Estimate for 2003. Bethesda, MD: Partners for Health Reform Plus Project, Abt Associates Inc., p. xv. 17. El-Zanaty, Fatma, and Ann Way. 2006. Egypt Demographic and Health Survey 2005. Cairo: Ministry of Health and Population, National Population Council, El-Zanaty and Associates, and ORC Macro, p. 209. 18. Salti, N., and J. Chaaban. 2008. "Health Equity in Lebanon: A Micro- economic Analysis." Unpublished, p. 28. 19. Regional Health Systems Observatory, WHO. 2006. Yemen: Health Systems Profile. Cairo: WHO Regional Office for the Eastern Mediterranean, p. 69. 20. Marmot, M. G. and others. 1991. "Health Inequalities among British Civil Servants: The Whitehall II Study." Lancet 337: 1387­93. Policy Package 1. Level the Playing Field and the Opportunity for Human Development in Lagging Areas 205 21. Salehi-Isfahani, D., S. Fesharaki, and M. Majbouri. 2006. "Analysis of Health and Education Expenditures in Iran." Unpublished, p. 9. 22. "2006 Household Expenditures and Income Survey." Hashemite Kingdom of Jordan, Department of Statistics, Amman. 23. Nicknamed "26-26," after the bank account to which citizens make volun- tary contributions. 24. GRUMP database and World Bank World Development Indicators. Con- sistent data are not available for Morocco; the GRUMP database aggregates Morocco with Western Sahara, while the World Bank WDI does not. 25. See http://www.habitat.org/gov/suppdocs/chapter4_strategies_that_improve_ tenure_security.pdf, accessed August 27, 2009; and World Bank 2005e. References Ades, A., and E. L. Glaeser. 1995. "Trade and Circuses: Explaining Urban Giants." Quarterly Journal of Economics 110 (1): 195­227. Aghajanian, A., A. H. Mehryar, S. Ahmadnia, and S. Kazemipour. 2008. "Impact of the Rural Health Development Programme in the Islamic Republic of Iran on Rural-Urban Disparities in Health Indicators." World Hospitals and Health Services 44 (3): 10. Alaedini, P., U. Deichmann, and H. Shahriari. 2007. Land Markets and Housing Dynamics in Low Income Settlements in Iran: Examining Data from Three Cities. http://www.worldbank.org/urban/symposium2007/ papers/alaedini.pdf. Ayubi, N. 1988. "Arab Bureaucracies: Expanding Size, Changing Roles." In Beyond Coercion: The Durability of the Arab State, ed. I. Zartman and A. Dawisan. London: Croom Helm. ------. 2008. Overstating the Arab State. London: I. B. Tauris. Beblawi, H., and G. Luciani. 1987. The Rentier State. New York: Croom Helm. Bengio, Ofra, and Gabriel Ben-Dor, eds. 1999. Minorities and the State in the Arab World. Boulder, CO: Lynne Rienner. Blomquist, J. 2008. "Social Safety Nets and the Response to Food, Fuel, and Other Price Shocks in the Middle East and North Africa Region." Unpublished. Blomquist, J., and N. Kabbani. 2007. "Providing Income Support to the Poor in Syria: Simulating Targeting Options." Unpublished draft report. February. Brown, C. 1997. Imperial Legacy: The Ottoman Imprint on the Balkans and the Middle East. New York: Columbia University Press. 206 Poor Places, Thriving People Chauvot de Beauchęne, X., and P. Mantovani. 2009. "Subsidies for the Poor: An Innovative Output-Based Approach. Providing Basic Services to Poor Periurban Neighborhoods in Morocco." In Water in the Arab World. Management Perspectives and Innovations, ed. V. Jagannathan, A. M. Shawky, and A. Kremer, 335­53. Washington, DC: World Bank. Cities Alliance. 2003. "2003 Annual Report." World Bank, Washington, DC. Cleveland, William L. 1999. A History of the Modern Middle East. Boulder, CO: Westview Press. Coady, D., M. Grosh, and J. Hoddinott. 2004. Targeting of Transfers in Developing Countries: Review of Lessons and Experience. Washington, DC: World Bank. CSDH (Commission on Social Determinants of Health). 2008. Closing the Gap in a Generation: Health Equity through Action on the Social Deter- minants of Health. Geneva: World Health Organization. Dasgupta, S., B. Laplante, S. Murray, and D. Wheeler. 2009. Climate Change and the Future Impacts of Storm-Surge Disasters in Developing Countries. Washington, DC: Center for Global Development. Dasgupta, S., C. Meisner, A. Makokha, and R. Pollard. 2009. "Commu- nity Management of Rural Water Supply: Evaluation of User Satis- faction in Yemen." In Water in the Arab World: Management Perspectives and Innovations, ed. V. Jagannathan, A. Shawky, and A. Kremer, 383­99. Washington, DC: World Bank. Dollar, D., M. Hallward-Driemeier, and T. Mengistae. 2005. "Investment Climate and Firm Performance in Developing Countries." Economic Development and Cultural Change 54 (1): 1­31. Douidich, M., A. Ezzrari, A., and P. Lanjouw. 2008. Simulating the Impact of Geographic Targeting on Poverty Alleviation in Morocco. Washington, DC: World Bank. Economist. 2005. The Economist Pocket World in Figures 2005. London. Eickleman, D. 1976. Moroccan Islam: Tradition and Society in a Pilgrimage Center. Austin: University of Texas Press. El-Batran, M., and C. Arandel. 1998. "A Shelter of Their Own: Informal Settlement Expansion in Greater Cairo and Government Responses." Environment and Urbanization 10 (1): 217­32. Policy Package 1. Level the Playing Field and the Opportunity for Human Development in Lagging Areas 207 FEMISE. 2005. Jordan Country Profile: The Road Ahead for Jordan. Cairo: Economic Research Forum and Institut de la Méditerranée. Findlay, C. 1980. Bureaucratic Reform in the Ottoman Empire: The Sublime Porte, 1789­1922. Princeton, NJ: Princeton University Press. ------. 1997. "The Ottoman Administrative Legacy and the Middle East." In Imperial Legacy: the Ottoman Imprint on the Balkans and the Middle East, ed. C. Brown. New York: Columbia University Press. Fiszbein, A., and M. Schady. 2009. Conditional Cash Transfers: Reducing Present and Future Poverty. Washington, DC: World Bank. Freedom House. 2009. "Freedom in the World." http://www.freedom house.org. Georgetown University. 2007. Engagement of Social Change Agents in Improving Women's Health in the Arab Republic of Egypt: Developing a Framework for Evaluation and Research. Washington, DC: Georgetown University, School of Nursing and Health Studies. Golnar, M. 1995. Girls' Drop-out from Primary Schooling in the Middle East and North Africa: Challenges and Alternatives. Amman: UNICEF. Goujon, A., H. Alkitkat, W. Lutz, and I. Prommer. 2007. Population and Human Capital Growth in Egypt: Projections for Governorates to 2051. Luxenburg: International Institute for Applied Systems Analysis. Hanigan, F. L., and M. M. Ibrahim. 1994. "Completing Egypt's Agricul- tural Cadaster." MN GIS/LIS News (St. Paul), 422­33. Henry, C. M. 2004. "Algeria's Agonies: Oil Rent Effects in a Bunker State." Journal of North African Studies 9 (2): 68­81. HM Treasury. 2009. "Analysis of Public Expenditure by Country and Region." http://www.hm-treasury.gov.uk/d/10(1).pdf. Huyer, S., N. Hafkin, H. Ertl, and H. Dryburgh. 2005. "Women in the Information Society." In From the Digital Divide to Digital Opportuni- ties: Measuring Infostates for Development, ed. G. Sciadis, chapter 6. Montreal: Orbicom. IDSC (Information and Decision Support Center). 2007. Cars in Egypt. Cairo: IDSC. INS (Institut National de la Statistique). 2000. Enquęte Nationale sur le Budget. La Consommation et le Niveau de Vie des Ménages. Tunis: INS. 208 Poor Places, Thriving People Jacobs, D., J.-L. Hélis, and D. Bouley. 2008. Budget Classification. Washington, DC: International Monetary Fund. Karaman, K. K. 2009. "Decentralized Coercion and Self-Restraint in Provincial Taxation: The Ottoman Empire, 15th­16th Centuries." Journal of Economic Behavior and Organization 71 (3): 690­703. Katayama, H. 2009. "Geographical Disparities in Enrollment, Attainment and Learning Achievements in the MENA Region." Unpublished. Kaufman, D., A. Kraay, and M. Mastruzzi. 2009. Governance Matters VIII. Aggregate and Individual Governance Indicators 1996­2008. Washington, DC: World Bank. Keivani, R., M. Mattingly, and H. Majedi. 2008. "Public Management of Urban Land, Enabling Markets, and Low-Income Housing Pro- vision: The Overlooked Experience of Iran." Urban Studies 45: 1825­53. Khelladi, M. 2009. "Inégalités spatiales en Algérie: tendances et poli- tiques publiques." Unpublished background paper for this report. Korany, B. 1987. "Alien and Besieged Yet Here to Stay: The Contradic- tion of the Arab Territorial State." In The Foundations of the Arab State, ed. G. Salamé, 47­74. London: Croom Helm. Kraiem, F. 2009. "Tunisia Slum Upgrading Note." Unpublished. Kuran, T. 2004. "Why the Middle East Is Economically Underdevel- oped: Historical Mechanisms of Institutional Stagnation." Journal of Economic Perspectives 18: 71­90. Lall, S. V., and T. Mengistae. 2005. The Impact of Business Environment and Economic Geography on Plant-Level Productivity: An Analysis of Indian Industry. Washington, DC: World Bank. Lall, S., C. Timmins, and S. Yu. 2009. Connecting Lagging and Leading Regions: The Role of Labor Mobility. Washington, DC: World Bank. Leite, P. 2008. "Implementing a Proxy Means-Test to Identify the Poor- est in Yemen on the Basis of the Household Budget Survey 05/06 (HBS)." Unpublished. Luciani, G. 1989. The Arab State. Berkeley: University of California Press. Mengistae, T., and F. Thompson. 2006. "Unbundling Location Effects in Productivity: The Case of Manufacturing Plants in Morocco." Policy Package 1. Level the Playing Field and the Opportunity for Human Development in Lagging Areas 209 Center for the Study of African Economies. http://www.csae.ox.ac.uk/ conferences/2006-EOI-RPI/papers/csae/Thompson-Mengistae.pdf. Ministčre de la Santé. 2003, 2007. Santé en Chiffres. Rabat. Ministry of Planning and Development Cooperation. 2009. Iraq House- hold Socio-Economic Survey. Baghdad, Iraq. Movahedi, M., B. Hajarizadeh, A. Rahimi, M. Arshinehi, K. Amirhos- seini, and A. A. Hadhgoost. 2009. "Trends and Geographical Inequal- ities of the Main Health Indicators for Rural Iran." Health Policy and Planning 24: 229­37. North, D. 1990. Institutions, Institutional Change, and Economic Perform- ance. New York: Cambridge University Press. Owen, R. 1992. State, Power, and Politics in the Making of the Modern Mid- dle East. London: Routledge. Ozbudun, E. 1997. "The Continuing Ottoman Legacy: The State Tra- dition in the Middle East." In Imperial Legacy: the Ottoman Imprint on the Balkans and the Middle East, ed. C. Brown, 133­57. New York: Columbia University Press. Pennell, C. R. 2001. Morocco since 1830: A History. London: Hurst. Political Instability Task Force: M. Marshall and G. Cole. 2009. "Global Report on Conflict, Governance, and State Fragility." George Mason University, Fairfax, Virginia. Rawlings, L. B., and G. M. Rubio. 2005. "Evaluating the Impact of Conditional Cash Transfer Programs." World Bank Research Observer 20 (1): 29­55. Rawlings, L., L. Sherburne-Benz, and J. van Domelen. 2004. Evaluating Social Funds--A Cross-Country Analysis of Community Investments. Washington, DC: World Bank. Richards, A., and J. Waterbury. 1990. A Political Economy of the Middle East: State, Class, and Economic Development. Boulder, CO: Westview Press. Sadrizadeh, B. 2004. "Strengthening Primary Health Care in the Islamic Republic of Iran." In Public Health in the Middle East and North Africa, ed. A. M. Pierre-Louis, F. A. Akala, and H. Karam, 115­25. Wash- ington, DC: World Bank Institute. Sapir, A. 2003. "An Agenda for a Growing Europe: Making the EU Eco- nomic System Deliver." Report of an Independent High-Level Study 210 Poor Places, Thriving People Group Established on the Initiative of the President of the European Commission. Schiavo-Campo, S., and D. Tommasi. 1999. Managing Public Expendi- ture. Manila: Asian Development Bank. Shah, A. 2007. "A Practitioner's Guide to Intergovernmental Fiscal Transfers." In Intergovernmental Fiscal Transfers: Principles and Practice, ed. R. Boadway and A. Shah, 1­54. Washington, DC: World Bank. Shlash, A. 2009. "Iraq High Committee for Strategy Reduction. Draft Strategy Review Presentation." June. Amman. Soliman, A., A. Shawky, M. Hamed, W. Wakeman, and M. Mehany. 2009. "Rural Sanitation within an IWRM Framework: Case Study of Application in the Delta Region, Egypt." In Water in the Arab World: Management Perspectives and Innovations, ed. V. Jagannathan, A. Shawky, and A. Kremer, 401­20. Washington, DC: World Bank. Stern, N. 2002. A Strategy for Development. Washington, DC: World Bank. Stewart, D. J. 1996. "Cities in the Desert: The Egyptian New-Town Program." Annals of the American Association of Geographers 86 (3): 459­80. Sutton, K., and W. Fahmi. 2001. "Cairo's Urban Growth and Strategic Master Plans in the Light of Egypt's 1996 Population Census Results." Cities 18 (3): 135­49. Thomas, M. 2005. The French Empire between the Wars: Imperialism, Politics, and Society. Manchester: Manchester University Press. Tosun, M. S., and S. Yilmaz. 2008. Centralization, Decentralization, and Con- flict in the Middle East and North Africa. Washington, DC: World Bank. ------. 2009. "Decentralization and Spatial Dimension of Public Expenditures in MENA Countries." Unpublished background paper for this report. UN Habitat. 2006. State of the World's Cities 2006/7: The Millennium Development Goals and Urban Sustainability. Nairobi: United Nations Human Settlements Program. UNDP (United Nations Development Programme). 2005. Egypt Human Development Report 2005. New York. UNICEF (United Nations Children's Fund). 2009. The State of the World's Children 2009: Maternal and Newborn Health. Annual Report. New York: UNICEF. Policy Package 1. Level the Playing Field and the Opportunity for Human Development in Lagging Areas 211 WHO (World Health Organization) and DFID (Department for Inter- national Development). 2003. The Analytic Review of the Integrated Management of Childhood Illness Strategy: Final Report. Geneva: WHO. World Bank. 2002. Algeria Low Income Housing Project. Midterm Review Mission Report. Washington, DC: World Bank. ------. 2004a. The Geographical Dimension of Public Expenditures and Its Link to Poverty: The Case of Madagascar. Retrieved August 24, 2009, from http://siteresources.worldbank.org/INTPGI/Resources/342674- 1092157888460/Galasso.GeographicalMadagascar.pdf. ------. 2004b. Implementation Completion Report, Morocco: Social Priorities Program Basic Health Project. Report No. 28390-MOR. Washington, DC: World Bank. ------. 2005a. Egypt--Toward a More Effective Social Policy: Subsidies and Social Safety Nets. Washington, DC: World Bank. ------. 2005b. Iraq--Social Protection in Transition: Labor Policy, Safety Nets and Pensions. Washington, DC: World Bank. ------. 2005c. Local Administration, Fiscal Intergovernmental Arrangements, and Fiscal Decentralization in Egypt. Washington, DC: World Bank. ------. 2005d. Decentralization and the Intergovernmental System in Iran. Washington, DC: World Bank. ------. 2005e. The Macroeconomic and Sectoral Performance of Housing Supply Policies in Selected MENA Countries: A Comparative Analysis. Washington, DC: World Bank. ------. 2005f. "Forum on Investment Climate, Egypt." Retrieved August 27, 2009, from http://siteresources.worldbank.org/INTMN AREGTOPPRVSEC/Resources/Land.pdf. ------. 2005g. Implementation Completion Report, West Bank and Gaza: Health System Development Project. Report No. 33572. Washington, DC: World Bank. ------. 2005h. Implementation Completion Report on a Credit in the Amount of US$17.2 Million to the Arab Republic of Egypt for a Popula- tion Project. Report No. 28300-IDA. Washington, DC: World Bank. ------. 2006a. Tunisia: Agricultural Policy Review. Washington, DC: World Bank. ------. 2006b. Upper Egypt: Challenges and Priorities for Rural Development. Washington, DC: World Bank. 212 Poor Places, Thriving People ------. 2006c. Implementation Completion Report, Yemen: Child Development Project. Report 35003. Washington, DC: World Bank. ------. 2007a. Kingdom of Morocco: Moving out of Poverty in Morocco. Washington, DC: World Bank. ------. 2007b. Morocco Skills Development and Social Protection within an Integrated Strategy for Employment Creation. Washington, DC: World Bank. ------. 2007c. Syria Arab Republic: Toward a Social Protection Strategy for a Social Market Economy. Washington, DC: World Bank. ------. 2007d. Hashemite Kingdom of Jordan: A Note on Strategy on Mod- ernization of the Social Safety Nets. Washington, DC: World Bank. ------. 2008a. The Road Not Traveled. Education Reform in the Middle East and North Africa. Washington, DC: World Bank. ------. 2008b. Using PISA to Understand the Determinants of Learning in the Middle East and North Africa. Washington, DC: World Bank. ------. 2008c. Kingdom of Morocco Policy Notes: Conditions for Higher and Inclusive Growth. Washington, DC: World Bank. ------. 2008d. For Protection and Promotion: The Design and Implementa- tion of Effective Safety Nets. Washington, DC: World Bank. ------. 2008e. Arab Republic of Egypt. Towards an Urban Sector Strategy. Washington, DC: World Bank. ------. 2009a. World Development Report 2009: Reshaping Economic Geog- raphy. Washington DC: World Bank. ------. 2009b. Tapping a Hidden Resource: Energy Efficiency in the Middle East and North Africa. Washington, DC: World Bank. ------. 2009c. "Yemen Synthesis Rural Land Tenure." Unpublished. ------. 2009d. From Privilege to Competition: Unlocking Private-led Growth in the Middle East and North Africa. Washington, DC: World Bank. World Bank and IFC (International Finance Corporation). 2002. Algeria Investment Climate Assessment. Washington, DC: World Bank and IFC. Zebardast, E. 2006. "Marginalization of the Urban Poor and the Expan- sion of the Spontaneous Settlements on the Tehran Metropolitan Fringe." Cities 23 (6): 439­54. Ziari, K. 2006. "The Planning and Functioning of New Towns in Iran." Cities 23 (6): 412­22. CHAPTER 5 Policy Package 2. Connecting Poor Places to the Poles of Development Policy Package 2 (Adapted from table 2.3) Package 2. "Infrastructure" For growth to Lagging areas Beware of Connecting ("spatially spillover close thinning- poor places connective from eco- enough to out effect to the poles policies") nomic poles density [and in low- of develop- into lagging with enough potential ment. areas, they economic areas. have to be density] to connected. benefit The Power of Spillovers spillover A. n. That which spills over; the process of spilling over; (an) incidental development; a consequence, a repercussion, a by- product. --Oxford English Dictionary Economic development spills out from leading areas into the areas around them. In terms of economic geography, the agglomeration economies from being close to an economic hub are second only to those of being within the hub itself. The "belt" around an agglomeration can sell its crops and its manufactured goods to the agglomeration without having to subtract long-distance haulage charges from its profit margin. Its workers can commute to the agglomeration and use the agglomeration's markets and health facilities. Its employers can use the agglomeration's financial institutions, shop around for the best deal there for inputs (and not have to worry about transport costs and delivery delays), do business face-to- face with suppliers and clients--and drop into a senior official's office if necessary. Waldo Tobler's first law of geography sums it up: "everything is related to everything else, but near things are more related than distant things" (Tobler 1970). 213 214 Poor Places, Thriving People Spillovers from agglomerations into their peripheries are therefore one of the most powerful forces in spatial development. A key predictor of how prosperous a place becomes is how prosperous the neighboring places are; this is true whether one is talking about the countries of Africa and the Middle East (Parent and Zouache 2008), the provinces of Indonesia (McCulloch and Sjahrir 2008), or the regions of Algeria (Belarbi and Zouache 2009). In the Arab Republic of Egypt, the city gov- ernorates of Cairo, Alexandria, and Port Said had (unweighted) average per capita household expenditures of LE 4,813 per annum in 2005. For the governorates that bordered these metro areas,1 average per capita household expenditures were LE 3,260. For the other governorates, the average was LE 2,632. Spillover effects, therefore, have the potential to reduce spatial dis- parities if they yoke lagging areas to the growth of leading regions. According to the principles of economic geography, to benefit most from these spillover effects, a lagging area must meet two criteria (World Bank 2009a, 76, 246): · Proximity to a leading area agglomeration · Population density, so that economies of agglomeration can build up within the lagging area A Geographical Information System (GIS) can give us an idea of where MENA's lagging areas fit these criteria. MENA has approximately 55 cities of 500,000 inhabitants or more. Using road maps, relief maps, and assumptions about the speed of road travel, the travel time from every location in MENA13 to a city of 500,000 or more inhabitants was calculated. Of course, these estimates are necessarily approximate. Digital data on minor roads and on road quality are not available, and the distribution of population within administrative divisions has to be derived from satellite images of nighttime lighting. But they still give us a rough overview of MENA's connectivity situation (which is shown in map 5.1). The analysis tells us that MENA has a proximity advantage (table 5.1): most of its population live in cities or close to them. It is estimated that 59 percent of MENA's population live within two hours of one of these cities and 72 percent within three hours (unweighted country average). Only in Algeria does more than half the population appear to live more than three hours away from a city. And some lagging areas benefit from proximity. Map 3.5 shows where they are. In Morocco, the area from Fčs moving northwest toward the sea is lagging, has a dense population, and is close to the hub of Rabat. In Algeria, the lagging wilaya of Mascara lies just inland from Oran; Médéa and Aďn Defla are situated just inland from the greater Algiers Policy Package 2. Connecting Poor Places to the Poles of Development 215 MAP 5.1 Trunk Road Connectivity Is Good Where the People Are: Travel Time in Hours to a City of More Than 500,000 Inhabitants, with Closed Borders IBRD 38185 NOVEMBER 2010 ESTIMATED TRAVEL TIME (HOURS) WITH STRICT BORDER 1 This map was produced by 2 the Map Design Unit of The World Bank. The boundaries, 3 colors, denominations and any other information shown 4 on this map do not imply, on the part of The World Bank 5 Group, any judgment on the 0 500 1000 1500 legal status of any territory, or any endorsement or 6 acceptance of such KILOMETERS boundaries. Source: Staff GIS analysis; sources detailed in appendix. TABLE 5.1 Most of MENA's Population Live Close to Agglomerations Travel time by hours to a city of 500,000 or more Country 1 2 3 4 5 6 % of population Algeria 18 13 12 12 13 32 Egypt, Arab Rep. 45 28 7 3 7 10 Iran, Islamic Rep. 37 12 15 16 9 11 Iraq 60 23 12 4 1 0 Jordan 64 28 4 2 1 1 Lebanon 67 27 5 1 0 0 Libya 41 21 13 3 5 16 Morocco 40 14 12 12 12 9 Syrian Arab Republic 52 20 9 4 5 10 Tunisia 28 19 21 12 13 7 West Bank and Gaza 85 13 0 0 2 0 Yemen, Rep. 17 14 37 17 8 7 Total 40 19 13 9 8 11 Sources: Staff calculations using GIS algorithms; Oak Ridge National Laboratory 2005; Nelson 2008. agglomeration; and Guelma abuts Annaba. In Tunisia, lagging, dense Béja and Kairouan are close to Tunis. In Libya, lagging, dense Al Khums province adjoins Tripoli itself. In Egypt, governorates with poverty rates greater than 20 percent (Beheira and Sharkeyya [UNDP 2008]) are close to the Cairo metropolis. In Iraq, Babil abuts Baghdad. In the Republic of Yemen, Amran lies just to the north of the cluster of leading 216 Poor Places, Thriving People governorates around Sana'a. In the Islamic Republic of Iran, the ostan of Qom lies just to the south of Tehran. There will be many other lagging areas with the potential to benefit from spillovers, which could be picked up by higher-detail country-level analysis. But for a lagging area to benefit from proximity to a leading area, it has to be well connected by infrastructure. In Indonesia, it has been shown that the spillover effect extends for 60 kilometers when roads are good and only 25 kilometers when they are poor (World Bank 2009a, 76). The growth of low-income areas around Jakarta in Indonesia, Seoul in the Republic of Korea, and Săo Paulo in Brazil followed the path of major transport axes (World Bank 2009a, 77). But the same could apply, on a smaller scale, to the agricultural track between a hamlet and the nearest weekly village market, or the dirt road between a village and the nearest market town. Investment in rural infrastructure, including rural roads, was found to be one of the main factors that determined how well a region of India fared in reducing poverty (Datt and Ravallion 1998). About a sixth of the Lao People's Democratic Republic's poverty reduc- tion from 1997­98 to 2002­03 was attributable to rural roads (Warr 2005). "Spillover connectivity" is different from long-distance connectivity. Spillover connectivity integrates a place into its nearest economic hub so that it can build economic links and share economies of agglomeration. The impact of spillover connectivity is, therefore, unambiguously posi- tive. Long-distance connectivity, in contrast, allows finished goods to be sold in distant markets. Improving long-distance connectivity in lagging areas can therefore be either positive or negative for lagging areas. On the positive side, it improves their access to markets. But on the negative side, it exposes their existing home markets to competition from more efficient producers in agglomerated areas. It is, therefore, very complex to predict whether the impact of better road connectivity will be positive or negative, but international evidence (Mu and Van de Walle 2007; OECD 2009) suggests that short-distance connectivity has a more posi- tive impact than long-distance connectivity. The question, then, is this: how can the Middle East and North Africa (MENA) connect its lagging areas to agglomeration hubs so that they benefit from spillover effects as much as possible? This chapter will deal with two kinds of connectivity: physical and virtual. It will begin with an examination of transport connectivity. But it will then go on to discuss how MENA can use information and communications technology (ICT) to bring some of the benefits of proximity to all locations, whether they be near to or distant from the economic hubs. Policy Package 2. Connecting Poor Places to the Poles of Development 217 Road Transport: An Extensive Long-Distance Network but an Underdeveloped Short-Distance Connectivity MENA's transport infrastructure is basically road-based. There is some rail connectivity, but railways represent only a twelfth of investment needs (Fay and Yepes 2003). Within-country sea and air transportation are of minor importance. Intercity (i.e., long-distance) connectivity in MENA is generally ade- quate, thanks to an extensive primary road network. The extent of the paved network naturally varies according to country size, with the Islamic Republic of Iran, Algeria, and Egypt having the largest paved networks (126,000 km, 76,000 km, and 75,000 km, respectively), while Lebanon, West Bank and Gaza, and Djibouti have the smallest networks (6,200 km, 5,200 km, and 1,200 km, respectively). The layout of the region's road network provides a good coverage of population centers (see table 5.1). Other measures of road network coverage, particularly the network density (per area of arable land and per population), also show that MENA generally performs well in comparison with other developing economies (figure 5.1). MENA's trunk road network fits the region's agglomerated pattern of population distribution. Map 5.1 shows how long it takes for people across the region to travel to a city of at least half a million people, based upon a model that reflects road speeds according to the class of road and the hilliness of the landscape. It is estimated that 72 percent of the MENA13 region's population (excepting Djibouti) live within three hours of such a city (table 5.2). However, the population coverage of road connectivity varies by country. In Lebanon, the West Bank and Gaza, Jordan, and Iraq, the share of the population living more than three hours from a city is 5 percent or less. At the other extreme, it is greater than 30 percent for the Republic of Yemen, Tunisia, Morocco, and the Islamic Republic of Iran. Algeria is an outlier, with more than half the population estimated to be living more than three hours away from a city of half a million inhabitants. As one would expect, lagging areas' populations are less likely to be well connected to cities. For the MENA13 economies (minus Djibouti), someone living in a lagging area is 77 percent more likely to be more than three hours from a major city than someone in a nonlagging area. Overall, however, 61 percent of MENA's lagging area populations do live within three hours of a city. Moreover, only 14 percent of MENA13's population live both in a lagging area and at least three hours away from 218 Poor Places, Thriving People FIGURE 5.1 Paved Roads Density: MENA Compares Well with Other Middle-Income Countries a. Road density 123 (km paved road/km2 arable land) 6 5 4 3 2 1 MENA average = 0.8 0 ria nd ti an n ya Ar stan Al p. am ile Tu . Th aq Tu d Sy me co Ro ep. nia Jo a ra p. y M ia p no an z rke k a ibou Re Re n A Re nis Ga Lib Ye oroc rd Isl Ch Ir ge bR ma ba ail yp aki ab ic ria n, Dj Le Eg P t, an n, tB Ira es W 9.5 b. Road density (km of paved road/1,000 people) 4.0 3.5 3.0 2.5 2.0 MENA average = 1.4 1.5 1.0 0.5 0.0 n, . p. me ep Ar lic Ro ya nia am ria p. Le q Dj n Th ti an Jo d nd n Pa co n e e za n, Al y M ia Ira no an k a rda ta b R il rke Re u Re nis yp pub Ye b R Ga Lib oc ra Ch Isl ge ibo ma kis ba ail Tu or Tu ic a t, nA Eg tB Ira ria es Sy W Sources: CIA 2006 (or latest available year); World Bank WDI 2003, 2006. Note: Paved road data for 2006 are available for most MENA13 economies. Where 2006 data are not avail- able, the latest available data are used, assuming that there have not been significant variations in paved road length between the latest year available and 2006. WDI data is for 2006 for all economies, except for Libya where it is 2003. a city of half a million people. This suggests that MENA economies should not necessarily assume that long-distance connectivity is a major component of the lagging areas' problem. But the good picture on long-distance connectivity masks a problem: there is a predisposition toward expanding higher-class roads in MENA at the expense of other key transport expenditures. The paved network represents around 60 percent of the total road network in MENA (figure 5.2). This is quite high for developing countries. The tendency Policy Package 2. Connecting Poor Places to the Poles of Development 219 TABLE 5.2 Percentage of Population Living More Than Three Hours' Road Travel from a City of at Least Half a Million Inhabitants % of nationwide % of lagging areas* % of nationwide population population population Lebanon 1 3 1 West Bank and Gaza 2 5 2 Jordan 4 5 2 Iraq 5 11 3 Syrian Arab Republic 18 12 4 Egypt, Arab Rep. 20 43 19 Libya 25 41 16 Yemen, Rep. 31 36 10 Tunisia 32 53 11 Morocco 34 53 12 Iran, Islamic Rep. 37 44 17 Algeria 57 63 19 Total 28 39 14 Source: Staff calculations from data sources in appendix. Note: *Lagging areas are defined as the two quintiles of administrative divisions with the lowest house- hold poverty or consumption indicators, according to data availability. FIGURE 5.2 MENA Region Concentrates on Paved Roads More Than in Comparator Economies 100 paved roads as a % of total road network 80 MENA average = 58% 60 40 20 0 ria Le an n Ira Ara q am p. Al . M tan co Th ya Tu d Ro outi nia ep le me ic Jo a p. Pa ia Dj y p no Ira an z rke Ye ubl Isl Re Re b R hi Re nis Ga oc Lib rd ge ma kis ba ail ra C ib or Tu b ic n, nd ka t, an yp nA n, tB Eg ria es W Sy Sources: CIA 2006 (or latest available year); World Bank WDI 2003, 2006. Note: Paved road data for 2006 are available for most MENA13 economies. Where 2006 data are not avail- able, the latest available data are used, assuming that there have not been significant variations in paved road length between the latest year available and 2006. WDI data is for 2006 for all economies, except for Libya where it is 2003. 220 Poor Places, Thriving People toward higher-class paved roads can be partially explained by the spatial concentration of the population and the resulting high traffic levels. However, it may also result from the centralization of road planning and management, the administrative complexity of managing many small secondary road projects for a centralized department, and the undervalu- ation of rural connectivity. Table 5.3 summarizes what country transport sector reviews by the European Union, the World Bank, and other institutions have identified as MENA's main problems in ensuring connectivity. The degradation of road surfaces, congestion, poor road safety, the low quality of passenger and freight services, and poor border procedures and facilities emerge as the main constraints. The following pages will discuss each of these in a little more detail. Degradation of Road Surfaces The drive toward the expansion of the network is taking place at the expense of road maintenance expenditures. The quality of the road TABLE 5.3 Road Connectivity Problems in MENA, as Highlighted in EuroMed and World Bank Sector Reviews Iran, Islamic Rep. Egypt, Arab Rep. Lebanon Morocco Djibouti Algeria Tunisia Jordan Libya Total Iraq Quality of the network Deteriorated surfaces of paved roads X X X X X 5 Traffic Intense traffic on main arteries X X X X X X X 7 Severe congestion in major cities X X X X X X X X 8 Safety Poor road safety X X X X X X X X 8 Transport services (passengers) Low quality of passenger transport services X X X X X 5 Transport services (freight) Low quality of freight transport services X X X X X X 6 Old trucking fleet X X X X X X 6 Logistics Underdeveloped logistics facilities X X X X X X X 7 Border-crossing Cumbersome customs procedures X X X X X X 6 Insufficient harmonization of procedures with neighbors X X X X X X X X 8 Source: Study Team from World Bank and EuroMed transport sector reviews, various years. Note: Empty cell = not reported as problem. Policy Package 2. Connecting Poor Places to the Poles of Development 221 network in MENA is, therefore, deteriorating. World Bank studies identified road deterioration as a major problem in Algeria, Egypt, the Islamic Republic of Iran, Iraq, Libya, and West Bank and Gaza. About 30 percent of the paved network in Egypt and Algeria is in bad condition, and up to 50 percent in West Bank and Gaza. No accurate measures are available for other economies in the region since they generally lack an effective asset management system--another indica- tion of the neglect of road maintenance. Meanwhile, few exact figures exist on the quality of the unpaved road network, although it is generally accepted among professionals that it is in worse shape than that of the paved network. In Algeria, for example, 70 percent of communal roads were only in poor or fair condition, compared with 46 percent of national roads (World Bank 2007a). Djibouti keeps about 70 percent of its paved roads in fair to very good condition, thanks to the use of high-quality construction materials and favorable geological and climatic conditions. A large share of Djibouti's unsurfaced roads, however, is deteriorating due to lack of maintenance (World Bank 2005a). Roads maintenance often suffers from underfunding. During the 15 years up to 2007, Algeria was allocating only 0.2 percent of GDP to road maintenance, compared with an international norm of 0.5­1 percent (World Bank 2007a). In 2004, Egypt was allocating only 0.05 percent of GDP, despite having an adequate roads network in terms of coverage (World Bank 2007b). High axle loads aggravate the lack of maintenance. In Jordan, annual spending on road maintenance averaged only US$12.8 million per year during 1998­2003, whereas estimated needs were dou- ble that figure (World Bank 2004). Libya spends most of its 0.8 percent of GDP roads budget on new construction and very little on mainte- nance. The resulting maintenance backlogs across MENA's roads sector entail severe economic losses, both for the government and for the road user. The South African government has estimated that repair costs are multi- plied by a factor of 6 after three years of neglect and by 18 after five years of neglect (World Bank 2007a). And it is reported that a $1 reduction in road maintenance expenditure can increase the cost of vehicle operation by $2 to $3 (World Bank 1989). Rural Connectivity Deficits Another contrast to the efforts to expand the trunk (long-distance) net- work is the comparative neglect of rural accessibility, despite some recent improvements. The rural index indicator for MENA, defined as the percentage of rural population with access to an all-season road, was esti- mated at about 60 percent in 2004 (figure 5.3), which is one of the low- est values in the world (outperforming only Sub-Saharan Africa and 222 Poor Places, Thriving People FIGURE 5.3 Rural Access: MENA versus Comparator Regions a. Rural access index, by world region, 2004 100 % of rural population with access to all-season road 80 60 40 20 0 ic sia an a * ** ric sia cif be ca lA Af Pa hA fri rib tra th nA he Ca ut en or So dt ra dN he dC an ha dt an an Sa ia an As pe st b- ica Ea Su ro st Eu er le Ea Am idd M tin La b. Rural access index, by country, 2004 100 % of rural population with access to 80 MENA average all-season road 60 40 20 0 nia Dj n Jo i an ab a p. T e Pa p. Al n ria ep q Tu ic Th cco n, d p.* am y M ia ut no t, Liby il ta Ira me an Isl urke l Re Re nis ub rd Ch ibo ge ma kis Re o ba Ye ail or ic Ro Le bR Ar ra yp nA n, Eg Ira ria Sy Source: Roberts, Shyan, and Cordula 2006. Note: Top: * Data for 2003­04, ** data for 1999­2004. Bottom: * Data for 1999. South Asia). The Republic of Yemen, Morocco, Tunisia, and the Syrian Arab Republic had the lowest indicators in MENA at respectively about 20, 38, 40, and 50 percent in 2004. However, significant improvements have been recently noted, partic- ularly in the Republic of Yemen and Morocco. Morocco launched its National Rural Roads Program (NRRP) in 1994. During the following Policy Package 2. Connecting Poor Places to the Poles of Development 223 10 years, the government constructed or improved more than 10,000 kilometers of rural roads. Many positive impacts were observed: increased frequency of public transport, a shift toward higher-value crops, dou- bled enrollment in private education, increased teacher retention rates, and more use of health facilities. Nonetheless, only 43 percent of the rural population was estimated to have "easy access" to roads by 2004. The government therefore launched its NRRP-2 for 2005­10. An important change is that the old plan set targets in terms of kilometers of road constructed whereas the NRRP-2 focused on maximizing the number of people served (Levy 2004). Despite the high headline costs of large-scale rural roads programs, they are actually quite cost-effective on a per capita basis. Even in the Republic of Yemen's mountainous landscape, the Rural Access Program is estimated to bring surfaced road connectivity at a cost of around $100 per rural inhabitant (World Bank 2009e). The Moroccan National Rural Roads Program (NRRP2) set a threshold criterion for connectiv- ity of $490 per beneficiary. The first phase of the NRRP2 is establishing rural road connections at a cost of around $350 per person (2.1 million beneficiaries for DH 5.6 billion). However, unit costs will rise steeply as rural connectivity rates improve and only the most isolated communities are left to connect. Urban Congestion: Too Many Private Cars MENA's urban road systems are seizing up. A sharp increase in motor- ization rates and the lack of public space for further infrastructure devel- opment are suffocating MENA's agglomerations. Excessive congestion is therefore constraining connectivity in most MENA cities, particularly Tehran, Cairo, Alexandria, Casablanca, and Beirut, with the associated increase in travel time, energy inefficiency, and poor air quality. Urban congestion has an economic and a human cost. In 1989, the average speed of Cairo traffic during the rush hour was 8.1 kilometers per hour in peak periods. Since then the vehicle fleet may have quadrupled (El Araby 2002). On the urban expressways, which were constructed in order to decongest the city, average speeds were found in 2000 to be as low as 10 kilometers per hour and traffic queues often extended 2 to 3 kilometers (World Bank 2000). Egypt's National Research Center found that the average noise level in Cairo from 7 a.m. until 10 p.m. was 85 deci- bels, equivalent to a freight train passing 5 meters away (Slackman 2008)! One estimate (Shokri 2005) is that Tehranis spend a total of 4.5 million hours per day waiting in traffic and that the space required to park the city's cars increases by 5 hectares per day. In Algiers, average journey times are reported to be as high as 80 minutes each day (World Bank 224 Poor Places, Thriving People 2007a). Traffic speeds in Beirut are as low as 12 kilometers per hour for much of the day (World Bank 2006a). Particulate matter concentrations in MENA2 cities are around 85 grams per cubic meter, compared with 60 grams for the world as a whole, and cost MENA about 0.9 percent of gross national income in terms of impaired health (World Bank 2009b), compared with 0.5 percent worldwide. Moreover, motorization rates are set to increase with economic growth. The problem is that motorization in MENA leans toward private cars at the expense of public transportation. Higher-income economies in MENA have a significantly higher share of cars in their fleets, indicating a consumer preference for personal car ownership, when income levels per- mit, to meet transportation needs and to compensate for the lack of pub- lic transportation and the poor quality of available services. The result is that buses' share of the total vehicle fleet is more than twice as high in com- parator economies with similar levels of income (Chile, Turkey, Pakistan, and Romania) as in the MENA region: 5 percent versus 2 percent. In Tehran and Isfahan, around 30 percent of all urban trips are made in pri- vate cars, which is unusually high for a country at this level of development (World Bank 2005b); and in Lebanon 68 percent of motorized trips are made by private car (World Bank 2006a). In 2006 the rate of private car ownership in Damascus was similar to that of Ireland (World Bank 2008). This model of urban transportation development in MENA, focused on private cars and low-quality bus services, is becoming unsustainable. The building blocks for an efficient (bus-based) urban public trans- portation system are well known and tested: · The establishment of a public transport regulatory authority · Integrated urban transport planning · Route planning and procurement of services by the public authority · Provision of services by several private operators · Performance monitoring and enforcement by the public authority · Allocation of some road space dedicated for public transport As this sequence suggests, a city region's ability to mobilize public trans- portation for spillover connectivity will depend critically upon the estab- lishment of an empowered regulatory authority. Difficulties faced by women in using public transportation hinder their participation in the workforce. As low female labor force participa- tion is part of the lagging area problem in MENA, it will be important to consider female-only transport services or other culturally acceptable transport modes, particularly in conservative areas. Policy Package 2. Connecting Poor Places to the Poles of Development 225 Some MENA cities lack adequate bypass roads for long-distance freight traffic, which worsens urban congestion. This situation occurs in Algiers, Beirut, and Amman. Despite the extensive spread of MENA's trunk road network, some countries also have a few critical stretches of intercity road where congestion is particularly high due to inadequate road quality and capacity. These roads include segments of Algeria's east-west northern axis and the International Corridor from the port of Djibouti to the Ethiopian border. Road Safety MENA's road safety record is bad and worsening. Road safety appeared as a major issue in all eight of the MENA13 economies for which the World Bank has conducted recent transport sector reviews. The region's average rate of road crash fatalities is currently one of the highest in the world, be it measured per capita or per registered vehicles (10 people killed per 100,000 people and 25 people killed per 100,000 vehicles per year; see figure 5.4). The situation is particularly worrisome in Morocco, Jordan, and the Islamic Republic of Iran (figure 5.5). In the Islamic Republic of Iran, the fatality rate of 50 deaths per 10,000 vehicles per year is 20 times the norm for developed countries, and road accidents have been estimated to cost 3 percent of GDP (Ayati and Behnood 2007). The causes of poor road safety in the Islamic Republic of Iran include poor driver training, slack enforcement of traffic rules (World Bank 2005b), dangerous intersections, lack of warning signs, worn pave- ment markings, bad quality road surfaces, and inadequate maintenance. FIGURE 5.4 MENA Is a Region of Dangerous Roads 30 25 persons killed per 10,000 registered vehicles 20 15 10 5 0 Europe and Latin America East Asia and Middle East and Central Asia and the Caribbean Pacific North Africa Source: International Road Federation 2008. 226 Poor Places, Thriving People FIGURE 5.5 Average Rates of Road Crash Fatalities per Capita and Registered Vehicles in MENA Region and in Comparator Countries Road fatalities in MENA persons killed per 10,000 vehicles 60 50 40 30 20 10 0 p. lic co n ia ria an za nia ile y rke ta Re nis Ga ub oc rd Ch ge ma kis Tu Jo or ep Tu ic Al nd Pa Ro M am bR ka Isl ra an nA n, tB Ira ria es Sy W Sources: For Iran: World Bank. "Technical Assistance Project for Road Safety. Concept Note." For all other countries, team calculations based on International Road Federation 2008. In 2002, about 5 percent of health loss in MENA, measured in terms of Disability-Adjusted Life Years (DALYs), was due to road crashes. This is higher than any other region and makes road crashes the third leading cause for health loss in MENA. In 2030, road crashes are expected to become the leading cause for health loss in MENA, accounting for about 8 percent of total health loss in the region. Low Efficiency of Road Use The return on MENA's new road investment is currently being under- mined by inappropriate regulation of the commercial operators who use them. The third key challenge facing MENA's roads sector is, therefore, to create a favorable business environment for transportation. While it is easy to criticize road sector regulation, it has proved extremely difficult for MENA economies to achieve a satisfactory mid- dle way between overregulation, which raises prices and restricts supply, and underregulation, which leads to poor service quality. Egypt, the Islamic Republic of Iran, Iraq, and Jordan are examples of overregulation keeping costs up. Algeria, Lebanon, and Tunisia are examples of the combination of low costs and low service quality in the transport sector that results from underregulation (box 5.1). Policy Package 2. Connecting Poor Places to the Poles of Development 227 BOX 5.1 Seeking the Balance between Underregulation and Overregulation of Freight and Passenger Transportation · In the Arab Republic of Egypt, the privatization of the roads sector has been under way since the early 1990s; however, the largest companies remain state-owned, are inef- ficient, and offer low service quality (Euro Med Transport Project 2004b). Regula- tions and duties on the import of trucks (especially refrigerated trucks) and spare parts help keep the operating costs for trucks 30­50 percent higher than in Lebanon and Jordan (World Bank 2006c). Road sector institutional responsibilities are frag- mented between three ministries (World Bank 2007i). · In the Islamic Republic of Iran, the formal passenger transport sector is largely under state control, and the low quality of services provided exacerbates reliance on private cars and informal collective taxis (World Bank 2005b). · In postwar Iraq, the trucking sector was essentially monopolized by government- controlled firms. The emphasis was placed on volumes rather than service quality. Truck maintenance was insufficient and the vehicle fleet was aging (United Nations and World Bank 2003). · In Jordan, the "turn system" is the main bottleneck affecting the trucking industry. Private operators cannot approach and negotiate with customers directly, but must go through UCOLT, a state-owned company, which allocates jobs among contrac- tors. This rigid system leads to an oversupply of trucks and long waiting periods. Operators do not have an incentive to offer better services and their returns are low, while prices are high (Euro Med Transport Project 2004c). · In Algeria, freight transport was completely liberalized in 1988, and the private sec- tor is now dominant for both passenger (90 percent) and freight (95 percent) trans- port. However, a lack of adequate regulation of private operators initially led to frag- mentation and a poor quality of service. More recently, entry regulations have been enforced, incentives for operators to group have been introduced, and the situation has improved somewhat (Euro Med Transport Project 2004a). · Lebanon's complete absence of entry regulations in the passenger transport sector since 1996 has led to the emergence of myriad small operators, 70 percent of whom use pri- vate cars, which contribute to congestion concerns. Oversupply has led to price-cutting at the expense of vehicle maintenance, service quality, and safety (World Bank 2006a). · Morocco's freight transport sector is another instance of atomization, with 90 percent of operators owning just one or two trucks, a predominance of informal carriers (70­75 percent), a lack of specialized vehicles such as refrigerated trucks, and an anti- quated vehicle fleet (on average 13 years) (World Bank 2006b). · Tunisia's freight transport has been completely liberalized, ensuring low tariffs (Euro Med Transport Project 2004d). 228 Poor Places, Thriving People Regional Trade Facilitation Matters Crucially for Lagging Areas Land borders can create artificial lagging areas. Put simply, if it is diffi- cult for freight and people to cross a land border, the areas on either side of it are cut off from part of their economic hinterland. Even in the highly integrated European Union (EU), border regions are twice as likely to be lagging regions than are nonborder regions (Van Well 2006, 80). In MENA, there are several lagging border areas that could benefit from their proximity to major cities in neighboring countries if travel across borders were as easy as within borders. There is also a tendency for governments to underinvest in transborder road and rail links, maybe because part of the benefits of such investment will go to another coun- try or simply because of a lack of coordination between each country's transport planners (Puga 2008). Cross-border trade within MENA is low. The Maghreb (Algeria, Libya, Mauritania, Morocco, and Tunisia) has one of the lowest figures for intraregional trade integration of any region in the world: regional trade represents only 1.3 percent of total Maghreb trade with the world. The Mashreq region is somewhat more integrated. The intraregional share of their total trade, at about 17 percent, is much less than the 65 percent plus of the two largest trading blocks (North American Free Trade Agreement [NAFTA] and EU), rather less than that of Association of Southeast Asian Nations (ASEAN) (26 percent) and rather more than the Southern Cone Common Market (Mercado Común del Sur) (Mercosur) (16 percent) and the Arab League countries as a whole (11 percent) (World Bank 2009d). A comparison of map 5.2 with map 5.1 shows how zones of economic proximity can spill across national boundaries to benefit lagging areas in some countries. The far northeast of Morocco benefits from proximity to Oran in Algeria. The lagging far northeast of Algeria benefits from proximity to Tunis. The poor northern tip of Lebanon benefits from proximity to Syria's Damascus-Aleppo axis. The disadvantaged region of northern Syria gains from proximity to Gaziantep, one of the most developed provinces of southeast Turkey, and Diyarbakir. Finally, the lagging ostan of Khuzistan in the southwest of the Islamic Republic of Iran gains from proximity to Basra in Iraq. The construction of transborder connections and the easing of border crossings are therefore not only trade promotion policies; they can also be spatial development strategies. A measure of the room for progress is the World Bank's Logistics Per- formance Index (LPI), which measures the ease of doing trade across bor- ders.3 Table 5.4 compares nine MENA13 economies' trade logistics with those from other parts of the world. Of the countries surveyed, Tunisia and Policy Package 2. Connecting Poor Places to the Poles of Development 229 MAP 5.2 Smooth Border Crossings Would Give Lagging Border Regions a Better Chance of Benefiting from Spillover Connectivity: Travel Time in Hours to a City of More Than 500,000 Inhabitants, with Zero-Hour Border Crossings IBRD 38184 NOVEMBER 2010 ESTIMATED TRAVEL TIME (HOURS) WITH NO BORDER 1 This map was produced by 2 the Map Design Unit of The World Bank. The boundaries, 3 colors, denominations and any other information shown 4 on this map do not imply, on the part of The World Bank 5 Group, any judgment on the 0 500 1000 1500 legal status of any territory, or any endorsement or 6 acceptance of such KILOMETERS boundaries. Source: Staff GIS analysis; sources detailed in appendix. TABLE 5.4 MENA's 2006­08 Logistics Performance Index (LPI) Benchmarked Efficiency of customs and Quality of Domestic other border transport and IT Logistics transportation Overall LPI procedures infrastructures competence costs Djibouti 1.94 1.92 1.64 2.00 2.80 Algeria 2.06 1.83 1.60 1.92 3.17 Syrian Arab Republic 2.09 2.17 1.91 1.80 2.89 Yemen, Rep. 2.29 2.18 2.08 2.22 2.67 Sub-Saharan Africa average 2.35 2.21 2.11 2.33 2.98 Egypt, Arab Rep. 2.37 2.00 2.08 2.38 2.83 Lebanon 2.37 2.14 2.17 2.40 3.40 Morocco 2.38 2.20 2.33 2.13 2.38 Lower-middle-income average 2.46 2.26 2.30 2.38 3.04 Iran, Islamic Rep. 2.51 2.44 2.50 2.69 2.93 Latin America and the Caribbean average 2.57 2.38 2.38 2.52 2.97 Tunisia 2.76 2.83 2.83 2.43 3.20 Upper-middle-income average 2.76 2.60 2.55 2.72 2.90 Jordan 2.89 2.62 2.62 3.00 2.92 Source: http://info.worldbank.org/etools/tracksurvey, accessed March 2009. 230 Poor Places, Thriving People Jordan measured up against the average performance rating for the upper- middle-income country peer group. Morocco, Lebanon, and Egypt were very close to the Sub-Saharan African average. Djibouti, Algeria, Syria, and the Republic of Yemen achieved a significantly lower score than the average for Sub-Saharan Africa. Breaking the LPI down, only the Islamic Republic of Iran, Jordan, and Tunisia rated higher than Sub-Saharan Africa on the efficiency of customs and other border procedures. Some MENA economies' customs performance has improved in recent years. Algeria's customs procedures have been streamlined, and now run around the clock with the introduction of the SIGAD auto- mated system. As a transit country because of its geographic location, Jordan was an early adopter of the concept of intermodal transport and logistics, although the limited capacity of the main customs depot in Amman remains a bottleneck. Since the end of the Civil War, Lebanon has made major changes to its customs procedures, including the imple- mentation of the computerized NAJM customs system. Morocco's cus- toms reforms reduced clearance times from 5.5 to 2 days by the early 2000s, by means of the simplification of procedures and forms, the use of information technology, and measures to reduce corruption (World Bank 2002). Tunisia has introduced the electronic SINDA customs man- agement system. In 2002, Syria, Lebanon, and Jordan adopted a com- mon set of customs procedures to facilitate border crossings. However, despite the improvements in customs procedures, the shortage of modern logistics platforms and infrastructure remains a handicap to international trade regionwide. Whereas a survey of enter- prises in 2001 found tariffs to be the top obstacle to cross-border trade, by 2008 infrastructure services and red tape had overtaken tariffs (Hoekman and Zarrouk 2009). The infrastructural aspect of logistics emerges in sector reviews as a top priority issue in Morocco and Tunisia, where progress has already been made on the procedural side of cross-border trade. Priorities for Improving Spillover Connectivity First, MENA's broad pattern of public expenditure in transport could benefit from rebalancing. The earlier analysis suggests that MENA economies should consider reallocating resources away from expansion of the domes- tic trunk road network toward (a) maintenance of the existing network, (b) improving rural access, (c) trans-border networks, and (d) improving traffic management. Second, urban and peri-urban mass transit systems are critical if people from lagging areas are to benefit from agglomerations' demand for Policy Package 2. Connecting Poor Places to the Poles of Development 231 labor. In Morocco, for example, the recent investments in peri-urban rail links have enabled workers to commute into the prime cities of Rabat and Casablanca from Salé, Kenitra, and Mohammedia (Er-Rouch 2009). As cities grow, so do commuting distances; in Egypt, the average time spent commuting increased by about a third between 1988 and 1998 (Assaad and Arntz 2005). While rarely justified on a purely finan- cial basis and therefore requiring substantial public contributions, mass transit investments support smooth urban expansion (chapter 4) and yield significant economic benefits. Many MENA cities will need new approaches to urban mass transit. At present, it is common for mass transit to consist largely of low-quality pub- licly owned bus services, and unaffordable public investment in light rail is seen as the only alternative. However, there are other potential strategies: · First, many MENA economies have informal urban shared taxi sec- tors, providing low-quality services. However, they often lack private bus services of high enough quality to compete with private cars for the middle-class market. Lifting barriers to private investment in urban transport can encourage the development of formal, high- quality private bus services. · Second, bus rapid transport (BRT) is proving itself a lower-cost alter- native to light rail for high-volume mass transit. At its peak, Bogotá's TransMillenio system can move 25,000 people per hour across a single point in space. It is estimated that dedicated bus lanes can handle 40 to 60 vehicles per hour (Levinson, Zimmerman, Clinger, and Rutherford 2002), competing with private cars for speed, comfort, and cost. The BRT model also permits the relatively simple participation of multiple operators on the network, some of which may be private. · Third, improving transborder connectivity will be particularly impor- tant in ensuring that lagging areas harness spillover connectivity ben- efits. A strategy to improve cross-border connectivity could contain the following components: ­ An analysis of the main obstacles to cross-border connectivity: procedures, regulations, the lack of regional logistics services, and missing infrastructure links (roads, railways, multimodal facilities, border infrastructure, and logistical platforms). The Transport and Trade Facilitation Audit (TTFA) is an established methodology for identifying such bottlenecks. ­ Concrete international cooperation in areas such as data, training, and information systems ­ Spatial planning to capture economies of scale and match invest- ments to transport flows 232 Poor Places, Thriving People ­ Bilateral and regional agreements on road and rail transport ­ A transit regime for long-distance trade (e.g., the TIR [Transports Internationaux Routiers] system) · Fourth, making MENA's road infrastructure more productive will depend on finding the appropriate mix of liberalization and regulation in the passenger and freight transportation sectors. According to the country experiences in box 5.1, key elements of a good regulatory and institutional approach include the following: ­ Creation of integrated and empowered municipal authorities for urban transport, to ensure a coordinated approach to urban trans- port planning (defining the roles of public and private operators) and to public transportation across administrative jurisdictions. ­ Open access to private operators and withdrawal of the public sec- tor from freight transportation. Freight transportation is a sector par excellence in which the private sector's dynamism and respon- siveness are essential. ­ Gradual and strict regulation of vehicles' safety standards and axle weights. Progressive implementation will give operators time to adapt and to avoid the disruption of services for low-income passengers. ­ Liberalization of importation of commercial vehicles and parts, including specialized vehicles such as refrigerated trucks. ­ Accepting that the fragmentation of the road freight sector is nor- mal. Even in developed countries, around 80 percent of road haulers operate one or two vehicles. If larger road freight compa- nies are not emerging in MENA, this may be because of an unfa- vorable general business environment for large formal companies or because of a lack of large, consolidated shipments in a frag- mented goods market. Atomization of the transportation sector does not mean that it needs additional regulation. (See box 5.2.) Electronic Proximity: New Tool for the Integration of Lagging Areas Isolation from information and relationships is a key part of the lagging area problem. It puts the populations of lagging areas at a disadvantage when it comes to interacting with bureaucracies and markets. Information and communication technology4 (ICT), therefore, give policy makers a whole new set of options for attacking spatial dispari- ties. They make it possible to create virtual proximity where physical Policy Package 2. Connecting Poor Places to the Poles of Development 233 BOX 5.2 Enhancing Spillover Connectivity in MENA: Four Ideas for Policy Makers · Identify lagging areas that are dense and close enough to agglomerations to benefit from spillover connectivity, including transborder connectivity. · Identify the constraints on spillover connectivity in these areas, paying specific atten- tion to infrastructure maintenance, short-distance urban and rural connections, and the efficiency of freight and passenger transport operators. · Work out which institutional arrangements are needed to underpin the integrated planning of connections within a city and its hinterland. · Within a territorial development framework, prepare a plan integrating the institu- tional, regulatory, and investment actions required to connect a city and its hinterland. Source: Study Team. proximity is impossible. Of course, electronic connectivity can never compensate lagging areas for high transport costs to agglomerations, but it can help mitigate the informational and relational penalties of dis- tance (table 5.5). ICT Can Help Lagging Areas Mobile phones have a special potential in rural areas in developing coun- tries; the next billion mobile subscribers will consist mainly of the rural poor (World Bank 2009c). Three-quarters of all mobile handsets are now used in developing countries (Economist 2008). Mobile phones are now the world's largest distribution platform, offering applications such as m-health, m-education, and m-banking. However, ICT infrastructure can be weak in lagging areas. In many rural parts of the world, such problems can be very basic: aging copper- based telephone lines, or local monopolies of telecommunication corpo- rations over exchanges and "last mile" infrastructure, with little incentive to open up to alternative providers. For every 10-percentage-point increase in broadband penetration in developing countries, there is a 1.4-percentage-point growth benefit, but the penetration of broadband connectivity into developing countries' rural areas has been slow. Less than 5 percent of low-income countries' populations were connected to broadband networks in 2007, and that was mostly in urban centers (World Bank 2009h). For digital subscriber 234 Poor Places, Thriving People TABLE 5.5 How ICT Can Benefit Lagging Areas Individuals Businesses Government Making transactions with businesses Making transactions with consumers Delivering time savings and reduc- and government quicker and or government quicker and tion of costs in transacting with cheaper, e.g., by saving a trip to cheaper citizens or businesses check school exam scores Improving access to public services Improving access to new markets; Enhancing community development through e-government, e-health, connecting to economic opportu- and e-education, if offered nities and integration at the national and international levels Increasing employment Stimulating development and facili- Emergency or disaster support opportunities tating business services needed by SMEs and home businesses Improving access to financial services, Improving the investment climate in Reducing tax collection costs such as banking, payments or lagging areas remittances and money transfers Relieving the need to migrate to Increasing productivity New sources of income urban centers Access to knowledge, skills, and Access to knowledge, skills, and net- Improved ability to deliver public networks to increase opportunities works to increase opportunities services Improving access to information Developing the local private sector services and lessening migration to the cities Source: Study Team. Note: SMEs = small and medium enterprises. line (DSL) technologies, which are the most popular platform for broad- band services worldwide, there is a particular problem of attenuation of signal with distance. This means that quality is lower and costs are higher in low-density areas. In the absence of a universal service obligation (USO) relating to broadband Internet, its provision is subject to market forces, which do not favor areas of dispersed population because of the high investment and low demand (Warren 2007). There is, therefore, a danger that nonusers of the Internet and mobile telephony in lagging areas will become disenfranchised. As ICT becomes progressively regarded as the default communication medium, those without access to it become gradually marginalized (Warren 2007). Table 5.5 summarizes how ICTs can benefit lagging areas. Examples abound of successful pilots demonstrating how ICTs have overcome problems of geographical isolation that were once thought to be insur- mountable. · The high cost of setting up bank branches used to be considered a barrier to providing formal banking services to rural and semirural Policy Package 2. Connecting Poor Places to the Poles of Development 235 areas of Brazil. At the end of 2000, 71 percent of Brazil's 5,636 munic- ipalities had bank services. But three years later, all of them did. The difference was the explosion of banking correspondents: small retail- ers such as kiosks or pharmacies equipped with a point-of-sale (POS) device, a keypad or maybe a personal computer, for customers to enter their personal identity numbers. The POS device might be as simple as a mobile phone. By the end of 2005, 90,000 banking correspon- dents were in action. They had become the core of retail financial services in the lagging areas of the north and northeast, where 51 and 42 percent of municipalities, respectively, were now served exclusively by correspondents (CGAP 2006a; CGAP 2006b). · There are many examples worldwide of SMS (Short Message Service) alert systems giving farmers, fisherfolk, and traders real-time infor- mation on market prices in different locations. · The Egyptian Education Initiative (EEI) has established information technology clubs nationwide, with a focus on rural areas. By pooling hardware, software, and connections, these clubs help reduce usage costs to about US$0.20 an hour, overcoming the financial barrier to Internet access in low-income areas (MCIT 2008). · Egypt's Ministry of Communications and Information Technology has also deployed public-private partnerships to set up information tech- nology (IT) clubs across the country. The majority of the 1,427 clubs are in low-income and rural areas, with the highest per capita coverage in lagging Upper Egypt and the frontier governorates. Local small enterprises are using the IT clubs to leverage scarce computing power. · Tunisia has established cyberparks for remote labor in 16 provincial towns. They host 395 enterprises (MCT 2009). However, informa- tion is not available on their impact in lagging areas or on their cost- effectiveness in terms of jobs created. · National Research and Education Networks (NRENs) can pool information resources (electronic library access) among universities to reduce disparities in resources. An example is the Egyptian Uni- versities Network (EUN: http://www.eun.eg), which was an early example of a regional NREN and which now links with NRENs in Europe and North America. · An innovative new project, supported by the UN World Food Pro- gramme (WFP), uses mobile phones to deliver food vouchers to Iraqi refugees living in the country. This system enables them to buy a wider range of food, including fish and eggs, rather than being dependent on government-run food shops. Significantly, the WFP 236 Poor Places, Thriving People reports that mobile ownership among refugees is almost universal, even though other assets might be in short supply. · In a context of movement restrictions, mobile phones and the Internet are widely used in the West Bank and Gaza to facilitate remittances and to stay in touch with family members. The West Bank and Gaza leads in the rural adoption of new ICTs: more than a quarter of rural house- holds own a computer and more than a tenth have Internet access. · Tunisia's remote medicine investments allow outlying hospitals to take advantage of remote pathology, consultation, and conference services. MENA's Rural ICT: A Strong Legacy of Fixed Lines, but Lagging Access to Mobile Telephony and the Internet Comparing MENA's rural areas with those of other regions, the good news is that MENA has a comparatively high rate of access to electricity (figure 5.6). MENA's rural areas, however, have much poorer access to FIGURE 5.6 The Spatial Coverage of ICT in Developing Regions' Rural Areas 120 99 100 % of rural localities with access 83 83 80 59 57 60 48 51 38 38 41 40 22 20 13 14 12 7 3 3 1.4 0 c sia n ca sia ca ifi ea ri ri lA hA ac ibb Af Af eP tra ut rth an ar So th en eC r No ha dC d th an Sa d an an nd b- sia Su aa pe st A Ea ro ric st Eu le Ea me idd A M in t La electricity telephone Internet Source: ITU 2008. Note: Based on data between 2000 and 2006 using simple averages. Policy Package 2. Connecting Poor Places to the Poles of Development 237 telephone and Internet services than their counterparts in South Asia and East Asia, as well as Europe and Central Asia. Mobile telephony's com- parative advantage over fixed telephony is strongest in low-density areas, where the cost per household of installing and maintaining cable con- nections is highest. However, the shift in rural areas from fixed-line to mobile telephony has been less pronounced in MENA than in other parts of the world. In all the other regions, the rate of rural mobile usage has overtaken that of fixed lines. In MENA, however, 37 percent of households have fixed lines and 26 percent have mobile lines. A probable reason is that MENA made significant early investments in extending fixed-line telephony to rural areas, achieving high rural usage rates com- pared with East Asia and the Pacific (10 percent), Latin America and the Caribbean (13 percent), South Asia (11 percent), and Sub-Saharan Africa (1 percent). Only Europe and Central Asia (46 percent) have higher rates of rural household fixed-line connections than MENA (ITU 2008). Expansion of ICT Use in Lagging Areas When designing strategies to extend ICTs to lagging areas, it is useful to break the task down into two components: (a) the market efficiency gap and (b) the access gap (Navas-Sabater and Dymond 2002). The market efficiency gap is the difference between the current level of serv- ice penetration and the level achievable in a liberalized and stable regu- latory environment. The access gap relates to the gap between leading and lagging areas that would remain under efficient market conditions because the lagging population cannot afford the market prices at which the service is offered. Sector reform will help close the market efficiency gap by increasing competition, reducing tariffs, and increasing supply. As can be seen in Table 5.6, fixed-line liberalization still has a way to go in MENA. The present and planned liberalization of the fixed-line and data mar- kets in some MENA economies should enhance infrastructure-based competition in these countries. Allowing for international connectivity competition in most MENA economies will also lower international bandwidth costs, which will allow the Internet and data providers to start offering more advanced services at cost-based rates, and hence much lower rates, across the country (Arab Advisors Group 2008). Figure 5.7 compares the spread of communications technology in MENA economies with global trends around 2007­08. In 2007, Jordan was ahead of the global trend for cellular subscriptions, with Egypt, the Islamic Republic of Iran, Libya, Lebanon, and Syria lagging behind. In 2008, most MENA economies' monthly cellular tariffs were below the global trend, but not Morocco's and Lebanon's. Broadband connections have been slow to grow in Libya and Syria. 238 Poor Places, Thriving People TABLE 5.6 Fixed-Line and Internet Market Status (According to Operational Licenses) in MENA Cellular Country Fixed lines Internet operators Algeria Duopoly Competitive 3 Egypt, Arab Rep. Monopoly Competitive 3 Iraq Monopoly Competitive 5 Jordan Competitive Competitive 4 Lebanon Monopoly Competitive 3 Morocco Competitive Competitive 3 West Bank and Gaza Monopoly Competitive 2 Syrian Arab Republic Monopoly Competitive 2 Tunisia Monopoly Competitive 3 Yemen, Rep. Monopoly Duopoly 4 Sources: Arab Advisors Group 2008; Internet search for cellular operators. To close the access gap, governments need to use a mix of approaches, including regulatory incentives to attract investors, smart subsidies, and a universal access fund (Navas-Sabater and Dymond 2002). Telecommunications sector liberalization will increase total serv- ice supply and reduce rates. However, policy makers should be aware that increased competition and reduced tariffs can discourage operators from investing in low-density areas where the cost of infrastructure per subscriber is higher. Governments may wish to set up specific incentives for ICT operators to expand into poorer areas. The economic rationale for such measures would be that information-related services create positive externalities of social integration and efficient markets. It is, therefore, important to identify incentives that encourage operators to meet the government's spatial objectives without distorting or stunting the overall growth of the sector. The traditional solution to spatial disparities in ICT access has been simple: instructing state-owned fixed-line monopolies to expand into rural areas, as in many MENA economies. This is a high-cost solution, as publicly owned fixed-line technologies are the least efficient technol- ogy in low-density areas. Another disadvantage of the fixed-line monop- oly approach to expanding lagging area coverage is that operators can use their social mandate as a cover for their suboptimal financial perform- ance; once a state-owned monopoly has such a noncommercial objective, it is very difficult for the government to distinguish the costs of the social provision from the costs of inefficiency. Policy Package 2. Connecting Poor Places to the Poles of Development 239 FIGURE 5.7 Telecommunications: Who Is Ahead and Who Is Behind the International Trend? a. Mobile cellular subscriptions, 2007 160 mobile cellular subscriptions per 100 people 140 120 100 80 Jordan Algeria Libya Morocco Tunisia 60 Egypt, Arab Rep. 40 Iran, Islamic Rep. Lebanon 20 Syrian Arab Republic 0 2.5 3.0 3.5 4.0 4.5 5.0 log10 of GNI per capita, PPP (current international $) b. Mobile cellular tariffs, 2008 40 35 mobile cellular prepaid tariff 30 25 ($ per month) Morocco Lebanon 20 15 Syrian Arab Republic 10 Tunisia Algeria 5 Yemen, Rep. Jordan Iran, Islamic Rep. Egypt, Arab Rep. 0 2.5 3.0 3.5 4.0 4.5 5.0 log10 of GNI per capita, PPP (current international $) c. Broadband connections in Libya and Syrian Arab Republic, 2006 100.00 fixed broadband connection per 10.00 Lebanon 100 people (log scale) 1.00 Morocco Jordan Tunisia Algeria Egypt, Arab Rep. Libya 0.10 Syrian Arab Republic 0.01 0.00 3 4 4 5 5 log10 of GNI per capita, PPP (current international $) Source: International Telecommunications (ITU) database. 240 Poor Places, Thriving People BOX 5.3 Bringing ICT Connectivity to Lagging Areas: Two Ideas for Policy Makers · Shifting from a telecoms policy for lagging areas based on monopolies' and oligopolies' "social mandates" to an approach founded on the market-based allocation of spatial incentives to competing private operators. · Implementing public-private partnerships for ICT-enabled services to lagging areas--for example, financial services, Internet clubs, distance education, and market information. Source: Study Team. A second solution has been for governments to levy fees on telecom operators in order to fund social ICT programs (see box 5.3). Morocco, for example, has adopted levies to finance the GENIE program (computers in schools), and levies on operators have subsidized the Smart Village Technopark, a public-private partnership in the outskirts of Cairo. But this approach carries some risks. First, it is difficult for special social programs to reach a broad range of disadvantaged people. Second, the fact that oper- ators are transferring a part of their oligopoly rents to government may strengthen their bargaining power in opposing the entry of new operators. A more attractive example of the incentives approach was to be found in Chile, an early telecoms reformer. In 1994, the government estab- lished a Telecommunications Development Fund, financed by the national budget, to catalyze private investment in pay phone service in low-income areas. Between 1995 and 2000, the fund supported the provision of pay phone service to more than 6,000 rural localities with about 2.2 million inhabitants, thereby reducing the proportion of Chile's population living in places without access to basic voice communication from 15 percent in 1994 to 1 percent in 2002. In addition, some 25,000 individual rural telephone lines were provided. The subsidies cost the government less than 0.3 percent of total telecommunications sector rev- enue during the funding period, and fund administration costs about 3 percent of the monies granted. The fund's success and efficiency was largely due to extensive reliance on market forces to determine and allo- cate subsidies, minimal regulatory intervention, simple and relatively expeditious processing, and effective government leadership. Competi- tion among operators for the rural market and subsidies led to substan- tial reductions in cost (Wellenius 2002). Policy Package 2. Connecting Poor Places to the Poles of Development 241 Endnotes 1. Giza, Gharbeyya, Suez, Ismailiya, Kalyoubiya, Sharkiya, Matrouh, and Beheira. 2. Including the Gulf Cooperation Council (GCC). 3. The Logistics Performance Index (LPI) reflects the overall perception of a country's logistics based on over 1,000 responses to a survey of logistics perform- ance evaluated in seven key subcategories. These categories include efficiency of customs and other border procedures, quality of transport and IT infrastructures, international and domestic transportation costs, ease of shipments and logistics competence, and tracking ability and timeliness of shipments. The value of the index ranges from 1 to 5, with a higher score representing a better performance. 4. For present purposes, ICT comprises the following subsectors: telecommu- nications services (fixed and mobile), telecom hardware, computer hardware, software, networking, ICT-enabled services (ITES), IT applications, and advanced media. The broadcasting and postal sectors, which are also considered part of the ICT sector by some countries, are not treated. References Arab Advisors Group. 2008. ADSL Rates in the Arab World: A Regional Comparison. Amman: Arab Advisors Group. Assaad, R., and M. Arntz. 2005. "Constrained Geographical Mobility and Gendered Labor Market Outcomes under Structural Adjustment: Evidence from Egypt." World Development 33 (3): 431­54. Ayati, E., and H. R. Behnood. 2007. "The Burden of Fatalities Resulting from Road Accidents: An Epidemiological Study of Iran." Paper presented at the 14th International Conference, "Road Safety on Four Continents," Bangkok. Belarbi, Y., and A. Zouache. 2009. Regional Employment Growth and Spatial Dependence in Algeria, 1998­2005. Cairo: Economic Research Forum. CGAP (Consultative Group to Assist the Poor). 2006a. FocusNote: Using Technology to Build Inclusive Financial Systems. Washington, DC: CGAP. ------. 2006b. Mobile Phones for Microfinance. Washington, DC: CGAP. CIA (Central Intelligence Agency). 2006-present. The World Factbook. Field Listing: Roadways. Washington, DC: CIA. Datt, G., and M. Ravallion. 1998. "Why Have Some Indian States Done Better than Others at Reducing Rural Poverty? Economica 65 (257): 17­38. Economist. 2008. "Halfway There: How to Promote the Spread of Mobile Technologies among the World's Poorest." May 29. 242 Poor Places, Thriving People El Araby, M. (2002). "Urban Growth and Environmental Degradation: The Case of Cairo, Egypt." Cities 19 (6): 389­400. Er-Rouch, O. 2009. "Train Commuting: A Necessary Pain." June 24, Agence Maghreb Arabe Presse. http://www.map.ma/eng/sections/ general/train_commuting__a_n/view. Euro Med Transport Project. 2004a. Algeria Diagnostic Study. ------. 2004b. Egypt Diagnostic Study. ------. 2004c. Jordan Diagnostic Study. ------. 2004d. Tunisia Diagnostic Study. Fay, M., and T. Yepes. 2003. Investing in Infrastructure: What Is Needed from 2000 to 2010. Washington, DC: World Bank. Hoekman, B., and J. Zarrouk. 2009. "An Assessment of the Changes in Cross-Border Trade Costs in the Pan-Arab Free Trade Area, 2001­2008." MNA Knowledge and Learning Fast Brief 30. World Bank, Washington, DC. IRF (International Road Federation). 2008. "World Road Statistics." IRF, Geneva. ITU (International Telecommunications Union). 2008. Measuring Infor- mation and Communication Technology Availabilty in Villages and Rural Areas. Geneva: ITU. Levinson, H., S. Zimmerman, J. Clinger, and S. Rutherford. 2002. "Bus Rapid Transit: An Overview." Journal of Public Transportation 5 (2): 1­29. Levy, H. 2004. "Rural Roads and Poverty Alleviation in Morocco." Paper presented at Scaling Up Poverty Reduction: A Global Learn- ing Process and Conference, Shanghai, May 25­27. World Bank, Washington, DC. McCulloch, N., and B. S. Sjahrir. 2008. Evaluating the Determinants of Subnational Growth in Decentralized Indonesia. Washington, DC: World Bank. MCIT (Ministry of Communications and Information Technology, Egypt). 2008. The Egyptian Education Initiative: Keys to Success. Cairo: MCIT. MCT (Ministry of Communication Technologies, Tunisia). 2009. "ICT in Tunisia: National Strategy, Main Achievements and Future Prospects." http://www.infocom.tn/index.php?id=199. Policy Package 2. Connecting Poor Places to the Poles of Development 243 Mu, R., and D. Van de Walle. 2007. Rural Roads and Local Market Devel- opment in Vietnam. Washington, DC: World Bank. Navas-Sabater, J., and A. Dymond. 2002. Telecommunications and Information Services for the Poor: Toward a Strategy for Universal Access. Washington, DC: World Bank. Nelson, A. 2008. Travel Time to Major Cities: A Global Map of Accessibility. Ispra, Italy: Global Environment Monitoring Unit--Joint Research Centre of the European Commission. http://gem.jrc.ec.europa.eu/. Oak Ridge National Laboratory. 2005. Landscan Global Population Data- base. http://www.ornl.gov/landscan. OECD (Organisation for Economic Co-operation and Development). 2009. How Regions Grow: Trends and Analysis. Paris: OECD. Parent, O., and A. Zouache. 2008. "Determinants of Growth in Africa and Middle East: How to Implement Geographical Features and Institutional Factors?" Paper presented at Economic Research Forum 15th Annual Conference, Cairo. Puga, D. 2008. "Agglomeration and Cross-Border Infrastructure." European Investment Bank Papers 13 (2): 102­24. Roberts, Peter, K. C. Shyan, and Rastogi Cordula. 2006. "Rural Access Index: A Key Development Indicator." Transport Papers, March. World Bank, Washington, DC. Shokri, F. 2005. "Congestion, Pollution Still Plague Tehran." Iran Daily, January 13. Slackman, M. 2008. "A City Where You Can't Hear Yourself Scream." New York Times, April 14. Tobler, W. 1970. "A Computer Movie Simulating Urban Growth in the Detroit Region." Economic Geography 46 (suppl.): 234­40. UNDP (United Nations Development Program). 2008. Poverty, Growth, and Income Distribution in Lebanon. Beirut: UNDP. United Nations and World Bank. 2003. Iraq Joint Needs Assessment. Transport and Telecommunications. Baghdad: United Nations and World Bank. Van Well, L. 2006. ESPON Project 1.1.3. Enlargement of the European Union and the Wider European Perspective as Regards Its Polycentric Spatial Structure. Part 1. Summary. Esch-sur-Alzette, Luxembourg: European Spatial Planning Observation Network. 244 Poor Places, Thriving People Warr, P. 2005. Road Development and Poverty Reduction: The Case of Lao PDR. Manila: Asian Development Bank Institute. Warren, J. 2007. "The Digital Vicious Cycle: Links between Social Disadvantage." Telecommunications Policy 13: 374­88. Wellenius, B. 2002. Closing the Gap in Access to Rural Telecommunications: Chile, 1995­2002. Washington, DC: World Bank. World Bank. 1989. Road Deterioration in Developing Countries: Causes and Remedies. Washington, DC: World Bank. ------. 2000. Cairo Urban Transport Note. Washington, DC: World Bank. ------. 2002. Best Practices in Customs Reform--Lessons from Morocco. Washington, DC: World Bank. ------. 2004. Hashemite Kingdom of Jordan. Amman Development Corridor Project Appraisal Document. Washington, DC: World Bank. ------. 2005a. Djibouti Transport Sector Review. Washington, DC: World Bank. ------. 2005b. Islamic Republic of Iran Transport Sector Review and Strategy Note. Washington, DC: World Bank. ------. 2006a. Lebanon Urban Transport Development Project. Project Appraisal Document. Washington, DC: World Bank. ------. 2006b. La Logistique du Commerce et la Competitivité du Maroc. Washington, DC: World Bank. ------. 2006c. Upper Egypt: Challenges and Priorities for Rural Development. Washington, DC: World Bank. ------. 2007a. People's Democratic Republic of Algeria: A Public Expenditure Review. Volumes I and II. Washington, DC: World Bank. ------. 2007b. Arab Republic of Egypt. Enhancing the Performance of the Road Sector. Washington, DC: World Bank. ------. 2008. Syrian Arab Republic. Technical Assistance to the Transport Sector. Concept Note. Washington, DC: World Bank. ------. 2009a. World Development Report 2009: Reshaping Economic Geog- raphy. Washington DC: World Bank. ------. 2009b. Tapping a Hidden Resource: Energy Efficiency in the Middle East and North Africa. Washington, DC: World Bank. Policy Package 2. Connecting Poor Places to the Poles of Development 245 ------. 2009c. Information and Communications for Development: Extending Reach and Increasing Impact. Washington, DC: World Bank. ------. 2009d. Regional Cross-Border Trade Facilitation and Infrastructure Study for Mashreq Countries. Washington, DC: World Bank. ------. 2009e. "Project Paper: Yemen Second Rural Access Project." World Bank, Washington, DC. CHAPTER 6 Policy Package 3. Underpin Private Sector Interest in Nonleading Areas Policy Package 3 (adapted from table 2.3) Package 3. "Incentives" Sometimes non- Areas where Beware of Underpin ("spatially leading areas the private incentives private sector focused do move ahead sector sees and induce- interest in policies") to leading-area unrealized ments to nonleading status (e.g., growth locate areas Tunisia's center- potential businesses east, Morocco's in lagging Tanger-Tétouan) areas Our first policy package was aimed at leveling the playing field for development. It was applicable to all lagging areas. The second policy package was about helping lagging areas to benefit from spillover con- nectivity, and only places close enough to growth poles stood to gain. Now we will look at the third policy package, where the aim is to realize a lagging area's economic potential. Regional economic development strategies have a poor track record and should not be entered into lightly. If the lagging-area problem is likely to resolve itself in an acceptable time frame through out-migration, a targeted regional economic development approach will be redundant. If the lagging area in question lacks the human and natural resource base to attract private investment, a targeted regional economic development approach will be unsuccessful. Investing in people and leveling the play- ing field for development (Policy Package 1) should, therefore, always be the foundation of a lagging-area strategy. However, sometimes it does make sense to layer a targeted regional development approach on top of Policy Package 1. It makes sense where out-migration is not going to solve the lagging-area problem and where there is a genuine prospect of competitive private cluster development. 247 248 Poor Places, Thriving People Although most economically lagging areas remain lagging areas because of the forces of agglomeration, a few do catch up: · In 1980, Tunisia's center-east region's mean per capita household con- sumption was 63 percent of that of Tunis. Since then, living standards have risen rapidly and converged, reaching 91 percent of the level of Tunis by 2000. But the south and the northeast, which had approxi- mately the same level of consumption as the center-east in 1980, got left behind; their levels of consumption were 61 and 68 percent (respectively) of Tunis's in 2000. · The Tanger-Tétouan region in Morocco had a mean per capita house- hold consumption in 1998 that was 60 percent of Greater Casablanca's. It ranked eighth out of Morocco's 15 regions in terms of per capita household consumption, alongside Fčs-Boulemane. As of the latest (2008) household income and expenditure survey, however, Tanger- Tétouan's mean per capita household consumption was 8 percent higher than Casablanca's, bringing Tanger-Tétouan to the number- two spot, behind the Rabat-Salé-Zemmour-Zaer capital region itself. Between 1994 and 2004, Tanger-Tétouan received 110,000 net in-migrants, equivalent to 4.5 percent of its 2004 population. Fčs- Boulemane, however, is now right at the bottom of the ranking. · The Islamic Republic of Iran saw an overall reduction in intergover- norate disparities between the pre- and post-Revolutionary period and 1985; the unweighted coefficient of variation of governorates' mean per capita household consumption fell from 0.27 to 0.22 dur- ing 1976­2005. But within that overall convergence, some individual lagging governorates made some remarkable progress. Bushehr, Ilam, and Mazandaran all rose at least 10 places in the governorate rankings, raising their per capita household consumption from 32­48 percent to 63­85 percent of Tehran's. These three provinces are far apart--Bushehr is on the Gulf, Ilam on the Iraqi border, and Mazan- daran on the Caspian Sea--so each has capitalized on a different latent competitive advantage to play catch-up. · Looking beyond MENA: In the United States, the states in the south- east and the south--such as the Carolinas, Georgia, Louisiana, and Mississippi--have reversed their status as senders of migrants from Reconstruction until the mid-20th century (World Bank 2009a) to become net receivers of internal migration (Suchan and others 2007; Perry 2006). Ireland was the second-poorest country in the European Union (EU) in 1995, in terms of purchasing-power-adjusted GDP per capita, but by 2002 it was the second-richest country. Policy Package 3. Underpin Private Sector Interest in Nonleading Areas 249 So lagging area catch-ups, though rare, can happen. The question for this chapter is: what role can governments play to facilitate the process? There is reason to believe that governments can facilitate the growth of agglomeration economies in lagging areas (World Bank 2009a). The Sapir Report on EU economic policy (Sapir 2003) was noncommittal on the impact of European Union regional and cohe- sion investments on lagging areas, but it came to the firm conclusion that national government had a strong role to play in determining a lagging area's economic prospects by establishing a favorable invest- ment climate and a level of administrative competence. Examples of successful programs for realizing the potential for locational compet- itiveness include Chile's Corporacion de Fomento, which focused on fruit exports; the Republic of Korea's Specialized Economic Zones (1972­81); and Ireland's promotion of high-technology investments nationwide (Dillinger 2007). The multisectoral South-West China Area Development Program found that the program investments, 43 percent of which were targeted at agricultural production, managed to steer the poorest villages of the region onto a sustained higher growth path (Jalan and Ravallion 1996). We will, therefore, start by looking at the performance of regional economic development policies in lagging areas around the world. We will show that these policies have generally failed because they have tried to fight against economies of agglomeration with spatially targeted pub- lic expenditures. The problem is that the economies of agglomeration are extremely powerful. So either public financial incentives are insuffi- cient to direct economic activity sustainably toward lagging areas, or they do so at a fiscal cost that is disproportionate to the benefits. We will then briefly describe regional economic development policies targeting lagging areas in the MENA13 economies. It will become apparent that the focus of these policies has also been on spatially tar- geted public expenditures, the formula that has proven unsuccessful else- where in the world. The chapter will then describe alternative approaches to regional eco- nomic development that are based not upon the distortion of economic forces but upon the realization of a lagging area's economic potential. In this model, the government's role is that of coordinating public and pri- vate players rather than that of using the public purse to influence the private sector's location decisions. In short, the old approach tries to combat agglomeration economies, but the new one tries to create coor- dination economies. Finally, we will ask how the new approach could be applied to a core sector for many lagging areas: agriculture. 250 Poor Places, Thriving People Difficulty and Expense of Using Public Money to Lure Development to Lagging Areas The impact of spatially targeted public expenditure is notoriously diffi- cult to evaluate. There are so many factors affecting a lagging area's development and so few similar regions to use for comparison that it is hard to work out what would have happened without the government's targeted spending. The Organisation for Economic Co-operation and Development (OECD) comments that "objective difficulties of evaluat- ing cross-sectoral policies, especially in quantitative terms, leave little material for research into the success or failures of policy" (OECD 2006a). Estimates of the impact of the United States's biggest regional development program, the Appalachian Regional Commission (ARC), vary from massive to negligible, depending on how comparator counties are selected (Glaeser and Gottlieb 2009). However, the available literature certainly has little positive to say about the potential of public expenditure to incentivize economic growth in lagging areas. One traditional approach has been to focus public infrastructural investment on lagging areas; however, a number of econometric stud- ies indicate that the returns to public investment are indeed highest in leading areas where they benefit from economies of agglomeration. Algeria's public investment into lagging areas since the end of the Civil War has had no apparent impact upon employment growth; the regions that grow fastest are those with rich neighbors (Belarbi and Zouache 2009). Infrastructure investment in the United States did not explain convergence between states (Lall and Yilmaz 2000). An econo- metric estimation of firms' location choices in Uganda showed that upgrading roads in a lagging province would increase its share of the food and beverage sector by only 0.3 or 0.4 percentage points, whereas a province adjoining the capital would benefit to the tune of 4.7 per- centage points (Lall, Schroeder, and Schmidt 2009). A review of the lit- erature (Eberts 1990) found evidence that public infrastructure invest- ment did support spatial economic convergence in the United States, Mexico, and Japan in certain circumstances: the investment had to be in a nonlagging (intermediate) area with some potential. Brakman and others (2005) show that the European Union's approach of connecting lagging areas to leading areas through improved infrastructure might actually have increased the leading-lagging divide by making it easier for leading areas to sell their products in lagging areas. Governments have also attempted to direct or steer corporate investment toward lagging areas, with similarly poor results. Despite massive public expenditures, Italy's Mezzogiorno's wages only briefly Policy Package 3. Underpin Private Sector Interest in Nonleading Areas 251 converged toward northern Italian levels--during the 1960s--and devel- opment subsidies stimulated rent-seeking behavior and corruption (Boltho, Carlin, and Scaramozzino 1997). In its time, the Soviet Union used to locate industrial plants in outlying areas for political or strategic reasons. Many of those factory towns are no longer viable in a market economy because of their distance from suppliers and consumers. The migration of workers back to denser areas is sometimes the best solution to their high unemployment rates (Bogetic 2008). The German govern- ment's use of spatially targeted investment subsidies to stimulate growth in the East German Länder in the 1990s distorted investment toward buildings and capital-intensive industries, despite the fact that capital was less productive in those lagging areas (Quehenberger 2000). In Brazil, a bidding war between the states of Bahia and Rio Grande do Sul for a Ford Motor Company investment left the taxpayer carrying a bill of $172,000 per job created, five times the cost of a job in Tennessee (Cal- mon 2003). The cost per job of the United Kingdom's Enterprise Zones has been estimated at $125,000 in 2007 prices, and that of the United States's Empowerment Zones was similarly high (Glaeser and Gottlieb 2009). Even where spatially targeted incentives have induced firms to invest in lagging areas, the gains have often been unsustainable. Brazil's subsidies for investment in the northeast were outweighed by national firms' preference for siting plants near their headquarters in the south (Carvalho, Lall, and Timmins 2006). Tunisia's experience of spatial investment subsidies was that it created a north-south "median band" of investment, as firms located as close as they could to the east coast without losing the subsidy. A notorious example in the United King- dom is Nissan's assembly plant in Sunderland, in England's lagging northeast, whose survival has been perennially insecure and dependent upon government subsidies (Castle 2001). MENA's Focus on Using Public Resources to Override Economies of Agglomeration So the global experience of using financial incentives to steer economic activity toward lagging areas has been negative. But in MENA, spatially targeted financial incentives are still widespread. Investment promotion law in most countries divides the national territory into different zones and establishes priority areas with preferential investment conditions. They consist primarily of tax exemptions and investment grants aimed at guiding private sector investments toward poorer locations. Table 6.1 summarizes the range of spatial incentives available in MENA. 252 Poor Places, Thriving People TABLE 6.1 Spatially Targeted Incentives for Business Are Prevalent in MENA ALG EGY JOR IRN IRQ LEB LIB MOR SYR TUN WBG YEM Fiscal incentives Extended corporate tax holidays ¸ ¸ ¸ ¸ ¸ ¸ ¸ ¸ ¸ ¸ ¸ ¸ Personal income tax holiday ¸ ¸ Exemption on social security contributions ¸ Exemptions on duties, customs, and indirect taxes ¸ ¸ ¸ ¸ ¸ ¸ ¸ ¸ ¸ ¸ Financial incentives General investment grant ¸ ¸ Discounted or free land ¸ ¸ ¸ ¸ ¸ ¸ Subsidized infrastructure costs ¸ ¸ ¸ ¸ ¸ ¸ Employment grants ¸ Technical assistance and training grants ¸ ¸ Source: Study Team survey of World Bank reports and Web documents. Note: Some incentives may not have been covered in the available documentation. Djibouti is not covered. ALG = Algeria, EGY = Egypt, Arab Rep., JOR = Jordan, IRN = Iran, Islamic Rep., IRQ = Iraq, LEB = Lebanon, LIB = Libya, MOR = Morocco, SYR = Syrian Arab Republic, TUN = Tunisia, WBG = West Bank and Gaza, YEM = Yemen, Rep. Location-specific exemptions from corporate taxation and duties cus- toms and indirect taxes are found throughout MENA (box 6.1). Partial or full corporate income tax holidays are granted to firms locating in the lesser-developed regions of virtually all economies in the MENA region. These holidays are often extensions of 5 to 10 years added to the general exemptions granted to investments in nonpriority locations. In the case of the Arab Republic of Egypt, corporate tax exemptions are granted for up to 20 years in the least-developed areas. Exemptions are usually extended to reinvested profits. Other fiscal incentives include the exemption of personal income taxes and social security contribu- tions for employees. Spatial exemptions on customs duties and other indirect taxes on the purchase of machinery, materials, and other items are also very common in the MENA region (Algeria, Jordan, Iraq, Lebanon, Libya, Morocco, Tunisia, and the West Bank and Gaza). Many countries also offer subsidies to investors in target areas. Investment grants cover a portion of the initial investment costs of a project and are often aimed at compensating for the lack physical infrastructure or human capital in remote areas. In Tunisia, the state allocates investment grants of up to 30 percent for projects locating in the priority investment zones. In Morocco, investment grants are offered to cover part of exter- nal infrastructure costs as well as professional training costs for projects Policy Package 3. Underpin Private Sector Interest in Nonleading Areas 253 BOX 6.1 Tax Breaks for Lagging Areas in MENA In Morocco, firms that locate in areas with low levels of economic activity benefit from a preferential fiscal treatment consisting of a 50 percent cut in corporate tax or income tax for five years. These preferences cover Morocco's periphery: the Mediterranean coast and the Rif in the north, the provinces close to the Algerian border in the east, and the far south (Source: Investment Charter, Morocco-ANPME). Tunisia offers tax incentives in the 140 "Zones de Développement Régional" (ZDR), in the western region and along the north-south central axis. Specifically, industry, crafts, and certain services are exempt from corporate tax and a personal income tax for their first 10 years and benefit from a 50 percent tax holiday for the following 10 years. Reinvested profits benefit from a complete tax deduction. Businesses are exempted from the Social Housing Promotion Fund for five years. In Jordan, projects initiated in Class C development areas, where economic activity is lowest, are granted a 75 percent corporate tax holiday for the first 10 years. Incentives in the more developed Class B and Class A development areas are 50 and 25 percent, respectively. In the Arab Republic of Egypt, firms located in remote areas benefit from a 10-year tax holiday (as opposed to the standard five-year period) for all investments. The exemp- tion is 20 years outside the Old Valley (World Bank 2008c). In Lebanon, investments in the poorer areas of northeast and southeast Lebanon, defined as Zone C, benefit from a complete exemption for a period of 10 years on income tax and taxes on dividends--as opposed to 50 percent for five years in Zone B areas and a maximum of two years in Zone A. In Libya, law no. 5 of 1997 on the promotion of investment of foreign capital allows projects constructed in "regional development areas" to receive an additional period of exemption by decision of the Libyan General People's Committee. In Algeria, law no. 1/2003 grants a 10-year exemption on various business taxes in the high plateaus and the south. Investments in underdeveloped areas also benefit from a real estate tax exemption and reduced duties on imports relevant to the investment (Khelladi 2009). In the Islamic Republic of Iran, the tax code grants a complete income tax exemption for a period of four years for extractive activities undertaken in "less developed regions," as opposed to the regular 80 percent exemption. In Iraq, investment law no. 13/2006 allows for an extension in the period of tax and fee exemption granted to all projects (from 10 years to a maximum of 15), depending on a project's geographical location. In the Republic of Yemen, projects in Investment Zone B benefit from an additional two years of profit taxes compared to projects in Zone A. (box continues on the following page.) 254 Poor Places, Thriving People BOX 6.1 (Continued) In the West Bank and Gaza, the law of 1998 grants extended tax exemptions for invest- ments in remote areas or for areas under the threat of settlement. In the Syrian Arab Republic, an extended two years of income tax exemption is granted for agricultural and industrial projects set up in developing governorates. Source: Study team, from Web sites and publications of national investment promotion bureaus. initiated in lagging areas. Technical assistance and training grants for investment in remote areas are also offered in Egypt. Some investment incentives are specifically aimed at facilitating access to land for investors in priority development areas. In Morocco, an additional investment grant is offered covering 20 percent of land costs for projects started in less-developed regions. The Egyptian gov- ernment grants land freely for investments in Upper Egypt and some areas of Middle Egypt. Access to land at particularly competitive prices is offered within special development zones in Jordan, the Islamic Republic of Iran, Morocco, and Tunisia. Other incentives support employment in lagging areas. In Egypt, a grant is offered for each per- son employed in Upper Egypt and certain areas of Middle Egypt for investments of a certain size. Special economic zones (SEZs) are often located in lagging areas as an instrument of regional economic development policy. This is the case of zones located in the oriental, northern, and southern regions of Morocco, in the inland governorate of Gafsa in Tunisia, and those in the southeast and northeastern regions of the Islamic Republic of Iran, where population is scarce and economic activity is, therefore, low (map 6.1). On top of fiscal exemptions and investment grants, SEZs usually offer streamlined business services, such as one-stop shops and simplified bureaucratic procedures. The physical infrastructure and the services provided reduce start-up costs for investors and to attract enter- prises of similar types, thereby favoring industrial clustering. Financial Incentives to Steer the Spatial Pattern of Development Have Not Been Successful Despite the prevalence of spatially targeted financial incentives in MENA economies, there is no clear evidence of their effectiveness. Policy Package 3. Underpin Private Sector Interest in Nonleading Areas 255 MAP 6.1 The Islamic Republic of Iran's Placement of Economic Zones in Low-Density Areas POPULATION DENSITY IBRD 38182 (POPULATION/SQ KM) 2005 NOVEMBER 2010 Armenia > 200 90 ­ 199 Ardabil 60 ­ 89 East Azerbaijan 30 ­ 59 < 20 Khorsn North Gjlan Golestan West Azerbaijan Zanjan Mazandaran Qazin Kordestan Khorsn Razavi Tehran Semnn Hamadan Kermanshah Qom Markazi Afghanistan Lorestan Ilm Esfahn Khorsn Chahar-Mahll Yazd Iraq Khuzestan Kohgluyeh Kerman FARS Pakistan Bushehr Hormozgan 0 50 100 150 200 250 KILOMETERS This map was produced by the Map Design Unit of The World Bank. The boundaries, colors, denominations and any other information shown on this map do not imply, on the part of The World Bank Group, any judgment on the legal status of any territory, or any endorsement or acceptance of such boundaries. Source: World Bank 2009b. Note: SEZ = special economic zone, FTZ = free trade zone. For a financial incentive program to be deemed a success, it would have to pass three tests: · The response test. Investors would have to be observed to locate in the targeted area; if they don't, the incentive did not work. · The sustainability test. Businesses would have to remain in the targeted area once the incentive is withdrawn; if they don't, the incentive's impact was only temporary. · The economic efficiency test. If businesses need a special incentive to locate in a target area, it must be because there is a profitability dis- advantage to locating there. So the incentive must be at least as big as the profitability disadvantage for businesses to locate in the target area. 256 Poor Places, Thriving People Therefore the incentive is encouraging inefficient production, thus creating an economic efficiency loss for the economy. However, if enterprises in the target area eventually achieve economies of scale and agglomeration, then the target area's profitability disadvantage will be reduced, and the incentive will no longer be necessary. So the economic efficiency test is that growth in the target area eventu- ally becomes self-sustaining. What little information is available suggests that MENA's spatial finan- cial incentives have been no more successful than those in other parts of the world. From the 1960s onward, Algeria implemented regional economic development programs for priority areas. These programs included investment in state-owned industries. The strategy involved spreading industrial activity beyond Algiers and Oran with the creation of new industrial poles in peripheral coastal sites: Arzew in the west and Annaba and Skikda in the east. The 1968 program for the Amazigh (Berber)-speaking Aurčs region in eastern Algeria, for example, envis- aged expansion of a textile factory, a brick and tile plant, a tannery, a powdered milk factory, a furniture factory, mines, data processing cen- ters, and large-scale artisanal units. However, the continued polariza- tion in Algiers remained (Sutton 1976), both for public and private investments. Between 1966 and 1972, 1,116 out of 1,821 private investment projects submitted for state approval were for Algiers wilaya provinces alone. The second-place wilaya (oasis) attracted only 144 applications. In short, spatial financial incentives in Algeria did not pass the impact test. The Syrian Arab Republic also experienced difficulties during the 1990s in using industrial zones as incentives to locate businesses in lag- ging areas. Although the main vocation of the country's industrial zones was to ensure the appropriate zoning of industrial activity, to minimize pollution, and to preserve the residential environment, they were also used to encourage economic activity in lagging rural areas. The direc- tor of Syria's Private Industrial Sector (Shahin 1999) describes how the government provided land in rural industrial zones at "symbolic prices" with the aim of "limiting migration and lessening the pressure on serv- ices in large cities." Rural areas benefited from concessional loans, tax exemptions, and assistance with inputs and marketing. The government also deliberately located large public industrial enterprises in rural areas with the aim of reducing the cost of raw materials for the private sector there. Despite these efforts, economies of agglomeration prevailed: the private sector persisted in locating in urban areas because of the avail- ability of inputs and services. The rural areas remained competitive only Policy Package 3. Underpin Private Sector Interest in Nonleading Areas 257 in traditional handicraft activities, which depended on rural skills and creativity. Again, the spatial financial incentives failed the impact test. In Tunisia, until 2007, the government offered investment grants of 25 percent for firms locating in the "Zones de Développement Régional" (ZDR). The aim was to steer investments westward, away from the coast. This policy initially led to the concentration of invest- ments along the very eastern strip of the ZDR area, as opposed to being evenly spread out across the underdeveloped areas. In other words, the incentives passed the impact test but probably failed the sustainability and efficiency tests, because businesses were clearly still seeking to locate as far east as possible. Recognizing the situation and wishing to counter it, the government reclassified the ZDRs into three groups according to employment rates and human development levels. Today, the areas closer to the coast receive an investment grant of 8 percent and the central axis 15 percent; the areas defined as "Zones de Développe- ment Prioritaires"--mainly the western regions--receive an investment grant of 25­30 percent (see map 6.2). These incentives seem to be pass- ing the impact test, at least; Tunisian officials cite a number of major investors moving west, including the Italian clothing company, United MAP 6.2 Tunisia's Priority Development Zones ???? IBRD 38183 NOVEMBER 2010 ??? ??? Bizerte Ariana Manouba TUNIS Jendouba Béja Ben Arous Nebaul Zaghouan Le kef Siliana Sousse Monastir Kairouan Mahdia Kasserine Sidi Bou Zid sfax Gafsa Tozer Gabés Kébili Medenine Tataouine This map was produced by the Map Design Unit of The World Bank. The boundaries, colors, denominations and any other information 0 25 50 75 100 shown on this map do not imply, on the part of The World Bank Group, any judgment on the legal status of any territory, or any endorsement or acceptance of such boundaries. KILOMETERS Source: Tunisia-API. Note: The grey areas are Priority Development Zones. 258 Poor Places, Thriving People Colors of Benetton. It is too early yet to say whether they pass the sus- tainability and efficiency tests. As map 6.2 shows, some of the subsidized zones are very close to the booming east coast, and the fact that some businesses choose to locate there could be due more to spillover effects from agglomerated areas than to the financial incentives. New Paradigm of Regional Economic Development Policy Because financial incentives for regional economic development have such a poor track record, governments in developed countries are adopt- ing a new approach to regional economic development policy. This approach accepts that the forces of agglomeration are too strong to be overridden by public money. So it goes with the flow of the market, helping lagging areas to capitalize on genuine competitive advantages. Rejecting centralized and top-down approaches to planning, it emphasizes the need for local ownership of initiatives and consistent involvement of the private sector. The role of the government becomes one of facilitator: building links among (a) the private sector, (b) central government departments, (c) local administrations, and (d) education and research institutions (see table 6.2). It may be impossible to overpower the forces of agglomeration, the thinking goes, but there is no reason governments cannot help overcome diseconomies of coordination. We will now look at the elements of the new paradigm as they appear in the global literature, much of it by the OECD. Infrastructure is effective only in combination with education. A recent study (OECD 2009, 82) modeled the relationship between road infra- structure (kilometers of roadway per capita) and subnational growth rates (gross regional domestic product, or GRDP) across developed countries TABLE 6.2 Old and New Paradigms of Regional Economic Development Policy Old paradigm New paradigm Objectives Trying to compensate for locational Tapping underused potential to enhance regional disadvantages of lagging regions competitiveness Unit of intervention Administrative units Functional economic areas Strategies Sectoral approach Integrated approach Tools Subsidies and state aids Mix of soft and hard capital (capital stock, labor market, business environment, social capital, and networks) Actors Central government Different levels of government (with private sector) Source: OECD 2009. Policy Package 3. Underpin Private Sector Interest in Nonleading Areas 259 between 1995 and 2005. It found that infrastructural investment influ- ences growth only when combined with human capital and innovation. This relationship held in several model specifications at both territorial and subterritorial (TL3 and TL2) levels of spatial aggregation when con- trolling for a number of other factors. Looking at panel data, analysts found a lagged effect of infrastructure three years after the investment, but only in the presence of "innovation" (represented by patenting activ- ity). Using another model, they found a five-year effect, but only in the presence of "human capital" (measured by primary and tertiary educa- tion attainment rates). The main caveats were that the definition of infra- structure was very narrow (excluding communications, energy, and so on) and that the analysis also ignores private stocks of capital. The study cautioned that a lagging area that invests exclusively in infrastructure may fall victim to the "leaking by linking" phenomenon (Hirschman 1958), as businesses continue to agglomerate in previously vital areas but now use the new trunk connectivity networks to crowd out local producers who before were protected by their inaccessibility (OECD 2009, 38). The construction of motorways into Poland's lagging eastern regions will, for example, expose their producers' home markets to competition from more efficient areas (OECD 2008). The need to combine infrastructure with human capital is particularly stark when the aim is to facilitate ICT-enabled growth poles, which are clusters of high-tech firms that benefit from synergies of collocation. Examples of planned growth poles in MENA include the Smart Village on the outskirts of Cairo1 and the Sidi Abdallah Technopole 30 kilome- ters south of Algiers. Conversely, raising levels of higher education without providing con- nectivity and employment may have little effect on economic perform- ance. The city of Ostrava in the Czech Republic has a large number of higher educational institutions, but its growth is hampered by its dismal access to the main highway network (OECD 2004). If regional development is to be based on genuine local potential, regions need a decentralized, bottom-up approach to policy design and implementation. Though decisions arrived at by top-down approaches may be coherent with other national policies and priorities, regional actors are able to make better use of information about local capacities and local needs than are central authorities. Local actors are also more likely to embrace projects and support their implementation when they feel that they personally have been a part of the projects' design. Mobilizing local stakeholders may require both decentralization and a structured dialogue between the different levels of government. An example of this is the European Union's LEADER Community Initia- tive, which has figured prominently in the EU's rural development 260 Poor Places, Thriving People strategies since the early 1990s. The program aims to promote endoge- nous growth within a finite territory through decentralized local action groups, which craft development plans and manage projects. The model has made important impacts on both development and local governance (OECD 2006b, 90). In Mexico, one of the pillars of rural development policy is the government's bottom-up territorial perspective, exemplified by its creation of Rural Development Councils. Through these organs, municipalities decide on the acceptability of locally proposed projects within the broader framework of the state's rural development strategy (OECD 2007, 86). The Netherlands demonstrates a flexible approach with its Agenda for a Vital Countryside, a document published in 2004 to define rural policy. Although the Agenda for a Vital Countryside articu- lates national priorities and targets, it devolves much of the responsibility over translating these policies into action to the communities involved (OECD 2006b, 83). The public-private interface is a key source of regional productivity growth and innovation. So the policy maker's role is not to "pick winners" but to facilitate interaction between public agencies, private firms, and other institutions. This advice draws on the observation that regional growth and productivity increases are driven by the formation of clusters: "geo- graphically proximate group[s] of interconnected companies and associ- ated institutions in a particular field, linked by commonalities and complementarities" (Porter 2000). Clustered firms create direct and indirect employment, generate spillovers and stimulate productivity, and diversify regional economies through innovation (Hotz-Hart 2000). As the Asian experience illustrated, it is "devilishly hard" to deliber- ately generate clusters--and, therefore, foolhardy to make that an explicit goal (Yusuf 2008). But policy makers can still play an important role in fomenting interchanges between regional actors and assembling the links on which clusters are built. They can promote informal net- works and formal trade associations where appropriate. They can also ensure that public educational institutions cultivate stronger ties with the private sector. The dynamic relationship between government and industry can be thought of as a "discovery process," one that is "as much about eliciting information from the private sector on significant externalities and their remedies as it is about implementing appropriate policies" (Rodrik 2004, 4). Italy's ERVET (Regional Institution for Territorial Economic Devel- opment) is an excellent example of this. Created in the 1970s to support regional development in the region of Emilia-Romagna, it has a share- holding structure that engages the local public sector, credit and financial institutions, and entrepreneurial organizations, among others. Through its extensive network, ERVET has supported the small and medium Policy Package 3. Underpin Private Sector Interest in Nonleading Areas 261 enterprises (SMEs) that dominate the core of the economy there (Cooke and Morgan 1995). The Republic of Korea is often cited as an example of the vital role of private sector leadership for regional development. In the 1960s, and 1970s, the Korean government created New Indus- trial Cities under the guidance of the private sector, and they have flourished in subsequent decades (Park 2009). Closer to home, Morocco's 1996 Le Maroc Compétitif initiative identi- fied four sectors as candidates for cluster development: tourism, sea products, electronics and Internet technology, and textiles and apparel. Le Maroc Compétitif fostered a continuous dialogue between public and private institutions aimed at identifying and coordinating the invest- ments and regulatory reforms that were critical for cluster growth. It is impossible to separate the impacts of Le Maroc Compétitif from other drivers of growth in the targeted sectors. After all, they were selected for Le Maroc Compétitif precisely because of their potential. However, in terms of the political economy of reform, Le Maroc Compétitif was part of the growth of a durable partnership between the state and the organized private sector in the formulation of economic and industrial policy (Cammett 2007). Although Le Maroc Compétitif did not have spatial development objectives, its diagnosis-through-dialogue approach could well prove applicable to MENA's lagging areas. Because cluster-driven growth is multisectoral, policy makers need to shift increasingly to a territorial perspective rather than a sectoral one. For exam- ple, Chile's National Innovation Council found it difficult to identify the synergies and spillovers that would achieve the necessary buy-in from local stakeholders by means of a sectoral approach (OECD 2009, 76). The developed countries' rural policy record is another case in point. In general, the focus of their strategies for rural development has been the subsidization of agriculture, which represents a mere 10 per- cent of the total rural workforce (OECD 2006b, 13). Governments were slow to identify other potential sources of rural growth because they conceptualized rural areas as solely agricultural. Recently, however, developed-country governments have broadened their sectoral focus. In Italy, the Territorial Pacts initiative promotes the grouping of rural municipalities to achieve a critical mass for attracting investment (Johnston 2005). The Mexican Micro-Regions Strategy is also applauded for its territorial perspective. This strategy addresses the needs of the most marginalized rural populations by coordinating devel- opment strategy with them in a participatory manner within a geo- graphic framework, providing a package of 11 basic services through decentralized Strategic Community Centers.2 What all these approaches share in common is the impetus to mobilize local assets through interin- stitutional coordination. 262 Poor Places, Thriving People Governments must, therefore, consider the best governance arrangements for territorial coordination. The choice of structures will depend upon the existing administrative culture and systems. Writing of Australia, one author described the challenge: "to coordinat[e] a multitude of autonomous agents with different, often competing interests, agendas and views . . . without formal power or actual possibility to force others into acting according to one's own agenda and without full information" (Huber 2005). The final chapter of this report explores the institutional dimension of spatial development and describes various options for coordination. Realizing Agricultural Potential through Clusters Although rural households often draw their livelihoods from diverse sources, agriculture is still central to MENA's rural economies and is a cornerstone of employment in many of MENA's lagging areas. In the Islamic Republic of Iran in 1996, where 21 percent of the population was employed in agriculture, 30 percent of the population in the six ostans with the lowest per capita consumption were defined as "skilled agricultural and fishery workers." In Morocco, 43 percent of employ- ment in 2007 was in agriculture, rising to 58 percent in the six regions with the lowest per capita household consumption (Haut Commissariat du Plan 2007). The association between agricultural employment and lagging areas is, of course, only to be expected because both are less likely to be metropolitan. Any increase in the returns to farmwork can be broken down into the following components: · An increase in the ratio of land to labor · An increase in the productivity of land in terms of crop and livestock yields · A shift in the product mix toward higher-value products · An increase in the unit value of a given product We will look at each of these sources of growth in turn. The conclusion will be that resource availability and domestic markets place limits on productivist strategies, but that a combination of high-value products and raising the unit value of output is the best bet for MENA's lagging areas. Increasing the ratio of land to labor, as table 3.2 illustrated, has to be a major source of agricultural labor productivity increases in MENA. Policy Package 3. Underpin Private Sector Interest in Nonleading Areas 263 Significant increases in the area of agricultural land, however, are not feasible. The area of arable land in the MENA13 increased by 0.2 per- cent per year over 1970­2005, while the number of people employed in agriculture increased by 0.8 percent per year (gross domestic income, or GDI). We saw in chapter 3 that MENA's agricultural workforce growth is far greater than it would normally be for countries at this level of development. Furthermore, it is widely recognized that the expansion of MENA's agricultural land area in recent years has been into marginal, environmentally fragile areas and has contributed to land degradation in the region's semi-arid areas. Increases in the ratio of agricultural land to agricultural labor, therefore, have to come from a shift in employment out of farming. At present, two MENA13 countries, Iraq and Lebanon, have begun to see a reduction in the absolute number of people employed in agriculture (GDI). Productivity-based strategies also have their limits. Most MENA economies' ministries of agriculture see their main role as promoting an increase in the productivity of land in terms of crop and livestock yields. Higher productivity has the added benefit of reducing MENA economies' dependence on food imports, a major political concern, par- ticularly after the leap in food prices in 2007­08. The MENA13 economies have achieved excellent yield increases over recent decades, including a 2.1 percent average annual increase in cereals yields between 1970 and 2005. However, much of agriculture's yield gains have been thanks to its increased drawdown of the region's scarce water resources. In Syria, for example, 66 percent of the average increase in wheat yields from 1985­89 to 2001­05 stemmed from an increase in the area of wheat under irrigation (World Bank 2008b). The region's statistics show a rapid increase in the share of cropped land under irrigation: over 1970­2000, it increased by 20 percent in Morocco; 50 percent in Lebanon; 60 percent in the Republic of Yemen; and 100 percent or more in Algeria, Iraq, Libya, and Syria (GDI). The current situation is one in which agriculture is consuming the lion's share of water withdrawals (see table 6.3), while demographic and eco- nomic growth are driving up nonagricultural water demand, all MENA economies are overexploiting water resources in some localities, some (Egypt, Jordan, and the Republic of Yemen) are surviving on nonre- newable resources at the national level, and climate change is already reducing water supply. A second generation of irrigation investments is, therefore, appear- ing in MENA, where the aim is not to divert more water onto more land but to raise the productivity of water. The basic components of such a strategy include the pressurization of water delivery systems; the conversion of on-farm irrigation systems from flood to drip or 264 Poor Places, Thriving People TABLE 6.3 Agriculture Is Straining Water Resources in MENA Agricultural water Total water Total renewable withdrawals withdrawals water resources as Country (m3 x 109, 2002) (m3 x 109, 2002) % of total water use Algeria 3.9 6.1 236 Djibouti 0.0 0.02 1579 Egypt, Arab Rep. 59.0 68.3 85 Iran, Islamic Rep. 66.2 72.9 189 Jordan 0.8 1.0 87 Lebanon 0.9 1.4 n.a. Morocco 11.0 12.6 230 Syrian Arab Republic 18.9 20.0 132 Tunisia 2.2 2.6 174 Yemen, Rep. 6.3 6.6 62 Source: World Bank 2007, 160­92. Note: n.a. = not applicable. sprinkler; the mobilization of Water Users' Associations (WUAs) to manage water distribution, operations, and maintenance; and links to extension and marketing arrangements. Tunisia was an early adopter of the irrigation efficiency approach in MENA, and Morocco, Egypt, the Islamic Republic of Iran, and Syria are all committed to scaling it up (Jagannathan, Mohamed, and Kremer 2009, 170, 433). Some countries, notably Egypt and Iraq, can increase yields through improved drainage, as waterlogging has led to salinification and land degradation. Agronomic and husbandry improvements will continue to produce modest yield growth. This change includes improvements in cultivars and breeds, farming practices and use of equipment, and purchased inputs. However, the biological capacity of MENA's crop and animal species to deliver higher yields is critically dependent upon water availability. Farmers are often well aware of the technical possibilities being offered, but they lack access to credit or (especially in rainfed areas) are wary of the associated risk. The shift to high-value agricultural products--fruit and vegetables-- looks like one of the main sources of growth for MENA's agricultural sector. Although fruits and vegetables occupy only 7 percent as much land as cereals (FAOSTAT), their area has been growing by 2.1 percent per year since 1970, compared with 0.2 percent per year for cereals. Fruit and vegetables use water much more cost-effectively than alternative products; the economic return to water use for vegetables is 6.3 times higher than for wheat and 10 times higher than for beef cattle (World Bank 2003). Policy Package 3. Underpin Private Sector Interest in Nonleading Areas 265 But the size of the domestic market sets another limit on high-value cropping. For cereals, the market is not a problem because increased MENA production will substitute for imports. But the problem with increasing the supply of other farm products, such as fruit and vegeta- bles, is that when farmers produce more, the increase in supply drives prices down so far that farmers' incomes fall instead of rise. Table 6.4 shows how the farmgate prices for fruit and vegetables have tended to fall over recent years: 50 of the 72 country-product prices dropped over the period. The greater the increase in farmers' production, therefore, the higher the risk of a drop in their incomes. An increase in the unit values of agricultural output, however, seems to be a promising potential source of growth for MENA's agri- culture sectors. Improving the quality and marketing of food products is critical. MENA's growing urban middle class, increased female labor force participation, the tourism industry, and the explosion of super- market retailing, are all drivers of demand for higher-quality--and higher-priced--food products. Participation in modern, quality- focused supply chains can boost farmers' income in middle-income countries by 10­100 percent (Flores, Reardon, and Hernandez 2006); however, the requirement for quality and standardization can easily exclude smaller or capital-scarce farming operations (World Bank 2008a, 126­28). Developing export markets for MENA's fruit and vegetables is one way of raising unit values. The prices achieved by those who TABLE 6.4 Price Decreases Were the Rule for MENA's Fruit and Vegetable Farmers, 1995­2007 Price decrease 1995­2007 (%) Beans, Cucumbers Eggplants Onions, Peaches and Apples Apricots green and gherkins (aubergines) dry nectarines Tomatoes Algeria ­31 ­18 ­27 ­29 ­39 16 ­16 ­3 Egypt, Arab Rep. ­27 ­27 31 ­48 ­29 14 ­27 ­21 Iran, Islamic Rep. 5 19 ­19 ­15 2 ­21 11 ­44 Jordan ­44 9 ­43 29 45 ­28 ­51 26 Lebanon ­27 ­35 ­68 ­44 ­51 5 ­41 ­46 Morocco 35 60 ­45 ­42 ­42 ­33 60 ­42 West Bank and Gaza 36 ­52 ­11 14 11 21 ­29 80 Tunisia ­27 ­41 ­20 ­40 ­37 ­37 ­48 ­24 Yemen, Rep. 37 ­62 82 ­62 ­44 ­56 ­52 ­48 Sources: FAOSTAT; World Development Indicators Database, accessed October 13, 2009. Note: Figures for the West Bank and Gaza and some for the Republic of Yemen are for 2000­07. 266 Poor Places, Thriving People penetrate export markets can be orders of magnitude higher than those available on the domestic market (World Bank 2006b, 42­44). MENA economies have a fourfold trade competitiveness advantage when it comes to high-value cropping. First, the southern Mediter- ranean countries are a natural agro-ecological home for the multi- tude of food plants that originated in the Middle East. Second, MENA economies are extremely cost-competitive in fruit and vegetables, as demonstrated, for example, by domestic resource cost calculations for Syria and Tunisia (World Bank 2008b; World Bank 2006a). Third, they are close to the importing countries of the EU, the Gulf Coopera- tion Council (GCC), or both. Fourth, MENA's climate allows it to sell into the off-season for EU producers, when import restrictions are fewer and retail prices are higher. The catch is that high-income export markets' demands for quality and standardization are even more rigor- ous than those of high-end national markets, and the demands include the added challenges of phytosanitary compliance and traceability. Successful quality-based supply chains depend on the links among a range of private and institutional actors: · The farmers themselves · Private operators involved in finance, collection, sorting, packing and refrigeration, transport, and marketing · Agricultural input and equipment suppliers (for their knowledge and experience as well as for their products) · Extension and farmer education services · National public institutions responsible for phytosanitary certifica- tion and export formalities · Public institutions responsible for water and transport infrastructure So agricultural growth is cluster-driven too. As in OECD countries, the main thrust of agricultural policy in most MENA economies has long been productivism: public spending on water diversion for agri- culture and on price subsidies (for energy, irrigation water, applied inputs, and producer prices). A subsidy and productivism policy stance brings a mixture of benefits and costs. It helps countries to reduce dependence on international markets for staple foods. But it has a sig- nificant fiscal cost and reduces agriculture's contribution to the economy by distorting farmers' production choices. Production subsidies can either exacerbate or mitigate spatial inequality. Although they transfer resources from urban to rural areas, often the Policy Package 3. Underpin Private Sector Interest in Nonleading Areas 267 better-off rural areas receive a disproportionate share of the production subsidy because they are involved in a disproportionate share of agricul- tural production. But the small part of subsidies that goes to smaller farms or poorer areas can also make a critical contribution to farm incomes there. In Tunisia, for example, it is the well-watered wheat- growing areas that have received the lion's share of wheat price support and water tariff discounts (see table 6.5), but it is the lagging rural areas of Le Kef and Béja in the northwest that would suffer the greatest per- centage decrease in incomes if output markets were to be liberalized (Ideaconsult 2006). In Syria, the cotton-based economy of irrigated areas in the lagging northeast is entirely dependent on subsidies (World Bank 2008b), and Lebanon's tobacco subsidy primarily benefits the lagging southern provinces. On the basis of the previous discussion, however, it would appear that the main potential engine of agricultural growth in MENA's lagging areas is cluster development around quality-driven supply chains. Pro- duction subsidies can support farm incomes as long as they are in place, but they cannot be a source of sustained growth. MENA provides several excellent examples of the cluster and supply- chain approach being applied to lagging-area agriculture. · The Agricultural Exports and Rural Income (AERI) project in the lagging area of Upper Egypt built links among farmers, SME proces- sors, financial institutions, farm advisers, the Assiut Business Associ- ation, private vets, dairy processors, and input suppliers. Its approach was based upon a market assessment of unmet demand for quality products in urban centers and Upper Egypt's potential cost compet- itiveness advantage. The project's outcomes were positive. For exam- ple, at a cost of about $180 per farmer, its livestock component raised TABLE 6.5 Forms of Subsidization of Agriculture, 2005 EGY IRN IRQ LEB MOR SYR TUN WBG Applied inputs price subsidy X X X X X Producer price support for some commodities X X X X X X X User charge for public irrigation schemes below X X X X X X operating and maintenance cost (Wheat) Major fuel subsidy X X X X Source: World Bank 2007. Note: EGY = Egypt, Arab Rep., IRN = Iran, Islamic Rep., IRQ = Iraq, LEB = Lebanon, MOR = Morocco, SYR = Syrian Arab Republic, TUN = Tunisia, WBG = West Bank and Gaza. 268 Poor Places, Thriving People buffalo and cow milk production by 63 and 153 percent, respectively, and increased net incomes by 20 percent within a target group of 50,000 farmers (ACDI/VOCA). · Exports of Tunisia's high-value olive oil took off after the removal of the public sector monopoly in 1994. Nearly the entire product is exported unbranded and in bulk. However, there is a rapidly growing number of virgin bottled oil brands, which command far higher prices. The emergence of high-value olive oil products depends on farmers' ability to supply quality raw materials, processor access to finance and modern equipment, market information supplied by pro- ducer groupings, and state-run testing and certification facilities (World Bank 2006a). · Morocco's agricultural development strategy, Plan Maroc Vert, will test several key principles of the new paradigm for regional economic development. Plan Maroc Vert's model of agricultural development is cluster or supply-chain driven, and the government has assigned itself the role of facilitating links among farmers, trader-processors (aggrégateurs), service organizations, research institutions, and local administrations. The institutional framework for Plan Maroc Vert is decentralized to the regional level so it can ease the coordination among local public and private institutions. By the same logic, responsibility for the ORMVAs (Offices Régionales des Mise en Valeur Agricole) has been transferred from the Ministry of Agriculture's headquarters to its deconcentrated offices in the regions. The purpose of this brief discussion has been to illustrate how cluster- based, private sector­driven model of regional economic development can be applied to agriculture. As the previous examples show, the start- ing point is potential investor interest, and the key role for government is that of coordinator and provider of information. Conclusion Leveling the playing field (Policy Package 1) is the foundation for lagging-area development, but sometimes it is appropriate to add the targeted economic development package. Spatially targeted economic development strategies are relevant where out-migration is not a solu- tion and where there is a genuine prospect of private sector­led growth. This chapter has argued that public financial incentives for regional economic development do not work; their impacts are insufficient, too costly, or unsustainable. MENA economies, like the developed world, Policy Package 3. Underpin Private Sector Interest in Nonleading Areas 269 have nonetheless focused on them. Using the state's convening power to facilitate cluster-driven development makes more sense. Active investor participation in the formulation of regional develop- ment plans is the critical test. Private sector leadership is the sign that a lagging area has genuine competitive potential. Without private sector involvement in the process, however, a government risks targeting expenditures on areas where there is no genuine prospect of cluster- driven growth. Endnotes 1. http://www.smart-villages.com. 2. OECD 2006a (OECD Rural Policy Reviews: Mexico). References ACDI/VOCA. "Egypt--Agricultural Exports and Rural Incomes: Dairy and Livestock." http://www.acdivoca.org/acdivoca/portalhub.nsf/ID/ egyptAERI. Belarbi, Y., and A. Zouache. 2009. Regional Employment Growth and Spatial Dependence in Algeria, 1998­2005. Cairo: Economic Research Forum. Bogetic, Z. 2008. "Advancing Regional Development in Russia." In The World Bank's Experience with Regional and Territorial Development: Pro- ceedings from the World Bank Session of the European Commission's Open Days, Brussels, October 8. Washington, DC: World Bank. Boltho, A., W. Carlin, and P. Scaramozzino. 1997. "Will East Germany Become a New Mezzogiorno?" Journal of Comparative Economics 112 (4): 241­64. Brakman, S., H. Garretsen, J. Gorter, A. Van der Horst, and M. Schramm. 2005. New Economic Geography, Empirics, and Regional Policy. The Hague: CPB Netherlands Bureau for Economic Policy Analysis. Calmon, P. C. 2003. Evaluation of Subsidies in Brazil. Washington, DC: World Bank. Cammett, M. 2007. "Business­Government Relations and Industrial Change: The Politics of Upgrading in Morocco and Tunisia." World Development 35 (11): 1889­903. 270 Poor Places, Thriving People Carvalho, A., S. Lall, and C. Timmins. 2006. Regional Subsidies and Industrial Prospects of Lagging Regions. Washington, DC: World Bank. Castle, S. 2001. "Nissan Poised to Make Micra Decision as Sunderland Wins GBP 40 Million in Subsidies." Independent (London), January 18. Cooke, P., and K. Morgan. 1995. "Growth Regions under Duress: Renewal Strategies in Baden Wurttemberg and Emilia-Romagna." In Globalization, Institutions, and Regional Development in Europe, eds. A. Amin and T. Nigel, 91­117. Oxford, U.K.: Oxford University Press. Dillinger, W. 2007. Poverty and Regional Development in Eastern Europe and Central Asia. Europe and Central Asia Chief Economist's Regional Working Paper Series. Washington, DC: World Bank. Eberts, R. W. 1990. "Public Infrastructure and Regional Economic Development." Federal Reserve Bank of Cleveland Economic Review (Q1): 15­27. Flores, L., T. Reardon, and R. Hernandez. 2006. Supermarkets, New- Generation Wholesalers, Farmers Organizations, Contract Farming, and Lettuce in Guatemala: Participation by and Effects on Small Farmers. East Lansing: Michigan State University, Department of Agricultural Economics. Glaeser, E. L., and J. D. Gottlieb. 2009. The Economics of Place-Making Policies. Washington, DC: National Bureau of Economic Research. Haut Commissariat du Plan. 2007. Activité, Emploi et Chômage. Rabat, Morocco: Direction de la Statistique. Hirschman, A. O. 1958. The Strategy of Development. New Haven, CT: Yale University Press. Hotz-Hart, B. 2000. "Innovation Networks, Regions, and Globalization." In Oxford Handbook of Economic Geography, ed. G. Clark, M. Gertler, and M. Feldman, 432­50. Oxford, U.K.: Oxford University Press. Huber, W. 2005. "Federal Co-ordination in Australia." In New Approaches to Rural Policy: Lessons from around the World, ed. OECD, 61­67. Paris: Organisation for Economic Co-operation and Development. Ideaconsult. 2006. République Tunisienne: Ministčre de l'Agriculture et des Ressources Hydrauliques. Actualisation de l'étude de la compétitivité de la production agricole. Tunis. Policy Package 3. Underpin Private Sector Interest in Nonleading Areas 271 Jagannathan, N. V., A. S. Mohamed, and A. Kremer. 2009. Water in the Arab World: Management Perspectives and Innovations. Washington, DC: World Bank. Jalan, J., and M. Ravallion. 1996. "Are There Dynamic Gains from a Poor Area Development Program?" Policy Research Working Paper 1695. World Bank, Washington, DC. Johnston, D. 2005. "The New Economic Imperative in Rural Regions." In New Approaches to Rural Policy: Lessons from Around the World, ed. OECD, 15­20. Paris: Organisation for Economic Co-operation and Development. Khelladi, M. 2009. Inégalités spatiales en Algérie: tendances et politiques publiques. Unpublished background paper for this report. Lall, S., and S. Yilmaz. 2000. "Regional Economic Convergence: Do Policy Instruments Make a Difference?" Annals of Regional Science 35 (1): 153­66. Lall, S., E. Schroeder, and E. Schmidt. 2009. Identifying Spatial Efficiency- Equity Trade-offs from Territorial Development Policies: Evidence from Uganda. Washington, DC: World Bank. OECD (Organisation for Economic Co-operation and Development). 2004. OECD Territorial Reviews: Czech Republic. Paris: OECD. ------. 2006a. Reinventing Rural Policy. Paris: OECD. ------. 2006b. OECD Rural Policy Reviews: The New Rural Paradigm-- Policies and Governance. Paris: OECD. ------. 2007. OECD Rural Policy Reviews: Mexico. Paris: OECD. ------. 2008. OECD Territorial Review of Poland. Paris: OECD. ------. 2009. "Investing for Growth. Building Innovative Regions." Policy Report. Meeting of the Territional Development Policy Com- mittee (TDPC) at the Ministerial Level. Park, S. O. 2009. "A History of the Republic of Korea's Industrial Struc- tural Transformation and Spatial Development." In Reshaping Eco- nomic Geography in East Asia, eds. Y. Huang and A. M. Bocchi, 320­36. Washington, DC: World Bank. Perry, M. 2006. Domestic Net Migration in the United States: 2000 to 2004. Washington, DC: U.S. Census Bureau. 272 Poor Places, Thriving People Porter, M. E. 2000. "Location, Clusters, and Company Strategy." In Oxford Handbook of Economic Geography, ed. G. Clark, M. Gertler, and M. Feldman, 253­74. Oxford, U.K.: Oxford University Press. Quehenberger, M. 2000. "Ten Years After: Eastern Germany's Conver- gence at a Halt?" EIB Papers 5 (7): 117­36. Rodrik, D. 2004. Industrial Policy for the Twenty-first Century. Cambridge, MA: John F. Kennedy School of Government. Sapir, A. 2003. "An Agenda for a Growing Europe: Making the EU Eco- nomic System Deliver." Report of an Independent High-Level Study Group Established on the Initiative of the President of the European Commission. Shahin, B. 1999. "Industrial Sites in Syria." Journal of Economic Coopera- tion among Islamic Countries 20 (1): 71­84. Suchan, T., M. Perry, J. Fitzsimmons, A. Juhn, A. Tait, and C. Brewer. 2007. Census Atlas of the United States. Washington, DC: U.S. Census Bureau. Sutton, K. 1976. "Industrialisation and Regional Development in a Centrally Planned Economy--The Case of Algeria. Tijdschrift voor Econ. en Soc. Geografie 67: 83­94. World Bank. 2003. Trade, Investment, and Development in the Middle East and North Africa: Engaging with the World. Washington, DC: World Bank. ------. 2006a. Tunisia: Agricultural Policy Review. Washington, DC: World Bank. ------. 2006b. West Bank and Gaza Country Economic Memorandum. Growth in West Bank and Gaza: Opportunities and Constraints. Volume II. Washington, DC: World Bank. ------. 2007. Making the Most of Scarcity: Accountability for Better Water Management in the Middle East and North Africa. Washington, DC: World Bank. ------. 2008a. Agriculture for Development: World Development Report 2008. Washington, DC: World Bank. ------. 2008b. Syrian Agriculture: Towards the Social Market. Washington, DC: World Bank. Policy Package 3. Underpin Private Sector Interest in Nonleading Areas 273 ------. 2008c. "Government PSD Policies and Industrial Strategy in Egypt." Unpublished Background Note to the World Bank MENA Flagship Report on Private Sector Development. ------. 2009a. World Development Report 2009: Reshaping Economic Geography. Washington DC: World Bank. ------2009b. Iran--Employment Dynamics and Job Creation in Iran's Special Zones: Evidence from an Employers' Survey. Washington, DC: World Bank. Yusuf, S. 2008. "Can Clusters Be Made to Order?" In S. Yusuf, K. Nabeshima, and S. Yamashita, Growing Industrial Clusters in Asia, 1­39. Washington, DC: World Bank. CHAPTER 7 Proximity for All: Public Institutions for Spatial Policies This final chapter of the report draws the threads together. It asks one last question: how can governments organize themselves to address the spatial dimension of development? The title of this final chapter is "Proximity for All." Some of the ben- efits of development traditionally have been reserved for people who live close to the center--close to the center of government and to national institutions. This proximity has led to the perception that citizens are included or excluded from the fruits of development according to where they live. These perceptions of spatially differentiated treat- ment, in turn, fuel political anxiety about spatial disparities. So this section is about how administrative reforms can ensure that even those who live many kilometers from the capital can share the bene- fits of proximity to decision making. Spatial Disparities: A Big Task for the State To recap, this report began with the iron laws of economic geography. The forces of agglomeration draw economic activity and workers into areas of high density. Investments in agglomerated areas give higher returns than those in lagging areas. So there is a trade-off between the objective of raising national growth and that of spreading growth geo- graphically. We saw that people living in agglomerated areas are better off in terms of just about any indicator of development: monetary consump- tion, poverty rates, health, and education. This report then described how the spatial dimension heated the politics of inequality. A sense of place affects how people identify them- selves and how they define their relationship with the state. Political 275 276 Poor Places, Thriving People organization and representation are geographical phenomena. When ethnicity and religious affiliation map onto geography, the spatial aspect of politics becomes even more potent. Spatial disparities have, therefore, become a politically charged issue in MENA. As elsewhere in the world, they are seen as an injustice. A government's political success depends upon it being able to demonstrate that its citizens share the benefits of development, irrespective of the accident of where they live. So the critical policy question is that of how MENA governments can reduce spatial disparities without losing the economic benefits of agglomeration. This report's analysis showed that MENA governments, fortunately, have three opportunities: · Opportunity number 1. Governments can raise people's living stan- dards in lagging areas without distorting market forces. The key is investing in people and leveling the playing field for development. · Opportunity number 2. Lagging areas near major agglomerations can "hitch a ride" on their growth, provided that the required connectiv- ity is in place. · Opportunity number 3. There are cases--in MENA and elsewhere-- of areas outside the main agglomerations moving up the rankings by mobilizing their unrealized economic potential. So there are three policy packages: one that can apply to any lagging area, a second that requires proximity to growth centers, and a third that requires some form of unrealized economic potential. Before one weighs in with a policy response, therefore, it is important to diagnose what kind of lagging-area problem a place is suffering from. The study showed how lagging areas come in several varieties: high- and low-density, connected and isolated. But we saw that lagging-area problems are not always as serious as they first appear. Statistics have a way of concealing the concentra- tions of human deprivation in leading areas. Also, some of the differ- ence in living standards between rich places and poor places is due to household characteristics rather than spatial disparities in economic opportunity. The high fertility rates in lagging areas have conferred a legacy of high dependency ratios, and internal migration rates are too low to redistribute population toward leading areas. Indeed, in some countries, nearly all the gap between leading and lagging provinces are based on household characteristics. Low accountability indicators worldwide are usually associated with high agglomeration. MENA's history since the 19th century has Proximity for All: Public Institutions for Spatial Policies 277 bequeathed a legacy of political and administrative centralization. The outcome was a spatially unequal climate for development in which met- ropolitan areas fared better than the periphery. MENA leaders are now striving to undo this inheritance. This report has given a glimpse of what MENA governments are doing and might do to level the playing field for development in the domains of public finance, the business environment, urban development, education, health, and social safety nets. A special place is given to girls' education in lagging areas, which is the key policy for reducing spatial disparities. Lagging areas can take advantage of spillover connectivity if they are close enough to economic agglomeration to benefit from it. MENA's connectivity infrastructure (essentially the road network) suffers from an imbalance between high spending on upper-grade trunk roads on the one hand and shortfalls in maintenance, rural roads, urban public transportation, and border infrastructure on the other. Improving lagging areas' spillover connectivity will involve addressing these transport sector deficits. Governments can now also encourage the innovative use of new information and communication technologies to give peripheral areas at least some of the benefits of proximity to eco- nomic density. Some nonleading areas have caught up with leading areas by cash- ing in on their economic potential. The question is how governments can encourage spatial catch-up. Most governments around the world have focused their efforts on using distortionary financial incentives to push private investment toward lagging areas. But the evidence is that the incentive-push approach almost always fails; it is dispro- portionately expensive, ineffective, and unsustainable. So now govern- ments around the world are adopting an alternative regional economic development model in which the role of the state is to facilitate private investment by coordinating public and private actors around emerg- ing clusters. This approach is particularly applicable to MENA's agribusiness sector, which has the potential to drive development in many lagging areas. But this all seems pretty daunting. It sounds as though any govern- ment wishing to address spatial disparities must activate the entire machinery of the state: ministries of finance, education, health, social affairs, municipalities, transport, telecommunications, industry, agricul- ture, and public works, not to mention local administrations and the research community. Therefore, the next question is how the state can best organize and mobilize itself to address the spatial dimension of development. What institutional frameworks can tie so many strands into a manageable whole? 278 Poor Places, Thriving People How States Organize and Mobilize to Address the Spatial Dimension of Policy Perhaps because of their common historical legacy, many MENA coun- tries and economies have similar government structures (World Bank 2007). Figure 7.1 is a stylized and simplified illustration of the structure of the state in MENA. Most of the business of government has tradition- ally been conducted by the central government. Deconcentrated offices of the sector ministries have performed their mandates in the provinces, accountable to the ministerial hierarchy in the capital city. Appointed governors--reporting to the internal security ministry--have represented the government in the provinces. Advised by an appointed executive council, they have overseen the activities of local administrations; how- ever, governors and local governments have had limited direct influence over the planning and implementation of sectoral development (box 7.1). They have depended upon indirect and sometimes ad hoc relationships with the deconcentrated line ministries. Managing local development within such a centralized structure is clearly a massive challenge. On the one hand, the institutions with the power and resources--the sector ministries--are centrally accountable FIGURE 7.1 A Stylized Portrait of the Multilevel State in the MENA Region Central government Functional deconcentrated units . . . social affairs Functional deconcentrated units . . . agriculture Functional deconcentrated units . . . education Functional deconcentrated units . . . and so on Functional deconcentrated units . . . health Intermediate governments. Governor + appointed executive council Appointed Elected executive mayor Elected council Elected council Source: Adapted from World Bank (2007), and based on the Arab Republic of Egypt, the Islamic Republic of Iran, Jordan, Lebanon, Morocco, Tunisia, the Republic of Yemen, and the West Bank and Gaza. Proximity for All: Public Institutions for Spatial Policies 279 BOX 7.1 Examples of Appointed Provincial and Elected Municipal Local Administration in MENA · In Jordan, 12 governorates have governors appointed by the Ministry of Interior. There are then 51 districts and 89 subdistricts, of which the head is also nominated by the center. At the lower level, 93 municipalities are governed by mayors and local councils elected directly by the people. · Morocco has 39 provinces and 8 urban prefectures. Their governors are appointed by the king. The administrations of the country's 1,503 communes are accountable to elected councils. Legislation of 1997 envisaged the creation of regional (i.e., multi- province) administrations accountable to elected regional councils. The mandate of the regions is under consideration. · In Tunisia, the governors of the 24 governorates are appointed by the Ministry of Interior and operate under its supervision. Urban areas and rural-urban centers are governed by municipalities with an elected municipal council, which, in turn, elects mayors from within its ranks. · The West Bank and Gaza has a similar system although there is often a mismatch between formal provisions and practice resulting from the particular situation on the ground. · In the Islamic Republic of Iran, provinces, districts, and rural counties are deconcen- trated units, for which governors and administrators are appointed from the center. There are then more than 1,000 municipalities and 68,000 rural municipalities. These are not completely decentralized because they are governed jointly by the Ministry of Interior and locally elected councils, and the role of the latter is limited. · The Arab Republic of Egypt and the Republic of Yemen do not have decentralized gov- ernment units. There are elected local councils that operate within the deconcen- trated levels of government, but with limited real powers. Source: Study Team. and vertically organized, making coordination at the local level very dif- ficult. On the other, the institutions with the multisectoral territorial mandate--the subnational administrations--have very limited resources and capacity and sometimes also a restricted geographical coverage. A review of global experience suggests that there are at least five basic approaches whereby states can integrate the management of the various sectoral strands of spatial policy. These are described, in turn, below: ter- ritorial development planning, needs-based spatial allocation of fiscal resources, the program mode, deconcentration, and decentralization. 280 Poor Places, Thriving People Territorial Development Planning Territorial development planning (aménagement du territoire) has been asso- ciated with the francophone administrative tradition. (In fact, it does not even have an agreed-upon name in English!) This approach involves a central institution with the responsibility for monitoring and analyzing spatial development trends and for feeding spatial orientations into the national public investment planning process. So the defining character- istic of the aménagement du territoire approach is the injection of spatial analysis into national investment planning. Of the MENA economies, · Algeria's 2001 law on aménagement du territoire1 allows for a Territo- rial Master Plan (Schéma National d'Aménagement du Territoire-- SNAT), published in 2009, as well as master plans covering specific regions, wilayate, metropolitan areas, and coastal zones; however, for these exercises to become fully operational, the development planning process would have to be revised to incorporate them. This revision would, in turn, open the wider question of how subnational governments--which contribute to the various master plans--would participate in national development planning alongside the central government (Khelladi 2009, 49). · In Morocco, arrangements for territorial planning are evolving rapidly, driven by the political commitment to regionalization. King Mohammed VI's Royal Address of December 12, 2006, restated in November 2008, announced the strengthening of regional administra- tions with responsibility for the territorial development of clusters of provinces. A new ministerial department in charge of aménagement du territoire has been established. A Consultative Commission has been created to develop a new approach to regionalization, and it is consid- ering a revision of the current provincial and prefectoral boundaries to establish more economically and geographically rational territories. Two pilot Regional Spatial Development Plans (for the Tadla-Azilal and Meknčs-Tafilalet regions) and two pilot Regional Development Strategies (for the Fčs-Boulemane and the Souss-Massa regions) have recently been adopted by the Regional Councils. As in Algeria, how- ever, the question remains of how to integrate territorial planning within the national development planning and budgeting process. · Tunisia is perhaps the country that has taken the aménagement du ter- ritoire approach to its logical conclusion: the developing planning process is conceived as a matrix of "vertical" sectoral plans with "hor- izontal" area-based plans, all within the framework of a Territorial Master Plan (box 7.2). Proximity for All: Public Institutions for Spatial Policies 281 BOX 7.2 How Tunisia Integrates the Spatial Dimension into National Development Planning The current territorial planning approach in Tunisia dates from the preparation of the Seventh Plan. In 1985, the government created the Commissariat Général de Développe- ment Régional (CGDR). Its mandate was to integrate the territorial planning approach into the development planning cycle. The Commissariat Général is responsible for · Preparing a regional development plan · Disseminating information on the regions · Designing development programs and coordinating their implementation The CGDR is represented by a deconcentrated directorate in each governorate. The analytical basis of regional policy is the Schéma National d'Aménagement du Territoire (Territorial Master Plan). The Interministerial Committee for Territorial Planning (CIAT) steers its preparation, under the oversight of the Ministry of Development and International Cooperation (MDCI). The drafting is the job of the Territorial Planning Directorate General of the Ministry of Environment, Habitat, and Territorial Planning (MEHAT), which, in turn, draws on data provided by the governorates. A national mas- ter plan (Schéma d'Orientation National) aggregates the subplans for each economic region, governorate, and municipality. So the preparation of the five-year plans involves the integration of the regional plan- ning process above with the sectoral planning process led by the line ministries in Tunis. The intersection takes place at the level of the governorate. The Regional Development Directorate, the deconcentrated counterpart of the Ministry of Development at the gov- ernorate level, coordinates the regional planning process, liaising with the officers of the sectoral ministries posted to their governorate. Source: Study Team. Needs-Based Spatial Allocation Needs-based spatial allocation of fiscal resources can come in two forms: as untied grants or as tied program funding. The formula-based approach to the spatial allocation of fiscal grants is common in federal states, where the central government's political credi- bility depends upon its resource transfers being seen as transparent and equalizing. In Belgium, for example, the "National Solidarity Measure" has transferred since 1990 a sum of 11.6 per inhabitant for each per- centage point difference between a lagging region's income tax receipts and the national average (Gérard 2001). The Swiss government has 282 Poor Places, Thriving People published an "index of fiscal capacity" for each canton since 1959; can- tons' transfers depend on their index. Every five years, India's constitution requires the president to establish an independent Finance Commission to recommend the allocation of resources from the center to the states. Assistance is transferred according to the Gadgil formula, a weighted index that is based on states' populations, tax performance, per capita income, and vulnerability to "special problems" such as natural disasters or urban congestion. The European Union's Cohesion Policy now accounts for around a third of the European Community budget. Alloca- tions were decided by the European Council in 2005 for 2007­13 on the basis of objective statistics and a formula known as the "Berlin method." Each member state is allocated a per capita amount so that the poorest receive the most. The amount is then capped so that it cannot exceed a maximum percentage of GDP (European Commission 2007). In the absence of simple and transparent allocation rules, however, political bargaining is likely to prevail. In Argentina, for example, where 65 percent of provincial expenditures are funded by transfers from the cen- ter, the allocation rules are theoretically set out in the Coparticipatión Fed- eral de Impuestos (federal tax-sharing agreement). Indeed, there is a general trend for resources to be transferred from richer states to poorer states. There are so many loopholes and exceptions in the rules, however, that provincial politicians have been able to influence the allocations. The outcome is significant variability among the receipts of provinces with similar needs (Saiegh and Tommasi 1999). Intragovernmental bargaining is currently not really an important feature in MENA, because subna- tional administrations are politically dependent upon the center. The second way in which a central government steers fiscal resources toward poor areas is as the financier of nationwide schemes in which the level of activity in a given location is based upon need. This approach is common in national programs to improve social services, such as India's Sarva Shiksa Abhiyan (Education for All Program) or Reproductive and Child Health Program, where special allocations are made to poor states, and incentives applied for raising human development indicators in tribal areas. In the United States, 89 percent of the federal develop- ment budget is spent through some 180 sectoral programs, each with its own allocation mechanism (Drabenstott 2005). In MENA, a number of nationwide human development schemes follow this logic--for exam- ple, Morocco's National Human Development Initiative (INDH), and Egypt's 1000 Villages Program, where eligibility is based on spatially disaggregated development indicators. Intergovernmental contracting, whereby funds are transferred across levels of government for the ful- fillment of certain predetermined goals, is an option under the scheme approach. In Mexico, these contracts (known as convenios) are reported Proximity for All: Public Institutions for Spatial Policies 283 to have contributed to capacity building within subnational govern- ments and to have fostered greater trust across levels of government (OECD 2009, 7). We saw in chapter 4 that much scope remains in MENA to use needs-based fiscal allocation rules to support lagging areas. Moreover, if decentralization continues and if new centers of power emerge out- side the central government, rules-based fiscal allocations will become even more important to prevent the emergence of intragovernmental political bargaining for funds. The Program Mode With area-based programs, the approach is to bring the spatial dimen- sion to policy by creating dedicated funds and institutions for specific areas. This approach has the advantage of being quickly implementable without any restructuring of existing administrative structures, but the risk is that they undermine local government by creating parallel systems. The main design issue, therefore, is how to integrate the area- based program with the activities of local governments and the decon- centrated provincial and district departments of line ministries. MENA governments are grappling with the coordination challenge in different ways. The planning (identification) and implementation of activities can be either local or central. · Central identification. The Islamic Republic of Iran's Bureau for Deprived Regions reports directly to the Office of the President. It has its own dedicated departmental budget, which it deploys by ear- marking transfers to sector ministries for investments in specific areas. Typical projects include rural electrification and roads. This approach may have the advantage of quick "campaign mode" results--the Bureau reported 1,576 projects completed in 10 days (IRNA 2003). However, it runs the risk of low local ownership of the investments. · Decentralized identification and central implementation. Algeria's 1998 Special Fund for the Development of the Southern Regions was funded by a 1 percent levy on fuel. It funded projects for the envi- ronment, infrastructure, renewable energy, and quality of life in border areas. Projects are identified by the wilayate, approved by the Ministry of the Interior and the cabinet, and then incorporated into the Plan for the Southern Regions. The budget for 2006 was DA 432 billion, or around US$6 billion (Khelladi 2009). · Decentralized identification and deconcentrated implementation. Egypt's Shrouk program (since 1994) and the World Bank­funded Sohag 284 Poor Places, Thriving People Rural Development Project financed small-scale projects, principally in rural areas. The projects were identified by local communities in the form of Village Priority Development Plans. Projects were implemented by the deconcentrated offices of sector ministries at governorate and district levels. In the case of Sohag, sector ministries directly supervised construction for projects of less than US$250,000 and delegated the supervision of larger construction works to con- sultants. The limited level of delegation from Cairo to the program location created some difficulties: each sector ministry followed its own procurement procedures and microcredit applications took a long time to process (World Bank 2008). · A shift toward more deconcentrated identification and implementation. Tunisia's Office for Forestry and Livestock Development in the northwest (ODESYPANO) was created in 1981, with funding from the European Union, the World Bank, Germany, and the Food and Agriculture Organization. It covered five governorates and its basic approach was to convert scrubland into permanent grazing through the provision of forest protection, livestock breeding, and rural tracks. In its first phase (1981­89), the Ministry of Agriculture initi- ated and delivered these services directly through its dedicated ODESYPANO office. By 1989, however, it was clear that a lack of community ownership was lowering the project's impacts: projects were unnecessarily costly, the new rangelands were being overgrazed, and the new roads were not being well maintained. ODESYPANO, therefore, invested five years (1989­94) in building communities' capacity to plan and implement investments. The community-led approach--in place since 1994--has brought about lower contract prices and has improved the sustainability of investments. The chal- lenge now is that communities' local development plans call for investments beyond the mandate of the Ministry of Agriculture. ODESYPANO is, therefore, signing protocols with each governor, whereby the Délégué (subgovernor) will convene the deconcentrated officials of other sector ministries, along with local NGOs, to imple- ment community development plans. In the design of area-based programs, therefore, there is a trade-off between the implementation capacity of sector ministries--especially where large investments are concerned--and the need for local owner- ship. In countries such as Brazil and Spain, regional governments serve as a bridge between the center and local administrations, providing a combination of local ownership and critical institutional mass. A com- mon theme is that long-term investment in local capacity creates the foundation for successful area-based programs. Proximity for All: Public Institutions for Spatial Policies 285 Deconcentration Deconcentration brings decision making physically closer to citizens and thus makes central governments more responsive to local needs. In the context of MENA's administrative systems, deconcentration means the delegation of power within central ministries from administrative units in the national capital to administrative units in provincial capitals or district towns. Deconcentration, therefore, does not increase formal external accountabil- ity, but it does reduce the bureaucratic burden of sending decisions up and down the management hierarchy between the locality and the capital. And deconcentration does bring administrators physically closer to end users, thereby increasing their exposure to citizens' voice. It makes for a more efficient implementation of policies and enhances internal accountability by strengthening internal oversight (World Bank 2003). Another advan- tage of deconcentration from the policy makers' point of view is that each sectoral ministry can deconcentrate at its own pace by adjusting its own internal administrative procedures. The daunting prospect of a nationally coordinated reform process can, therefore, be avoided. Deconcentration is on the agenda across the MENA13, with propos- als under consideration in Algeria, Egypt, the Islamic Republic of Iran, Iraq, Jordan, Morocco, and Tunisia. Given the similarities of the region's administrative structures and political objectives, there is great scope for MENA's policy makers to learn from each other's experience. Some of the key lessons have been the following: · Deconcentration involves the delegation of three kinds of decision making--on activities, on finance, and on human resource manage- ment (HRM)--and it works best when all three are deconcentrated in parallel. If we take the case of Egypt's Ministry of Agriculture, local governors have authority over field staff. But the governorates' ability to act independently--for example, in the case of a pest invasion--is less than it appears, because the Ministry of Agriculture in Cairo still controls nonstaff budget releases. This arrangement leads to a weak double line of accountability (World Bank 2009b). The Moroccan gov- ernment is making progress on the delegation of decisions concerning activities and finance (box 7.3). But its HRM remains centralized--to the detriment of performance. · The central ministry's policy making and oversight functions remain critical when policy execution is deconcentrated. For example, the deconcentration of many of the functions of the Islamic Republic of Iran's Management and Planning Organisation (MPO) to ostan gov- ernors' offices in July 2007 may have weakened the organization's capacity to monitor and manage the economy from the center. In 286 Poor Places, Thriving People BOX 7.3 Morocco's Experimentation with Deconcentration Illustrates the Challenges Morocco's government has made a strong drive toward deconcentration in recent years though sometimes in an ad hoc and evolutionary manner. Morocco's choices illustrate the practical challenges of deconcentration and how one MENA country is addressing them. Supraministerial leadership. The monarchy has been a driving and coordinating authority behind deconcentration. The royal decree of December 2, 2005, established common rules for the organization of the central administrations and required central admin- istrations to conduct organizational audits with a view to prepare Deconcentration Master Plans (Schémas Directeurs de Déconcentration). Five line ministries, including the Ministry of Agricultural and Maritime Fisheries (MAPM) and the Ministry of Health (MH), prepared their master plans, but the preparation of additional plans has been put on hold since the Palace ordered the production of a National Deconcentra- tion Charter. The charter is to define principles for the deconcentration process with a view to support the regionalization agenda laid out by the Palace in the Royal Statements delivered on December 12, 2006, and November 6, 2008. The question, however, is whether a sufficiently strong leadership and coordination function for deconcentration is emerging within the government itself. The enabling framework. The Moroccan government is underpinning sectoral deconcen- tration with cross-cutting financial and HRM reforms. A recent circular by the Ministry of Economy and Finance simplified the procedures for transferring budget credits to deconcentrated units within sector ministries. The government has increased budget management autonomy for the sector ministries and their deconcentrated services by contrast, the Tunisian model of deconcentrated planning retains a leadership role for planning departments in Tunis (box 7.2). And under the Moroccan model, central departments exercise control over decentralized units by means of contract programs (box 7.3). · It becomes extremely important that the mandates of elected local authorities (municipalities, communes, and so on) are clearly defined when the deconcentrated units of sector ministries are reinforced. A lack of clarity over the demarcation of responsibilities between elected local authorities and appointed ministry officials can cause friction. · Deconcentration does need a strong central champion and coordi- nator to create a favorable environment for financial and HRM reforms. Morocco and Egypt have both adopted a sector-wide Proximity for All: Public Institutions for Spatial Policies 287 aggregating budget categories and allowing the ordonnateurs and sous-ordonnateurs to relocate within these categories without the ex ante approval of the Finance Ministry. The government has also taken steps to reduce HRM constraints on deconcentration: a government-wide Strategic Staffing Frameworks process; draft public service legislation establishing the principles of mobility, secondment, and external assignment; concours procedures that would increase autonomy of the line ministries; and financial incentives for redeployment to remote regions. (However, the legislation has been opposed by employee interests and has been stalled in the Senate since December 5, 2005.) The oversight function. Deconcentration of responsibilities is being backed by strength- ened internal accountability and by the reinforcement of each ministry's headquarters policy and oversight role. The Moroccan government's instrument of choice is the performance-based contract program between central and branch offices. As of mid- 2009, these programs existed in the Health, Agriculture, Transport, and Education ministries. Again, the question remains of where strong coordination takes place at the supraministerial level. The need for mandate clarity. In parallel with deconcentration within ministries, the gov- ernment is setting up regional (supraprovincial) administrations and is reinforcing the financial and technical capacities of elected local authorities. The government's social and economic program identifies the regions as a key interlocutor in the government's new spatial development strategy and the socioeconomic development of urban and rural territories. The government, therefore, realizes that it is now critical to clarify the man- dates and relationships among the three sets of administrative bodies.2 The Palace has indicated a preference for making the deconcentrated services of the sector ministries accountable to the (appointed) central government representatives at the regional level. Source: Study Team. approach to deconcentration, with sector ministries reforming their own internal decision-making rules at their own pace. This approach avoids the political challenge of setting one ministry in a coordi- nation role over the others. However, it has its limits: without a centrally agreed-upon legal and managerial framework for the del- egation of financial and HRM responsibilities, subnational units' powers may be more theoretical than real. Decentralization Decentralization is not a panacea. It can make spatial disparities either big- ger or smaller. What matters for reducing spatial disparities is narrowing the resource and capacity gap between local administrations in leading and lagging areas. 288 Poor Places, Thriving People Decentralization is a formal arrangement to improve external accountability at the local level. Through local elections, citizens acquire a direct stake in decision making (World Bank 2003). Decentralization can enhance transparency and discourage corruption (Fisman and Gatti 2000). Another motive for decentralization is the hope that lagging areas' citizens will consider themselves more included in the political process. Decentralization is not advanced in MENA. Instead, the region's sub- national government structures fall into one of two classes: · In the typical case, the country or economy has a two-tier subnational administrative structure. Intermediate-tier deconcentrated administra- tive units, appointed by and accountable to the center, have a substan- tial public service delivery function. Locally elected authorities have a much more limited mandate: to cover road maintenance, solid waste management, recreation services, markets, and so forth (e.g., the Islamic Republic of Iran, Jordan, Tunisia, and the West Bank and Gaza). · In the alternative case, elected local councils themselves share power with the deconcentrated units of centrally controlled ministries (e.g., Egypt and the Republic of Yemen). The high level of centralization in MENA's government structures is evidenced by local governments' low share of public spending. Figures for MENA economies are well below those of comparator middle- income countries and the world average (figure 7.2). Moreover, central governments often impinge on local administrations' financial management. FIGURE 7.2 MENA's Local Governments Spend Far Less Than Those in Comparable Economies 40 35 30 percentage of total public spending 25 20 15 10 5 0 . co p. za n n a an l ry ep ra Ira no isi Re ita Ga oc rd de bR n ba Jo Un or Tu n, Fe nd Le a M me Ar ld ka ld or or Ye t, an W W yp tB Eg es W MENA economies World Sources: World Bank 2007; Tosun and Yilmaz 2008. Proximity for All: Public Institutions for Spatial Policies 289 In some countries, revenues collected at the local level come under partial control of the center (e.g., the West Bank and Gaza, or Egypt). And, as noted in table 4.5, intergovernmental transfer rules are usually ad hoc and unpredictable, a condition that can keep local governments in a permanent state of financial and political dependency and can make it difficult for them to address local development priorities (Shah 1999). Would more decentralization and fiscal decentralization reduce spa- tial disparities in MENA? Maybe, but maybe not. There is a vast literature to the effect that the transfer of power and resources to lower levels of government helps match resources to needs (Rodriguez-Pose, Tijmstra, and Bwire 2009). If, as suggested earlier, MENA's lagging areas have suffered from a lack of political voice, surely decentralization can help level the playing field for development by giv- ing people in poor places more influence over policy making. The problem is that fiscal decentralization increases the lagging-area citizens' influence only over the area's local administration's resources, not over the national allocation of resources. It does not address the resource and capacity disparities among local administrations. Indeed, there is a "dark side" to fiscal decentralization. There is a risk (a) of a mismatch between fiscal responsibilities and spending mandates and (b) of the overloading of subnational governments' limited manage- rial capacity. Mismatches between roles and resources can lead to deteri- oration in service quality and the quality of spending. In a sample of middle- and high-income countries, greater fiscal decentralization was associated with a shift in public spending from investment to current expenditures and a slowdown in growth (Rodriguez-Pose, Tijmstra, and Bwire 2009). Decentralization of expenditures in post-1990 Central and Eastern European countries was also linked to lower growth, although decentralization of revenue powers seems to have been more positive (Rodriguez-Pose and Kroijer 2008). Fiscal decentralization is usually a political response to spatial disparities or ethnic division and is often, therefore, implemented without sufficient attention to the potential neg- ative economic consequences (IMF 2009). Moreover, it is far from fact that decentralization or fiscal decentraliza- tion per se is a remedy for spatial disparities (IMF 2009). Decentralization can improve or worsen the quality of local services provided to citizens, depending on the quality of the accountability systems in place. Transfer- ring powers from the center to an unaccountable--and maybe technically and managerially less competent--local elite can lead to worse outcomes. Local authorities' capacity is likely to be weakest in lagging areas, and the power of elites may also be more concentrated (World Bank 2006). Cross-country analysis (Ezcurra and Rodriguez-Pose 2009) reveals that decentralization in low- and middle-income countries is linked to, 290 Poor Places, Thriving People if anything, a faster increase in spatial disparities over time. This occurence might be because decentralization gives richer territories more political scope to control revenues, resulting in less horizontal redistribution toward poorer administrative divisions. Another possi- ble reason is that less-developed areas--particularly in developing countries--lack the technical and administrative capacity to take advantage of the powers devolved to them (Rodriguez-Pose and Gill 2004). In Argentina, for example, poorer states were unable to take advantage of the decentralization of the education sector because they lacked the resources to do so (World Bank 2006). In Brazil, decentralization burdened poorer regions and municipalities with mandates that they could not fulfill, leading to an increase in health disparities (Baiocchi 2006). In China, spatial disparities in public expenditures increased during the decentralization period of 1978­93 (Lin, Tao, and Liu 2006). So what is the verdict on decentralization and spatial disparities in MENA? We can sum up the global evidence in three propositions: · There are good reasons to continue decentralization in MENA: to improve accountability, responsiveness, and service quality, and to democratize local decision making. The empowerment of municipal authorities will strengthen their hand in planning for growing popu- lations (see chapter 4) and public transportation needs (see chapter 5). · But decentralization will not necessarily reduce spatial disparities, and it may increase them. Decentralizing mandates alone will probably increase spatial disparities by putting extra strain on the limited finan- cial and managerial capacities of local administrations in lagging areas. · So what matters for spatial disparities is creating transparent and pre- dictable arrangements for equalizing intragovernmental fiscal transfers, and building local governments' capacity in lagging areas (see box 7.4). In conclusion, although most MENA governments have a strong commitment to the development of lagging areas, few of them have decided what kind of institutional platform is needed for a regional- spatial approach to development policy. Governments in the MENA region have inherited centralized, vertical, sectoral administrative hier- archies. If they are to work spatially, they must be adapted to local, multisectoral policy making and implementation. Conclusion: Keeping Spatial Development Simple Spatial development is messy. It respects no sectoral boundaries. It is characterized by the diversity of places and of their inhabitants. Proximity for All: Public Institutions for Spatial Policies 291 BOX 7.4 Mobilizing Institutions for Spatial Policy Making: Seven Ideas for Policy Makers · Identify the institutional "home" or "champion" for spatial development--an institution with the influence within government to monitor and coordinate across ministries, and with the planning capacity to develop spatial policy. · Integrate a territorial planning (aménagement du territoire) exercise into the formula- tion of the national development plan. · Develop a transparent needs- or formula-based system for intragovernmental fiscal transfers. · Strengthen the formal role of deconcentrated and decentralized administrations in investment planning. · Review the "enabling environment" for deconcentration. Is there a strong "cham- pion" monitoring and coordinating the process? Even though activities may have been delegated, do regulations for the management of public finances and staff per- mit adequate delegation of financial and HRM decision making? · Review in an integrated way the mandates and resourcing of the different levels of government to ensure the overall logic and coherence of the whole. · Identify the capacity-building needs of decentralized administrations. Source: Study Team. Matching Policy Packages to Lagging Areas World Development Policy packages Report 2009 terminology Why? Where? Package 1. Level the "Institutions" Spatial disparities in well- Applicable to all lagging playing field and the ("spatially blind policies") being can narrow even areas opportunity for human while production is development in lagging agglomerating, so long as areas policies are appropriate. MENA has a historical legacy of policies' widening disparities. Package 2. Connect poor "Infrastructure" ("spatially For growth to spill over Lagging areas close places to the poles of connective policies") from economic poles into enough to density development lagging areas, they have to benefit to be connected. Package 3. Underpin "Incentives" Sometimes nonleading Areas where the private private sector interest in ("spatially focused areas do move ahead sector sees unrealized nonleading areas policies") to leading-area status growth potential (e.g., Tunisia's center-east, Morocco's Tanger-Tétouan) Sources: Study team and World Bank 2009a. 292 Poor Places, Thriving People It is full of contrasts between rich and poor, dense and sparse, near and distant. All things are interlinked and interacting: rainfall, eco- nomics, fertility, migration, public expenditure, institutional archi- tecture, and politics. Generalizations that work at one spatial level break down when one zooms in to a finer level of disaggregation or zooms out to a broader overview. It may all seem overwhelmingly complicated. This is probably why development policy institutions--the World Bank included--usually prefer to treat development sectorally rather than spatially. Sectors have neater boundaries, fewer institutions, and fewer relationships to think about. But MENA governments do not have the luxury of ignoring the spa- tial dimension of development. Geography is intrinsic to the everyday business of politics from Casablanca to Baluchistan. The aim of this report has, therefore, been to offer MENA's policy makers a structure within which to arrange the untidy reality of spatial development (table 2.3). It is a framework that begins with diagnosis and ends with three policy packages. The report has explored how the framework would apply in practice to MENA's specific lagging-area challenges. The findings can be sum- marized as eight simple axioms for spatial policy making in MENA. 1. Spatial development begins with clear institutional roles and processes. Spatial policy making is multisectoral and multi-institutional. The "process platform" for spatial development includes arrangements for monitor- ing and coordination, resource allocation, planning, and role definition. 2. Analysis is vital. MENA needs better spatially disaggregated statistics, especially broken down by gender. It is critical to check that spatial disparities really are a major development issue; very often, they are not as important as politics portrays them. Diagnosing the lagging area will identify the right policy package(s). 3. Good development is good spatial development. First and foremost, lagging areas need a level playing field for development. Simply by fulfilling their existing mainstream mandates to provide a positive business environment and services to the poor, sector ministries will be making a full contribution to national spatial development objectives. Girls' education in lagging areas is the top priority for spatial development. Erasing spatial disparities in the business environment should offer a high benefit at a particularly low cost. 4. Bringing social services to lagging areas usually means (a) finding the incentives for high-quality personnel to work in underprivileged areas and for poor people to use public services, and (b) tailoring Proximity for All: Public Institutions for Spatial Policies 293 delivery mechanisms to lower-density areas. Simply building more facilities might not solve the problem. 5. Commodity subsidies channel a lot of public money to leading areas. Tar- geted social protection schemes are more spatially equitable. 6. A short-distance spillover connection is essential. Short-distance spillover connectivity between leading and lagging areas is the transportation priority for MENA's lagging areas. Depending on a lagging area's specific geography, this connection could mean investment in rural feeder roads, roads maintenance, intra-urban public transportation, border logistics, or peri-urban radial connections, as well as better regulation of transport operators. Electronic connectivity offers new potential for lagging areas to obtain the benefits of proximity. 7. The best-practice approach to regional economic development emphasizes the coordination of multiple actors to facilitate private sector interest in an area's latent potential. Attempts to use financial incentives to shunt pri- vate investment toward lagging areas have a very bad track record. 8. Finally, spatial development is about choices. Agglomeration may be an inevitable part of growth and development, but governments can choose to welcome the benefits of agglomeration or to discourage them. They can choose between the paths of smooth and disruptive agglomeration. Agglomeration creates spatial disparities in develop- ment, but governments can choose between policies that exacerbate them or those that mitigate them. This report has succeeded if it has shone light on these choices and made their implications clearer, thus helping the Middle East and North Africa to ensure proximity for all. Endnotes 1. Law no. 01-20 of December 12, 2001 (Government of Algeria 2001). 2. There is a telling parallel with Chile, which created such a system of Regional Development Associations in 2006. The main weaknesses documented thus far include (a) their lack of capacity to conduct a thorough analysis of local needs and capacities, (b) the weak effect that their recommendations have on public invest- ment, and (c) their poorly defined mandate (OECD 2009, 24). References Baiocchi, G. 2006. "Inequality and Innovation: Decentralization as an Opportunity Structure in Brazil. In Decentralisation and Local 294 Poor Places, Thriving People Governance in Developing Countries: A Comparative Perspective, ed. P. Bardhan and D. Mookherjee, 53­80. Cambridge: MIT Press. Drabenstott, M. 2005. "A Review of the Federal Role in Regional Devel- opment." Federal Reserve Bank of Kansas City, Kansas City. European Commission. 2007. Growing Regions, Growing Europe: Fourth Report on Economic and Social Cohesion. Luxembourg: Office for Official Publications of the European Communities. Ezcurra, R., and A. Rodriguez-Pose. 2009. Does Decentralization Matter for Regional Disparities: A Cross-Country Analysis. London: SERC, London School of Economics. Fisman, R., and R. Gatti. 2000. Decentralization and Corruption: Evidence across Countries. Washington, DC: World Bank. Gérard, M. 2001. "Fiscal Federalism in Belgium." Conference on Fiscal Imbalance, Quebec City, September 13­14. IMF (International Monetary Fund). 2009. Macro Policy Lessons for a Sound Design of Fiscal Decentralization. Washington, DC: IMF. IRNA (Islamic Republic News Agency). 2003. "Iranian Government to Privatize Printing Houses, Postal Services." February 5. http://www. payvand.com/news/03/feb/1033.html. Khelladi, M. 2009. "Inégalités spatiales en Algérie: tendances et poli- tiques publiques." Unpublished background paper for this report. Lin, J., R. Tao, and M. Liu. 2006. "Decentralization and Local Gover- nance in China's Economic Transition." In Decentralization and Local Governance in Developing Countries: A Comparative Perspective, eds. P. Bardhan and D. Mookherjee, 305­27. Cambridge: MIT Press. OECD (Organisation for Economic Co-operation and Development). 2009. Policy Brief: Regional Innovation in 15 Mexican States. Paris: OECD. Rodriguez-Pose, A., and N. Gill. 2004. "Is There a Global Link between Spatial Disparities and Devolution?" Environment and Planning. 36: 2097­117. Rodriguez-Pose, A., and A. Kroijer. 2008. Fiscal Decentralization and Eco- nomic Growth in Central and Eastern Europe. Madrid: Instituto Madrileńo de Estudios Avanzados. Rodriguez-Pose, A., S. Tijmstra, and A. Bwire. 2009. "Fiscal Decen- tralization, Efficiency, and Growth." Environment and Planning. http://www.envplan.com/abstract.cgi?id=a4087. Proximity for All: Public Institutions for Spatial Policies 295 Saiegh, S. M., and M. Tommasi. 1999. "Why Is Argentina's Fiscal Fed- eralism So Inefficient? Entering the Labyrinth." Journal of Applied Economics 2 (1): 169­209. Shah, A. 1999. Balance, Accountability, and Responsiveness: Lessons about Decentralization. Washington, DC: World Bank. Tosun, M. S., and S. Yilmaz. 2008. Centralization, Decentralization and Conflict in the Middle East and North Africa. Washington, DC: World Bank. World Bank. 2003. Better Governance for Development in the Middle East and North Africa: Enhancing Inclusiveness and Accountability. Washington, DC: World Bank. ------. 2006. World Development Report 2006: Equity and Development. Washington, DC: World Bank. ------. 2007. Decentralization and Local Governance in MENA: A Survey of Policies, Institutions, and Practices. Washington, DC: World Bank. ------. 2008. Sohag Rural Development Project: Implementation Completion and Results Report. Washington, DC: World Bank. ------.2009a. World Development Report 2009: Reshaping Economic Geog- raphy. Washington, DC: World Bank. ------. 2009b. Egypt--Linking Funding to Outputs: Expenditures of the Min- istry of Agriculture and Land Reclamation. Washington, DC: World Bank. Appendix: MENA Region ALGERIA Country MENA13a Source General Population (millions), 2007 34 334 WDI Database GDP per capita, PPP (constant 2005 international dollars), 2008 7,422 6,198b WDI Database GDP growth (average annual %), 1999­2008 3.8 4.7b WDI Database Physical geography Surface area (1,000 km2), 2008 2,381 8,774 WDI Database Arable land (% of surface area), 2005 3 6 WDI Database Average precipitation (mm/year), 2008 89 137 FAO AQUASTAT Renewable water resources per capita (m3/inhab/year), 2008 340 1,014 FAO AQUASTAT Human geography Population growth (annual), 2008 1.5 1.9 WDI Database Labor force participation rate (% of population ages 15­64), 2007 60 52 WDI Database Population density (people/km2), 2008 14 38 WDI Database Population in cities > 1 million (% of total population), 2007 10 22c WDI Database Population in largest city (% of urban population), 2007 15 24d WDI Database Agglomeration index (settlement size: 100,000) 50 63 Uchida and Nelson 2008 Gender parity index for net primary enrollment rate, 2007 0.98 0.98e UNESCO Structure of the economy Agriculture, value added (% of GDP), 2008 9 11b WDI Database Industry, value added (% of GDP), 2008 69 44b WDI Database Energy wealth (net energy exports per capita), ktoe, 2007 3.75 1.28f IEA (International Energy Agency) Governance Voice and accountability, 2008 (percentile rank 0­100, 100 = best) 18 23.7 World Bank WGI Connectivity Paved roads (% of total road network), 2004 70 58 CIA Factbook Road density (km of paved roads/km2 of arable land), 2004 1 0.8 WDI Database/CIA Factbook Road density (km of paved roads/1,000 people), 2004 2.3 1.4 WDI Database/CIA Factbook Travel time to city index (%) 57 28 Author's calculations (source in Appendix) Rural access index (%), 2004 59 58 Roberts and others 2006 National average distance to capital city (km), 2000 1,108 436 World Bank 2009 Rail density (km of track/km2 of arable land), 2005 0.06 0.05 WDI Database/Euromed Rail density (km of track/1,000 people), 2005 0.13 0.09 WDI Database/Euromed (continued on next page) 297 298 Poor Places, Thriving People ALGERIA (continued) Country MENA13a Source Mobile cellular subscriptions (per 100 people), 2007 81 58g WDI Database Population covered by mobile cellular network (%), 2007 82 89 WDI Database Internet users (per 100 people), 2007 10 18 WDI Database Spatial disparities Mean per capita consumption ratio (capital region/national) .. 1.51h Mean per capita consumption ratio (urban/rural) .. 1.82i Urban Rural Source Population growth (% annual), 2008 2.5 ­0.3 WDI Database Fertility rate (births per woman), 2005 2.7 3.4 MICS3 Poverty rate (based on national poverty line), 1995 14.7 13.3 World Bank 1999 Improved water source (% of population with access), 2005 88 82 MICS3 2005 Improved sanitation (% of population with access), 2005 98 87 MICS3 2005 Stunting (%), 2005 10 12 MICS3 2005 Infant mortality rate, 2000 23 44 MICS2 2000 Under-5 mortality rate, 2000 40 56 MICS2 2000 Geographic character of Country MENA13 lagging area population (%) Sparse Dense Sparse Dense Far from major cities 15 48 6j 33 j Close to major cities 4 33 5j 56 j Algeria: Spatial Concentration of Economic Activity Source: G-Econ project, Yale, 2010. Appendix: MENA Region 299 Algeria: Lagging Area Typology Sources: Elevation, SRTM30 (2000); land cover, GLC2000 (2000); population, Landscan (2005); rail, VMAP0 (1997); rivers, CIA World Data Bank II (1980s); roads, Euro-med (2000) and VMAP0 (1997); slope, in degrees, SRTM30 (2000); urban areas, GPW3-GRUMP (2000); water bodies, Global Lakes and Wetlands Database layer 1 (2004). 300 Poor Places, Thriving People Algeria: Spatial Integration Diamond accountability 1.0 0.5 spatial equality connectivity 0.0 in living standards spatial equality in human development country MENA 13 Sources: Accountability: World Bank World Governance Index 2008; spatial equality in human develop- ment: Multiple Indicator Cluster Survey2 (MICS2) 2000; connectivity: author's calculation based on Nelson (2008). Sources in Appendix. Notes: a. Averages are weighted, whenever possible. b. Data unavailable for Islamic Republic of Iran and the West Bank and Gaza. c. Data unavailable for Djibouti, Tunisia, and the West Bank and Gaza. d. Data unavailable for Djibouti, Jordan, Tunisia, and the West Bank and Gaza. e. Data unavailable for Islamic Republic of Iran, Iraq, Libya, Syrian Arab Republic, and the Republic of Yemen. f. Data unavailable for Djibouti and the West Bank and Gaza. g. Data unavailable for Islamic Republic of Iraq. h. Data unavailable for Algeria, Libya, and the West Bank and Gaza. i. Data unavailable for Algeria, Lebanon, and Libya. j. Data unavailable for Djibouti. .. = not available; MICS = Multiple Indicator Cluster Survey. Because of missing data, the spatial concentration of economic activity maps are not available for Jordan, Lebanon, and Tunisia, and the West Bank and Gaza, and the lagging area map is not available for the Islamic Republic of Iran. Definition of Spatial Integration Diamond Indicators: Accountability: Voice and accountability index, reversed (percentile rank 0­100, 100 = best). Spatial equality in living standards: rural per capita consumption as percentage of urban per capita con- sumption (%). Spatial equality in human development: urban infant mortality as percentage of rural infant mortality (%). Connectivity: Travel time to city index (% of population). See indicator definitions for more details. Appendix: MENA Region 301 DJIBOUTI Country MENA13a Source General Population (millions), 2008 0.8 334 WDI Database GDP per capita, PPP (constant 2005 international dollars), 2008 1975 6198b WDI Database GDP growth (average annual %), 1999­2008 3 4.7b WDI database Physical geography Surface area (1,000 km2), 2008 23 8,774 WDI Database Arable land (% of surface area), 2005 0 6 WDI Database Average precipitation (mm/year), 2008 220 137 FAO AQUASTAT Renewable water resources per capita (m3/inhab/year), 2008 353 1,014 FAO AQUASTAT (estimate) Human geography Population growth (annual), 2008 0.8 1.9 WDI Database Labor force participation rate (% of population ages 15­64), 2007 70 52 WDI Database Population density (people/km2), 2008 37 38 WDI Database Population in cities > 1 million (% of total population) .. 22c Population in largest city (% of urban population) .. 24d Agglomeration index (settlement size: 100,000) 50 63 Uchida and Nelson 2008 Gender parity index for net primary enrollment rate, 2007 0.86 0.98e UNESCO Structure of the economy Agriculture, value added (% of GDP), 2007 4 11b WDI Database Industry, value added (% of GDP), 2007 17 44b WDI Database Energy wealth (net energy exports/per capita), ktoe .. 1.28f Governance Voice and accountability, 2008 (percentile rank 0­100, 100 = best) 16.3 23.7 World Bank WGI Connectivity Paved roads (% of total road network), 2006 39 58 World Bank 2006a Primary road density (km of paved roads/km2 of arable land), 2006 122.6 0.8 WDI database/World Bank 2006a Primary road density (km of paved roads/1,000 people), 2006 1.5 1.4 WDI database/World Bank 2006a Travel time to city index (%) 0 28 Author's calculations (source in Appendix) Rural access index (%), 2004 81 58 Roberts and others 2006 National average distance to capital city (km), 2000 85 436 World Bank 2009 Rail density (km of track/km2 of arable land), 2006 10.6 0.05 World Bank 2006b Rail density (km of track/1,000 people), 2006 0.13 0.09 World Bank Railway Database Mobile cellular subscriptions (per 100 people), 2005 6 58g WDI Database Population covered by mobile cellular network (%), 2006 75 89 WDI Database Internet users (per 100 people), 2007 1 18 WDI Database Spatial disparities Mean per capita consumption ratio (capital region/national), 2002 1.24 1.51h Djibouti National HH Survey 2002 Mean per capita consumption ratio (urban/rural), 2002 1.69 1.82i Djibouti National HH Survey 2002 Urban Rural Source Population growth (% annual), 2008 2.2 1.0 WDI Database Fertility rate (births per woman), 2006 2.19 2.38 MICS3 2006 Poverty rate (based on national poverty line) .. .. Improved water source (% of population with access), 2006 98 54 MICS3 2006 Improved sanitation (% of population with access), 2006 70 19 MICS3 2006 Stunting (%), 2006 32 44 MICS3 2006 Infant mortality rate, 2006 68 54 MICS3 2006 Under-5 mortality rate, 2006 95 73 MICS3 2006 302 Poor Places, Thriving People Geographic character of Country MENA13 lagging area population (%) Sparse Dense Sparse Dense Far from major cities .. .. 6j 33 j Close to major cities .. .. 5j 56 j Djibouti: Spatial Integration Diamond accountability 1.0 0.5 spatial equality connectivity 0.0 in living standards spatial equality in human development country MENA 13 Sources: Accountability: World Bank World Governance Index 2008; spatial equality in living standards: Djibouti National Household Survey 2002; spatial equality in human development: MICS3 2006; connec- tivity: author's calculation based on Nelson (2008). Sources in Appendix. Notes: a. Averages are weighted, whenever possible. b. Data unavailable for Islamic Republic of Iran and the West Bank and Gaza. c. Data unavailable for Djibouti, Tunisia, and the West Bank and Gaza. d. Data unavailable for Djibouti, Jordan, Tunisia, and the West Bank and Gaza. e. Data unavailable for Islamic Republic of Iran, Iraq, Libya, Syrian Arab Republic, and the Republic of Yemen. f. Data unavailable for Djibouti and the West Bank and Gaza. g. Data unavailable for Islamic Republic of Iraq. h. Data unavailable for Algeria, Libya, and the West Bank and Gaza. i. Data unavailable for Algeria, Lebanon, and Libya. j. Data unavailable for Djibouti. .. = not available; MICS = Multiple Indicator Cluster Survey. Because of missing data, the spatial concentration of economic activity maps are not available for Jordan, Lebanon, and Tunisia, and the West Bank and Gaza, and the lagging area map is not available for the Islamic Republic of Iran. Definition of Spatial Integration Diamond Indicators: Accountability: Voice and accountability index, reversed (percentile rank 0­100, 100 = best). Spatial equality in living standards: rural per capita consumption as percentage of urban per capita con- sumption (%). Spatial equality in human development: urban infant mortality as percentage of rural infant mortality (%). Connectivity: Travel time to city index (% of population). See indicator definitions for more details. Appendix: MENA Region 303 ARAB REPUBLIC OF EGYPT Country MENA13a Source General Population (millions), 2008 82 334 WDI Database GDP per capita, PPP (constant 2005 international dollars), 2008 5,011 6,198b WDI Database GDP growth (average annual %), 1999­2008 5 4.7b WDI database Physical geography Surface area (1,000 km2), 2008 1,001 8,774 FAO AQUASTAT Arable land (% of surface area), 2007 3 6 WDI Database Average precipitation (mm/year), 2008 51 137 FAO AQUASTAT Renewable water resources per capita (m3/inhab/year), 2008 703 1,014 FAO AQUASTAT Human geography Population growth (annual), 2008 1.8 1.9 WDI Database Labor force participation rate (% of population ages 15­64), 2007 50 52 WDI Database Population density (people/km2), 2008 82 38 WDI Database Population in cities > 1 million (% of total population), 2007 20 22c WDI Database Population in largest city (% of urban population), 2007 35 24d WDI Database Agglomeration index (settlement size: 100,000) 93 63 Uchida and Nelson 2008 Gender parity index for net primary enrollment rate, 2007 (estimate) 0.96 0.98e UNESCO Structure of the economy Agriculture, value added (% of GDP), 2008 14 11b WDI Database Industry, value added (% of GDP), 2008 36 44b WDI Database Energy wealth (net energy exports/per capita), ktoe, 2007 0.16 1.28f IEA Governance Voice and accountability, 2008 (percentile rank 0­100, 100 = best) 14.4 23.7 World Bank WGI Connectivity Paved roads (% of total road network), 2009 73 58 CIA Factbook Primary road density (km of paved roads/km2 of arable land), 2009 1.6 0.8 WDI Database/CIA Factbook Primary road density (km of paved roads/1000 people), 2009 0.6 1.4 WDI Database/CIA Factbook Travel time to city index (%) 20 28 Author's calculations (source in Appendix) Rural access index (%), 2004 77 58 Roberts and others 2006 National average distance to capital city (km), 2000 558 436 World Bank 2009 Rail density (km of track/km2 of arable land), 2005 0.17 0.05 WDI/World Bank Railway Database Rail density (km of track/1,000 people), 2005 0.07 0.09 WDI/World Bank Railway Database Mobile cellular subscriptions (per 100 people), 2008 51 58g WDI Database Population covered by mobile cellular network (%), 2007 94 89 WDI Database Internet users (per 100 people), 2008 15 18 WDI Database Spatial disparities Mean per capita consumption ratio (capital region/national), 2005 1.75 1.51h Demography and Health Survey 2005 Mean per capita consumption ratio (urban/rural), 2005 1.92 1.82i Demography and Health Survey 2005 Urban Rural Source Population growth (% annual), 2008 1.9 1.7 WDI Database Fertility rate (births per woman), 2005 2.7 3.4 Demography and Health Survey 2005 Poverty rate (based on national poverty line), 2004 and 2005 10.1 26.8 World Bank 2007 Improved water source (% of population with access), 2005 99 97 Demography and Health Survey 2005 Improved sanitation (% of population with access), 2005 85 52 Demography and Health Survey 2005 Stunting (%), 2005 16 18 Demography and Health Survey 2005 (continued on next page) 304 Poor Places, Thriving People ARAB REPUBLIC OF EGYPT (continued) Country MENA13a Source Infant mortality rate, 2005 32 45 Demography and Health Survey 2005 Under-5 mortality rate, 2005 39 56 Demography and Health Survey 2005 Geographic character of Country MENA13 lagging area population (%) Sparse Dense Sparse Dense Far from major cities 0 43 6j 33 j Close to major cities 0 56 5j 56 j Arab Republic of Egypt: Spatial Concentration of Economic Activity Source: G-Econ project, Yale, 2010. Arab Republic of Egypt: Lagging Area Typology Sources: Elevation, SRTM30 (2000); land cover, GLC2000 (2000); population, Landscan (2005); rail, VMAP0 (1997); rivers, CIA World Data Bank II (1980s); roads, Euro-med (2000) and VMAP0 (1997); slope, in degrees, SRTM30 (2000); urban areas, GPW3-GRUMP (2000); water bodies, Global Lakes and Wetlands Database layer 1 (2004). Appendix: MENA Region 305 Arab Republic of Egypt: Spatial Integration Diamond accountability 1.0 0.5 spatial equality connectivity 0.0 in living standards spatial equality in human development country MENA 13 Sources: Accountability: World Bank World Governance Index 2008; spatial equality in living standards: Demographic and Health Survey 2005; spatial equality in human development: Demographic and Health Survey 2005; connectivity: author's calculation based on Nelson (2008). Sources in Appendix. Notes: a. Averages are weighted, whenever possible. b. Data unavailable for Islamic Republic of Iran and the West Bank and Gaza. c. Data unavailable for Djibouti, Tunisia, and the West Bank and Gaza. d. Data unavailable for Djibouti, Jordan, Tunisia, and the West Bank and Gaza. e. Data unavailable for Islamic Republic of Iran, Iraq, Libya, Syrian Arab Republic, and the Republic of Yemen. f. Data unavailable for Djibouti and the West Bank and Gaza. g. Data unavailable for Islamic Republic of Iraq. h. Data unavailable for Algeria, Libya, and the West Bank and Gaza. i. Data unavailable for Algeria, Lebanon, and Libya. j. Data unavailable for Djibouti. .. = not available; MICS = Multiple Indicator Cluster Survey. Because of missing data, the spatial concentration of economic activity maps are not available for Jordan, Lebanon, and Tunisia, and the West Bank and Gaza, and the lagging area map is not available for the Islamic Republic of Iran. Definition of Spatial Integration Diamond Indicators: Accountability: Voice and accountability index, reversed (percentile rank 0­100, 100 = best). Spatial equality in living standards: rural per capita consumption as percentage of urban per capita con- sumption (%). Spatial equality in human development: urban infant mortality as percentage of rural infant mortality (%). Connectivity: Travel time to city index (% of population). See indicator definitions for more details. 306 Poor Places, Thriving People ISLAMIC REPUBLIC OF IRAN Country MENA13a Source General Population (millions), 2008 72 334 WDI Database GDP per capita, PPP (constant 2005 international dollars), 2007 10,346 6,198b WDI Database GDP growth (average annual %), 1999­2008 5.4 4.7b WDI database Physical geography Surface area (1,000 km2), 2008 1,745 8,774 WDI Database Arable land (% of surface area), 2005 10 6 WDI Database Average precipitation (mm/year), 2008 228 137 FAO AQUASTAT Renewable water resources per capita (m3/inhab/year), 2008 1,876 1,014 FAO AQUASTAT Human geography Population growth (annual), 2008 1.3 1.9 WDI Database Labor force participation rate (% of population ages 15­64), 2007 55 52 WDI Database Population density (people/km2), 2008 44 38 WDI Database Population in cities > 1 million (% of total population), 2007 23 22c WDI Database Population in largest city (% of urban population), 2007 16 24d WDI Databse Agglomeration index (settlement size: 100,000) 55 63 Uchida and Nelson 2008 Gender parity index for net primary enrollment rate .. 0.98e Structure of the economy Agriculture, value added (% of GDP), 2008 10 11b WDI Database Industry, value added (% of GDP), 2008 45 44b WDI Database Energy wealth (net energy exports/per capita), ktoe 1.89 1.28f IEA Governance Voice and accountability, 2008 (percentile rank 0­100, 100 = best) 8.2 23.7 World Bank WGI Connectivity Paved roads (% of total road network), 2006 73 58 CIA Factbook Primary road density (km of paved roads/km2 of arable land), 2006 0.8 0.8 WDI Database/CIA Factbook Primary road density (km of paved roads/1,000 people), 2006 1.8 1.4 WDI Database/CIA Factbook Travel time to city index (%) 37 28 Author's calculations (source in Appendix) Rural access index (%), 2004 66 58 Roberts and others 2006 National average distance to capital city (km), 2000 654 436 World Bank 2009 Rail density (km of track/km2 of arable land), 2005 0.04 0.05 WDI Database/World Bank 2005 Rail density (km of track/1,000 people), 2005 0.11 0.09 WDI Database/World Bank 2005 Mobile cellular subscriptions (per 100 people), 2008 60 58g WDI Database Population covered by mobile cellular network (%), 2007 95 89 WDI Database Internet users (per 100 people), 2008 32 18 WDI Database Spatial disparities Mean per capita consumption ratio (capital region/national), 2005 1.54 1.51h Iran Nationa HH Survey, 2005 Mean per capita consumption ratio (urban/rural), 2005 1.87 1.82i Iran Nationa HH Survey, 2005 Urban Rural Source Population growth (% annual), 2008 2.1 ­0.3 WDI Database Fertility rate (births per woman), 1998 1.8 2.4 Abbasi-Shavazi and McDonald 2005 Poverty rate (based on national poverty line), 2005 0.8 7.2 World Bank 2008c Improved water source (% of population with access), 2000 99 86 WHO-UNICEF Joint Monitoring Report Improved sanitation (% of population with access), 2000 86 78 WHO-UNICEF Joint Monitoring Report Stunting (%), 2000 11 22 Health Systems Profile: Iran Infant mortality rate, 2000 28 30 Mehryar 2004 Under-5 mortality rate, 2000 37 35 Mehryar 2004 Appendix: MENA Region 307 Geographic character of Country MENA13 lagging area population (%) Sparse Dense Sparse Dense Far from major cities 8 36 6j 33 j Close to major cities 6 50 5j 56 j Islamic Republic of Iran: Spatial Concentration of Economic Activity Source: G-Econ project, Yale, 2010. 308 Poor Places, Thriving People Islamic Republic of Iran: Spatial Integration Diamond accountability 1.0 0.5 spatial equality connectivity 0.0 in living standards spatial equality in human development country MENA 13 Sources: Accountability: World Bank World Governance Index 2008; spatial equality in living standards: Iran National Household Survey 2005; spatial equality in human development: Mehryar (2004); connectiv- ity: author's calculation based on Nelson (2008). Sources in Appendix. Notes: a. Averages are weighted, whenever possible. b. Data unavailable for Islamic Republic of Iran and the West Bank and Gaza. c. Data unavailable for Djibouti, Tunisia, and the West Bank and Gaza. d. Data unavailable for Djibouti, Jordan, Tunisia, and the West Bank and Gaza. e. Data unavailable for Islamic Republic of Iran, Iraq, Libya, Syrian Arab Republic, and the Republic of Yemen. f. Data unavailable for Djibouti and the West Bank and Gaza. g. Data unavailable for Islamic Republic of Iraq. h. Data unavailable for Algeria, Libya, and the West Bank and Gaza. i. Data unavailable for Algeria, Lebanon, and Libya. j. Data unavailable for Djibouti. .. = not available; MICS = Multiple Indicator Cluster Survey. Because of missing data, the spatial concentration of economic activity maps are not available for Jordan, Lebanon, and Tunisia, and the West Bank and Gaza, and the lagging area map is not available for the Islamic Republic of Iran. Definition of Spatial Integration Diamond Indicators: Accountability: Voice and accountability index, reversed (percentile rank 0­100, 100 = best). Spatial equality in living standards: rural per capita consumption as percentage of urban per capita con- sumption (%). Spatial equality in human development: urban infant mortality as percentage of rural infant mortality (%). Connectivity: Travel time to city index (% of population). See indicator definitions for more details. Appendix: MENA Region 309 IRAQ Country MENA13a Source General Population (millions), 2009 (estimate) 29 334 CIA Factbook GDP per capita, PPP (constant 2005 dollars), 2009 (estimate) 3,600 6,198b CIA Factbook GDP growth (average annual %), 1999­2008 .. 4.7b WDI database Physical geography Surface area (1,000 km2), 2008 438 8,774 WDI Database Arable land (% of surface area), 2005 13 6 WDI Database Average precipitation (mm/year), 2008 216 137 FAO AQUASTAT Renewable water resources per capita (m3/inhab/year), 2008 2,512 1,014 FAO AQUASTAT (estimate) Human geography Population growth (annual), 2009 (estimate) 2.5 1.9 CIA Factbook Labor force participation rate (% of population ages 15­64), 2007 43 52 WDI Database/CIA Factbook Population density (people/km2), 2008 66 38 CIA Factbook Population in cities > 1 million (% of total population) 21 22c WDI Database Population in largest city (% of urban population) 25 24d WDI Database Agglomeration index (settlement size: 100,000) 71 63 Uchida and Nelson 2008 Gender parity index for net primary enrollment rate .. 0.98e Structure of the economy Agriculture, value added (% of GDP), 2009 (estimate) 10 11b CIA Factbook/WDI Database Industry, value added (% of GDP), 2009 (estimate) 63 44b CIA Factbook/WDI Database Energy wealth (net energy exports/per capita), ktoe 2.44 1.28f IEA Governance Voice and accountability, 2008 (percentile rank 0­100, 100 = best) 12 23.7 World Bank WGI Connectivity Paved roads (% of total road network), 2003 84 58 World Bank-UN 2003 Primary road density (km of paved roads/km2 of arable land), 2003 0.6 0.8 WDI Database/World Bank-UN 2003 Primary road density (km of paved roads/1,000 people), 2003 1.4 1.4 WDI Database/World Bank-UN 2003 Travel time to city index (%) 5 28 Author's calculations (source in Appendix) Rural access index (%), 2004 58 58 Roberts and others 2006 National average distance to capital city (km), 2000 281 436 World Bank 2009 Rail density (km of track/km2 of arable land), 2005 0.04 0.05 WDI Database/World Bank-UN 2003 Rail density (km of track/1,000 people), 2005 0.1 0.09 WDI Database/World Bank-UN 2003 Mobile cellular subscriptions (per 100 people) .. 58g Population covered by mobile cellular network (%), 2006 72 89 WDI Database Internet users (per 100 people), 2008 1 18 CIA Factbook Spatial disparities Mean per capita consumption ratio (capital region/national), 2007 1.02 1.51h Shlash 2009 Mean per capita consumption ratio (urban/rural), 2007 1.58 1.82i Shlash 2009 Urban Rural Source Population growth (% annual) .. .. Fertility rate (births per woman), 2006 4 5.1 MICS3 2006 Poverty rate (based on national poverty line), 2007 16.1 31.3 Iraq National HH Survey 2007 Improved water source (% of population with access), 2006 92 57 MICS3 2006 Improved sanitation (% of population with access), 2006 98 82 MICS3 2006 Stunting (%), 2006 19 24 MICS3 2006 Infant mortality rate, 2006 35 35 MICS3 2006 Under-5 mortality rate, 2006 41 41 MICS3 2006 310 Poor Places, Thriving People Geographic character of Country MENA13 lagging area population (%) Sparse Dense Sparse Dense Far from major cities 2 9 6j 33 j Close to major cities 9 80 5j 56 j Iraq: Spatial Concentration of Economic Activity Source: G-Econ project, Yale, 2010. Iraq: Lagging Area Typology Sources: Elevation, SRTM30 (2000); land cover, GLC2000 (2000); population, Landscan (2005); rail, VMAP0 (1997); rivers, CIA World Data Bank II (1980s); roads, Euro-med (2000) and VMAP0 (1997); slope, in degrees, SRTM30 (2000); urban areas, GPW3-GRUMP (2000); water bodies, Global Lakes and Wetlands Database layer 1 (2004). Appendix: MENA Region 311 Iraq: Spatial Integration Diamond accountability 1.0 0.5 spatial equality connectivity 0.0 in living standards spatial equality in human development country MENA 13 Sources: Accountability: World Bank World Governance Index 2008; spatial equality in living standards: Shlash (2009); spatial equality in human development: MICS3 2006; connectivity: author's calculation based on Nelson (2008). Sources in Appendix. Notes: a. Averages are weighted, whenever possible. b. Data unavailable for Islamic Republic of Iran and the West Bank and Gaza. c. Data unavailable for Djibouti, Tunisia, and the West Bank and Gaza. d. Data unavailable for Djibouti, Jordan, Tunisia, and the West Bank and Gaza. e. Data unavailable for Islamic Republic of Iran, Iraq, Libya, Syrian Arab Republic, and the Republic of Yemen. f. Data unavailable for Djibouti and the West Bank and Gaza. g. Data unavailable for Islamic Republic of Iraq. h. Data unavailable for Algeria, Libya, and the West Bank and Gaza. i. Data unavailable for Algeria, Lebanon, and Libya. j. Data unavailable for Djibouti. .. = not available; MICS = Multiple Indicator Cluster Survey. Because of missing data, the spatial concentration of economic activity maps are not available for Jordan, Lebanon, and Tunisia, and the West Bank and Gaza, and the lagging area map is not available for the Islamic Republic of Iran. Definition of Spatial Integration Diamond Indicators: Accountability: Voice and accountability index, reversed (percentile rank 0­100, 100 = best). Spatial equality in living standards: rural per capita consumption as percentage of urban per capita con- sumption (%). Spatial equality in human development: urban infant mortality as percentage of rural infant mortality (%). Connectivity: Travel time to city index (% of population). See indicator definitions for more details. 312 Poor Places, Thriving People JORDAN Country MENA13a Source General Population (millions), 2008 6 334 WDI Database GDP per capita, PPP (constant 2005 international dollars), 2008 5,055 6,198b WDI Database GDP growth (average annual %), 1999­2008 5.6 4.7b WDI database Physical geography Surface area (1,000 km2), 2008 89 8,774 WDI Database Arable land (% of surface area), 2005 2 6 WDI Database Average precipitation (mm/year), 2008 111 137 FAO AQUASTAT Renewable water resources per capita (m3/inhab/year), 2008 153 1,014 FAO AQUASTAT estimate Human geography Population growth (annual), 2008 3.2 1.9 WDI Database Labor force participation rate (% of population ages 15­64), 2007 47 52 WDI Database Population density (people/km2), 2008 67 38 WDI Database Population in cities > 1 million (% of total population), 2007 18 22c WDI Database Population in largest city (% of urban population), 2005 .. 24d Agglomeration index (settlement size: 100,000) 77 63 Uchida and Nelson 2008 Gender parity index for net primary enrollment rate, 2007 1.02 0.98e UNESCO Structure of the economy Agriculture, value added (% of GDP), 2008 4 11b WDI Database Industry, value added (% of GDP), 2008 32 44b WDI Database Energy wealth (net energy exports/per capita), ktoe ­1.22 1.28f IEA Governance Voice and accountability, 2008 (percentile rank 0­100, 100 = best) 26.9 23.7 World Bank WGI Connectivity Paved roads (% of total road network), 2006 100 58 CIA Factbook Primary road density (km of paved roads/km2 of arable land), 2006 4.2 0.8 WDI Database/CIA Factbook Primary road density (km of paved roads/1,000 people), 2006 1.4 1.4 WDI Database/CIA Factbook Travel time to city index (%) 4 28 Author's calculations (source in Appendix) Rural access index (%), 2004 79 58 Roberts and others 2006 National average distance to capital city (km), 2000 171 436 World Bank 2009 Rail density (km of track/km2 of arable land), 2005 0.34 0.05 WDI Database/Euromed Rail density (km of track/1,000 people), 2005 0.11 0.09 WDI Database/Euromed Mobile cellular subscriptions (per 100 people), 2008 90 58g WDI Database Population covered by mobile cellular network (%), 2008 99 89 WDI Database Internet users (per 100 people), 2008 25 18 WDI Database Spatial disparities Mean per capita consumption ratio (capital region/national), 2002 1.54 1.51h Jordan National HH Survey 2002 Mean per capita consumption ratio (urban/rural), 2002 1.32 1.82i Jordan National HH Survey 2002 Urban Rural Source Population growth (% annual), 2008 3.3 3.0 WDI Database Fertility rate (births per woman), 2007 3.6 3.7 Population and Family Health Survey 2007 Poverty rate (based on national poverty line), 2002 12.9 18.7 World Bank 2004 Improved water source (% of population with access), 2007 98 94 Population and Family Health Survey 2007 Improved sanitation (% of population with access), 2007 92 76 Population and Family Health Survey 2007 Stunting (%), 2007 12 13 Population and Family Health Survey 2007 (continued on next page) Appendix: MENA Region 313 JORDAN (continued) Country MENA13a Source Infant mortality rate, 2007 20 23 Population and Family Health Survey 2007 Under-5 mortality rate, 2007 22 27 Population and Family Health Survey 2007 Geographic character of Country MENA13 lagging area population (%) Sparse Dense Sparse Dense Far from major cities 4 1 6j 33 j Close to major cities 4 91 5j 56 j Jordan: Lagging Area Typology Sources: Elevation, SRTM30 (2000); land cover, GLC2000 (2000); population, Landscan (2005); rail, VMAP0 (1997); rivers, CIA World Data Bank II (1980s); roads, Euro-med (2000) and VMAP0 (1997); slope, in degrees, SRTM30 (2000); urban areas, GPW3-GRUMP (2000); water bodies, Global Lakes and Wetlands Database layer 1 (2004). 314 Poor Places, Thriving People Jordan: Spatial Integration Diamond accountability 1.0 0.5 spatial equality connectivity 0.0 in living standards spatial equality in human development country MENA 13 Sources: Accountability: World Bank World Governance Index 2008; spatial equality in living standards: Jordan National Household Survey 2002; spatial equality in human development: Population and Family Health Survey 2007; connectivity: author's calculation based on Nelson (2008). Sources in Appendix. Notes: a. Averages are weighted, whenever possible. b. Data unavailable for Islamic Republic of Iran and the West Bank and Gaza. c. Data unavailable for Djibouti, Tunisia, and the West Bank and Gaza. d. Data unavailable for Djibouti, Jordan, Tunisia, and the West Bank and Gaza. e. Data unavailable for Islamic Republic of Iran, Iraq, Libya, Syrian Arab Republic, and the Republic of Yemen. f. Data unavailable for Djibouti and the West Bank and Gaza. g. Data unavailable for Islamic Republic of Iraq. h. Data unavailable for Algeria, Libya, and the West Bank and Gaza. i. Data unavailable for Algeria, Lebanon, and Libya. j. Data unavailable for Djibouti. .. = not available; MICS = Multiple Indicator Cluster Survey. Because of missing data, the spatial concentration of economic activity maps are not available for Jordan, Lebanon, and Tunisia, and the West Bank and Gaza, and the lagging area map is not available for the Islamic Republic of Iran. Definition of Spatial Integration Diamond Indicators: Accountability: Voice and accountability index, reversed (percentile rank 0­100, 100 = best). Spatial equality in living standards: rural per capita consumption as percentage of urban per capita con- sumption (%). Spatial equality in human development: urban infant mortality as percentage of rural infant mortality (%). Connectivity: Travel time to city index (% of population). See indicator definitions for more details. Appendix: MENA Region 315 LEBANON Country MENA13a Source General Population (millions), 2008 4 334 WDI Database GDP per capita, PPP (constant 2005 international dollars), 2008 10,877 6,198b WDI Database GDP growth (average annual %), 1999­2008 3.5 4.7b WDI database Physical geography Surface area (1,000 km2), 2008 10 8774 WDI Database Arable land (% of surface area), 2005 18 6 WDI Database Average precipitation (mm/year), 2008 661 137 FAO AQUASTAT Renewable water resources per capita (m3/inhab/year), 2008 1,074 1,014 FAO AQUASTAT Human geography Population growth (annual), 2008 1.0 1.9 WDI Database Labor force participation rate (% of population ages 15­64), 2007 54 52 WDI Database Population density (people/km2), 2008 405 38 WDI Database Population in cities > 1 million (% of total population), 2007 45 22c WDI Database Population in largest city (% of urban population), 2007 52 24d WDI Databse Agglomeration index (settlement size: 100,000) 76 63 Uchida and Nelson 2008 Gender parity index for net primary enrollment rate, 2007 0.99 0.98e UNESCO Structure of the economy Agriculture, value added (% of GDP), 2008 6 11b WDI Database Industry, value added (% of GDP), 2008 22 44b WDI Database Energy wealth (net energy exports/per capita), ktoe ­0.98 1.28f IEA Governance Voice and accountability, 2008 (percentile rank 0­100, 100 = best) 35.6 23.7 World Bank WGI Connectivity Paved roads (% of total road network), 2000 85 58 WDI database Primary road density (km of paved roads/km2 of arable land), 2000 3.3 0.8 WDI Database/SESRIC database Primary road density (km of paved roads/1,000 people), 2000 1.6 1.4 WDI Database/CIA Factbook Travel time to city index (%) 1 28 Author's calculations (source in Appendix) Rural access index (%), 2004 87 58 Roberts and others 2006 National average distance to capital city (km), 2000 60 436 World Bank 2009 Rail density (km of track/km2 of arable land), 2005 0 0.05 Rail density (km of track/1,000 people), 2005 0 0.09 Mobile cellular subscriptions (per 100 people), 2008 35 58g WDI Database Population covered by mobile cellular network (%), 2006 100 89 WDI Database Internet users (per 100 people), 2007 38 18 WDI Database Spatial disparities Mean per capita consumption ratio (capital region/national), 2004 1.34 1.51h Lebanon National HH Survey 2004 Mean per capita consumption ratio (urban/rural) .. 1.82i Lebanon National HH Survey 2004 Urban Rural Source Population growth (% annual), 2008 1.2 0.1 WDI Database Fertility rate (births per woman) .. .. Poverty rate (based on national poverty line), 1995 8 >25 Haddad 1996 Improved water source (% of population with access), 2006/1999 100 100 WDI Database Improved sanitation (% of population with access), 2006 100 87 WDI/WHO-UNICEF Joint Monitoring Rep. Stunting (%), 2005 .. .. Infant mortality rate, 2005 .. .. Under-5 mortality rate, 2005 .. .. 316 Poor Places, Thriving People Geographic character of Country MENA13 lagging area population (%) Sparse Dense Sparse Dense Far from major cities 1 3 6j 33 j Close to major cities 5 91 5j 56 j Lebanon: Lagging Area Typology Sources: Elevation, SRTM30 (2000); land cover, GLC2000 (2000); population, Landscan (2005); rail, VMAP0 (1997); rivers, CIA World Data Bank II (1980s); roads, Euro-med (2000) and VMAP0 (1997); slope, in degrees, SRTM30 (2000); urban areas, GPW3-GRUMP (2000); water bodies, Global Lakes and Wetlands Database layer 1 (2004). Appendix: MENA Region 317 Lebanon: Spatial Integration Diamond accountability 1.0 0.5 spatial equality connectivity 0.0 in living standards spatial equality in human development country MENA 13 Sources: Accountability: World Bank World Governance Index 2008; connectivity: author's calculation based on Nelson (2008). Sources in Appendix. Notes: a. Averages are weighted, whenever possible. b. Data unavailable for Islamic Republic of Iran and the West Bank and Gaza. c. Data unavailable for Djibouti, Tunisia, and the West Bank and Gaza. d. Data unavailable for Djibouti, Jordan, Tunisia, and the West Bank and Gaza. e. Data unavailable for Islamic Republic of Iran, Iraq, Libya, Syrian Arab Republic, and the Republic of Yemen. f. Data unavailable for Djibouti and the West Bank and Gaza. g. Data unavailable for Islamic Republic of Iraq. h. Data unavailable for Algeria, Libya, and the West Bank and Gaza. i. Data unavailable for Algeria, Lebanon, and Libya. j. Data unavailable for Djibouti. .. = not available; MICS = Multiple Indicator Cluster Survey. Because of missing data, the spatial concentration of economic activity maps are not available for Jordan, Lebanon, and Tunisia, and the West Bank and Gaza, and the lagging area map is not available for the Islamic Republic of Iran. Definition of Spatial Integration Diamond Indicators: Accountability: Voice and accountability index, reversed (percentile rank 0­100, 100 = best). Spatial equality in living standards: rural per capita consumption as percentage of urban per capita con- sumption (%). Spatial equality in human development: urban infant mortality as percentage of rural infant mortality (%). Connectivity: Travel time to city index (% of population). See indicator definitions for more details. 318 Poor Places, Thriving People LIBYA Country MENA13a Source General Population (millions), 2008 16 334 WDI Database GDP per capita, PPP (constant 2005 international dollars), 2008 1,4970 6,198b WDI Database GDP growth (average annual %), 1999­2008 4 4.7b WDI database Physical geography Surface area (1,000 km2), 2008 1,759 8,774 WDI Database Arable land (% of surface area), 2005 1 6 WDI Database Average precipitation (mm/year), 2007 56 137 FAO AQUASTAT Renewable water resources per capita (m3/inhab/year), 95 1,014 FAO AQUASTAT 2008 (estimate) Human geography Population growth (annual), 2008 1.9 1.9 WDI Database Labor force participation rate (% of population ages 15­64), 2007 55 52 WDI Database Population density (people/km2), 2008 4 38 WDI Database Population in cities > 1 million (% of total population), 2007 55 22c WDI Database Population in largest city (% of urban population), 2007 46 24d WDI Databse Agglomeration index (settlement size: 100,000) 77 63 Uchida and Nelson 2008 Gender parity index for net primary enrollment rate .. 0.98e Structure of the economy Agriculture, value added (% of GDP), 2009 (est.) 4 11b CIA Factbook/WDI Database Industry, value added (% of GDP), 2009 (est.) 78 44b CIA Factbook/WDI Database Energy wealth (net energy exports/per capita), ktoe 5.22 1.28f IEA Governance Voice and accountability, 2008 (Percentile Rank 0­100, 100 = best) 2.4 23.7 World Bank WGI Connectivity Paved roads (% of total road network), 2003 57 58 CIA Factbook Primary road density (km of paved roads/km2 of arable land), 2003 3.3 0.8 WDI Database/CIA Factbook Primary road density (km of paved roads/1,000 people), 2003 9.5 1.4 WDI Database/CIA Factbook Travel time to city index (%) 25 28 Author's calculations (source in Appendix) Rural access index (%), 2004 78 58 Roberts and others 2006 National average distance to capital city (km), 2000 910 436 World Bank 2009 Rail density (km of track/km2 of arable land), 2005 0 0.05 World Bank Railway Database Rail density (km of track/1,000 people), 2005 0 0.09 World Bank Railway Database Mobile cellular subscriptions (per 100 people), 2007 73 58g WDI Database Population covered by mobile cellular network (%), 2006 71 89 WDI Database Internet users (per 100 people), 2007 5 18 WDI Database Spatial disparities Mean per capita consumption ratio (capital region/national) .. 1.51h Mean per capita consumption ratio (urban/rural) .. 1.82i Urban Rural Source Population growth (% annual), 2008 2.2 1.1 WDI Database Fertility rate (births per woman), 2001 3.8 4.9 Health Systems Profile: Libya Poverty rate (based on national poverty line) .. .. Improved water source (% of population with access), 1995 98 99 WHO-UNICEF Joint Monitoring Report Improved sanitation (% of population with access), 1995 99 98 WHO-UNICEF Joint Monitoring Report Stunting (%), 2001 4 6 Health Systems Profile: Libya Infant mortality rate, 2001 21 31 Health Systems Profile: Libya Under-5 mortality rate, 2005 27.5 36.3 Health Systems Profile: Libya Appendix: MENA Region 319 Geographic character of Country MENA13 lagging area population (%) Sparse Dense Sparse Dense Far from major cities 4 37 6j 33 j Close to major cities 5 54 5j 56 j Libya: Spatial Concentration of Economic Activity Source: G-Econ project, Yale, 2010. Libya: Lagging Area Typology Sources: Elevation, SRTM30 (2000); land cover, GLC2000 (2000); population, Landscan (2005); rail, VMAP0 (1997); rivers, CIA World Data Bank II (1980s); roads, Euro-med (2000) and VMAP0 (1997); slope, in degrees, SRTM30 (2000); urban areas, GPW3-GRUMP (2000); water bodies, Global Lakes and Wetlands Database layer 1 (2004). 320 Poor Places, Thriving People Libya: Spatial Integration Diamond accountability 1.0 0.5 spatial equality connectivity 0.0 in living standards spatial equality in human development country MENA 13 Sources: Accountability: World Bank World Governance Index 2008; spatial equality in human develop- ment: Health Systems Profile: Libya 2007; connectivity: author's calculation based on Nelson (2008). Sources in Appendix. Notes: a. Averages are weighted, whenever possible. b. Data unavailable for Islamic Republic of Iran and the West Bank and Gaza. c. Data unavailable for Djibouti, Tunisia, and the West Bank and Gaza. d. Data unavailable for Djibouti, Jordan, Tunisia, and the West Bank and Gaza. e. Data unavailable for Islamic Republic of Iran, Iraq, Libya, Syrian Arab Republic, and the Republic of Yemen. f. Data unavailable for Djibouti and the West Bank and Gaza. g. Data unavailable for Islamic Republic of Iraq. h. Data unavailable for Algeria, Libya, and the West Bank and Gaza. i. Data unavailable for Algeria, Lebanon, and Libya. j. Data unavailable for Djibouti. .. = not available; MICS = Multiple Indicator Cluster Survey. Because of missing data, the spatial concentration of economic activity maps are not available for Jordan, Lebanon, and Tunisia, and the West Bank and Gaza, and the lagging area map is not available for the Islamic Republic of Iran. Definition of Spatial Integration Diamond Indicators: Accountability: Voice and accountability index, reversed (percentile rank 0­100, 100 = best). Spatial equality in living standards: rural per capita consumption as percentage of urban per capita con- sumption (%). Spatial equality in human development: urban infant mortality as percentage of rural infant mortality (%). Connectivity: Travel time to city index (% of population). See indicator definitions for more details. Appendix: MENA Region 321 MOROCCO Country MENA13a Source General Population (millions), 2008 31 334 WDI Database GDP per capita, PPP (constant 2005 international dollars), 2008 3,938 6,198b WDI Database GDP growth (average annual %), 1999­2008 4.3 4.7b WDI database Physical geography Surface area (1,000 km2), 2008 446 8,774 WDI Database Arable land (% of surface area), 2005 19 6 WDI Database Average precipitation (mm/year), 2008 346 137 FAO AQUASTAT Renewable water resources per capita (m3/inhab/year), 918 1,014 FAO AQUASTAT 2008 (estimate) Human geography Population growth (annual), 2008 1.2 1.9 WDI Database Labor force participation rate (% of population ages 15­64), 2007 54 52 WDI Database Population density (people/km2), 2008 70 38 WDI Database Population in cities > 1 million (% of total population), 2007 19 22c WDI Database Population in largest city (% of urban population), 2007 18 24d WDI Databse Agglomeration index (settlement size: 100,000) 47 63 Uchida and Nelson 2008 Gender parity index for net primary enrollment rate, 2007 0.95 0.98e UNESCO Structure of the economy Agriculture, value added (% of GDP), 2008 16 11b WDI Database Industry, value added (% of GDP), 2008 20 44b WDI Database Energy wealth (net energy exports/per capita), ktoe ­0.45 1.28f IEA Governance Voice and accountability, 2008 (percentile rank 0­100, 100 = best) 27.9 23.7 World Bank WGI Connectivity Paved roads (% of total road network), 2006 62 58 CIA Factbook Primary road density (km of paved roads/km2 of arable land), 2006 0.4 0.8 WDI Database/CIA Factbook Primary road density (km of paved roads/1,000 people), 2006 1.2 1.4 WDI Database/CIA Factbook Travel time to city index (%) 34 28 Author's calculations (source in Appendix) Rural access index (%), 2004 36 58 Roberts and others 2006 National average distance to capital city (km), 2000 369 436 World Bank 2009 Rail density (km of track/km2 of arable land), 2005 0.02 0.05 WDI/World Bank Railway Database Rail density (km of track/1,000 people), 2005 0.06 0.09 WDI/World Bank Railway Database Mobile cellular subscriptions (per 100 people), 2008 73 58g WDI Database Population covered by mobile cellular network (%), 2006 98 89 WDI Database Internet users (per 100 people), 2008 33 18 WDI Database Spatial disparities Mean per capita consumption ratio (capital region/national), 1998 1.16 1.51h Morocco National HH Survey 1998 Mean per capita consumption ratio (urban/rural), 1998 2.17 1.82i Morocco National HH Survey 1998 Urban Rural Source Population growth (% annual), 2008 1.8 0.4 WDI Database Fertility rate (births per woman), 2007 2.1 3 Santé en chiffres 2007 Poverty rate (based on national poverty line), 2007 4.8 14.5 Haut Commissariat au Plan Improved water source (% of population with access), 2004 98 57 WHO-UNICEF Joint Monitoring Report Improved sanitation (% of population with access), 2004 98 58 WHO-UNICEF Joint Monitoring Report Stunting (%), 2003 13 24 EPSF 2003 Infant mortality rate, 2007 33 55 Santé en chiffres 2007 Under-5 mortality rate, 2007 38 69 Santé en chiffres 2007 322 Poor Places, Thriving People Geographic character of Country MENA13 lagging area population (%) Sparse Dense Sparse Dense Far from major cities 19 34 6j 33 j Close to major cities 11 36 5j 56 j Morocco: Spatial Concentration of Economic Activity Source: G-Econ project, Yale, 2010. Morocco: Lagging Area Typology Sources: Elevation, SRTM30 (2000); land cover, GLC2000 (2000); population, Landscan (2005); rail, VMAP0 (1997); rivers, CIA World Data Bank II (1980s); roads, Euro-med (2000) and VMAP0 (1997); slope, in degrees, SRTM30 (2000); urban areas, GPW3-GRUMP (2000); water bodies, Global Lakes and Wetlands Database layer 1 (2004). Appendix: MENA Region 323 Morocco: Spatial Integration Diamond accountability 1.0 0.5 spatial equality connectivity 0.0 in living standards spatial equality in human development country MENA 13 Sources: Accountability: World Bank World Governance Index 2008; spatial equality in living standards: Morocco National Household Survey 1998; spatial equality in human development: Santé en chiffres 2007; connectivity: author's calculation based on Nelson (2008). Sources in Appendix. Notes: a. Averages are weighted, whenever possible. b. Data unavailable for Islamic Republic of Iran and the West Bank and Gaza. c. Data unavailable for Djibouti, Tunisia, and the West Bank and Gaza. d. Data unavailable for Djibouti, Jordan, Tunisia, and the West Bank and Gaza. e. Data unavailable for Islamic Republic of Iran, Iraq, Libya, Syrian Arab Republic, and the Republic of Yemen. f. Data unavailable for Djibouti and the West Bank and Gaza. g. Data unavailable for Islamic Republic of Iraq. h. Data unavailable for Algeria, Libya, and the West Bank and Gaza. i. Data unavailable for Algeria, Lebanon, and Libya. j. Data unavailable for Djibouti. .. = not available; MICS = Multiple Indicator Cluster Survey. Because of missing data, the spatial concentration of economic activity maps are not available for Jordan, Lebanon, and Tunisia, and the West Bank and Gaza, and the lagging area map is not available for the Islamic Republic of Iran. Definition of Spatial Integration Diamond Indicators: Accountability: Voice and accountability index, reversed (percentile rank 0­100, 100 = best). Spatial equality in living standards: rural per capita consumption as percentage of urban per capita con- sumption (%). Spatial equality in human development: urban infant mortality as percentage of rural infant mortality (%). Connectivity: Travel time to city index (% of population). See indicator definitions for more details. 324 Poor Places, Thriving People SYRIAN ARAB REPUBLIC Country MENA13 a Source General Population (millions), 2008 21 334 WDI Database GDP per capita, PPP (constant 2005 international dollars), 2008 4,232 6,198b WDI Database GDP growth (average annual %), 1999­2008 3.5 4.7b WDI database Physical geography Surface area (1,000 km2), 2008 185 8,774 WDI Database Arable land (% of surface area), 2005 27 6 WDI Database Average precipitation (mm/year), 2008 252 137 FAO AQUASTAT Renewable water resources per capita (m3/inhab/year), 2008 791 1,014 FAO AQUASTAT Human geography Population growth (annual), 2008 3.5 1.9 WDI Database Labor force participation rate (% of population ages 15­64), 2007 51 52 WDI Database Population density (people/km2), 2008 116 38 WDI Database Population in cities > 1 million (% of total population), 2007 30 22c WDI Database Population in largest city (% of urban population), 2007 25 24d WDI Databse Agglomeration index (settlement size: 100,000) 57 63 Uchida and Nelson 2008 Gender parity index for net primary enrollment rate, 2007 .. 0.98e Structure of the economy Agriculture, value added (% of GDP), 2008 20 11b WDI Database Industry, value added (% of GDP), 2008 35 44b WDI Database Energy wealth (net energy exports/per capita), ktoe 0.21 1.28f IEA Governance Voice and accountability, 2008 (percentile rank 0­100, 100=best) 4.8 23.7 World Bank WGI Connectivity Paved roads (% of total road network), 2006 20 58 CIA Factbook Primary road density (km of paved roads/km2 of arable land), 2006 0.4 0.8 WDI Database/CIA Factbook Primary road density (km of paved roads/1,000 people), 2006 1 1.4 WDI Database/CIA Factbook Travel time to city index (%) 18 28 Author's calculations (source in Appendix) Rural access index (%), 2004 49 58 Roberts and orthers 2006 National average distance to capital city (km), 2000 289 436 World Bank 2009 Rail density (km of track/km2 of arable land), 2005 0.06 0.05 WDI/World Bank Railway Database Rail density (km of track/1,000 people), 2005 0.14 0.09 WDI/World Bank Railway Database Mobile cellular subscriptions (per 100 people), 2008 33 58g WDI Database Population covered by mobile cellular network (%), 2007 96 89 WDI Database Internet users (per 100 people), 2008 17 18 WDI Database Spatial Disparities Mean per capita consumption ratio (capital region/national), 2004 1.55 1.51h Syria National HH Survey 2004 Mean per capita consumption ratio (urban/rural), 2004 1.52 1.82i Syria National HH Survey 2004 Urban Rural Source Population growth (% annual), 2008 4.1 2.7 WDI Database Fertility rate (births per woman), 2007 3.4 4.4 MICS 2007 Poverty rate (based on national poverty line), 2004 7.9 14.5 El Laithy and Abu-Ismail 2005 Improved water source (% of population with access), 2006 94 80 MICS 2007 Improved sanitation (% of population with access), 2006 100 94 MICS 2007 Stunting (%), 2007 22 23 MICS 2007 Infant mortality rate, 2007 16 20 MICS 2007 Under-5 mortality rate, 2007 19 23 MICS 2007 Appendix: MENA Region 325 Geographic character of Country MENA13 lagging area population (%) Sparse Dense Sparse Dense Far from major cities 4 8 6j 33 j Close to major cities 8 80 5j 56 j Syrian Arab Republic: Spatial Concentration of Economic Activity Source: G-Econ project, Yale, 2010. Syrian Arab Republic: Lagging Area Typology Sources: Elevation, SRTM30 (2000); land cover, GLC2000 (2000); population, Landscan (2005); rail, VMAP0 (1997); rivers, CIA World Data Bank II (1980s); roads, Euro-med (2000) and VMAP0 (1997); slope, in degrees, SRTM30 (2000); urban areas, GPW3-GRUMP (2000); water bodies, Global Lakes and Wetlands Database layer 1 (2004). 326 Poor Places, Thriving People Syrian Arab Republic: Spatial Integration Diamond accountability 1.0 0.5 spatial equality connectivity 0.0 in living standards spatial equality in human development country MENA 13 Sources: Accountability: World Bank World Governance Index 2008; spatial equality in living standards: Syria National Household Survey 2004; spatial equality in human development: MICS 2007; connectivity: author's calculation based on Nelson (2008). Sources in Appendix. Notes: a. Averages are weighted, whenever possible. b. Data unavailable for Islamic Republic of Iran and the West Bank and Gaza. c. Data unavailable for Djibouti, Tunisia, and the West Bank and Gaza. d. Data unavailable for Djibouti, Jordan, Tunisia, and the West Bank and Gaza. e. Data unavailable for Islamic Republic of Iran, Iraq, Libya, Syrian Arab Republic, and the Republic of Yemen. f. Data unavailable for Djibouti and the West Bank and Gaza. g. Data unavailable for Islamic Republic of Iraq. h. Data unavailable for Algeria, Libya, and the West Bank and Gaza. i. Data unavailable for Algeria, Lebanon, and Libya. j. Data unavailable for Djibouti. .. = not available; MICS = Multiple Indicator Cluster Survey. Because of missing data, the spatial concentration of economic activity maps are not available for Jordan, Lebanon, and Tunisia, and the West Bank and Gaza, and the lagging area map is not available for the Islamic Republic of Iran. Definition of Spatial Integration Diamond Indicators: Accountability: Voice and accountability index, reversed (percentile rank 0­100, 100 = best). Spatial equality in living standards: rural per capita consumption as percentage of urban per capita con- sumption (%). Spatial equality in human development: urban infant mortality as percentage of rural infant mortality (%). Connectivity: Travel time to city index (% of population). See indicator definitions for more details. Appendix: MENA Region 327 TUNISIA Country MENA13a Source General Population (millions), 2008 10 334 WDI Database GDP per capita, PPP (constant 2005 international dollars), 2008 7,348 6,198b WDI Database GDP growth (average annual %), 1999­2008 5 4.7b WDI Database Physical geography Surface area (1,000 km2), 2008 163 8,774 WDI Database Arable land (% of surface area), 2005 18 6 WDI Database Average precipitation (mm/year), 2008 207 137 FAO AQUASTAT Renewable water resources per capita (m3/inhab/year), 2008 452 1,014 FAO AQUASTAT Human geography Population growth (annual), 2008 1 1.9 WDI Database Labor force participation rate (% of population ages 15­64), 2007 51 52 WDI Database Population density (people/km2), 2008 67 38 WDI Database Population in cities > 1 million (% of total population) .. 22c Population in largest city (% of urban population) .. 24d Agglomeration index (settlement size: 100,000) 39 63 Uchida and Nelson 2008 Gender parity index for net primary enrollment rate, 2007 1.01 0.98e UNESCO Structure of the economy Agriculture, value added (% of GDP), 2008 10 11b WDI Database Industry, value added (% of GDP), 2008 28 44b WDI Database Energy wealth (net energy exports/per capita), ktoe ­0.11 1.28f IEA Governance Voice and accountability, 2008 (percentile rank 0­100, 100 = best) 11.5 23.7 World Bank WGI Connectivity Paved roads (% of total road network), 2006 66 58 CIA Factbook Primary road density (km of paved roads/km2 of arable land), 2006 0.5 0.8 WDI Database/CIA Factbook Primary road density (km of paved roads/1,000 people), 2006 1.3 1.4 WDI Database/CIA Factbook Travel time to city index (%) 32 28 Author's calculations (source in Appendix) Rural access index (%), 2004 39 58 Roberts and others 2006 National average distance to capital city (km), 2000 335 436 World Bank 2009 Rail density (km of track/km2 of arable land), 2005 0.08 0.05 EU DESTIN Project Rail density (km of track/1,000 people), 2005 0.22 0.09 EU DESTIN Project Mobile cellular subscriptions (per 100 people), 2008 83 58g WDI Database Population covered by mobile cellular network (%), 2007 100 89 WDI Database Internet users (per 100 people), 2008 27 18 WDI Database Spatial disparities Mean per capita consumption ratio (capital region/national), 2000 1.33 1.51h INS 2000 Mean per capita consumption ratio (urban/rural), 2000 1.86 1.82 i INS 2000 Urban Rural Source Population growth (% annual), 2008 1.6 ­0.2 WDI Database Fertility rate (births per woman), 2006 2.5 3.3 World Bank 2006c Poverty rate (based on national poverty line), 2000 1.6 8.3 INS 2000 Improved water source (% of population with access), 2006 99 84 MICS3 2006 Improved sanitation (% of population with access), 2006 96 64 MICS3 2006 Stunting (%), 2006 4 10 MICS3 2006 Infant mortality rate, 2006 16 30 MICS3 2006 Under-5 mortality rate .. .. 328 Poor Places, Thriving People Geographic character of Country MENA13 lagging area population (%) Sparse Dense Sparse Dense Far from major cities 16 37 6j 33 j Close to major cities 13 34 5j 56 j Tunisia: Lagging Area Typology Sources: Elevation, SRTM30 (2000); land cover, GLC2000 (2000); population, Landscan (2005); rail, VMAP0 (1997); rivers, CIA World Data Bank II (1980s); roads, Euro-med (2000) and VMAP0 (1997); slope, in degrees, SRTM30 (2000); urban areas, GPW3-GRUMP (2000); water bodies, Global Lakes and Wetlands Database layer 1 (2004). Appendix: MENA Region 329 Tunisia: Spatial Integration Diamond accountability 1.0 0.5 spatial equality connectivity 0.0 in living standards spatial equality in human development country MENA 13 Sources: Accountability: World Bank World Governance Index 2008; spatial equality in living standards: INS 2000; spatial equality in human development: MICS3 2006; connectivity: author's calculation based on Nelson (2008). Sources in Appendix. Notes: a. Averages are weighted, whenever possible. b. Data unavailable for Islamic Republic of Iran and the West Bank and Gaza. c. Data unavailable for Djibouti, Tunisia, and the West Bank and Gaza. d. Data unavailable for Djibouti, Jordan, Tunisia, and the West Bank and Gaza. e. Data unavailable for Islamic Republic of Iran, Iraq, Libya, Syrian Arab Republic, and the Republic of Yemen. f. Data unavailable for Djibouti and the West Bank and Gaza. g. Data unavailable for Islamic Republic of Iraq. h. Data unavailable for Algeria, Libya, and the West Bank and Gaza. i. Data unavailable for Algeria, Lebanon, and Libya. j. Data unavailable for Djibouti. .. = not available; MICS = Multiple Indicator Cluster Survey. Because of missing data, the spatial concentration of economic activity maps are not available for Jordan, Lebanon, and Tunisia, and the West Bank and Gaza, and the lagging area map is not available for the Islamic Republic of Iran. Definition of Spatial Integration Diamond Indicators: Accountability: Voice and accountability index, reversed (percentile rank 0­100, 100 = best). Spatial equality in living standards: rural per capita consumption as percentage of urban per capita con- sumption (%). Spatial equality in human development: urban infant mortality as percentage of rural infant mortality (%). Connectivity: Travel time to city index (% of population). See indicator definitions for more details. 330 Poor Places, Thriving People WEST BANK AND GAZA Country MENA13a Source General Population (millions), 2008 4 334 WDI Database GDP per capita, PPP (2009 $), 2008 (estimate) 2,900 6,198b CIA Factbook GDP growth (average annual %), 1999­2008 .. 4.7b WDI Database Physical geography Surface area (1,000 km2), 2008 6 8,774 WDI Database Arable land (% of surface area), 2005 18 6 WDI Database Average precipitation (mm/year), 2008 402 137 FAO AQUASTAT Renewable water resources per capita (m3/inhab/year), 2008 202 1,014 FAO AQUASTAT Human geography Population growth (annual), 2008 3.4 1.9 WDI Database Labor force participation rate (% of population ages 15­64), 2007 43 52 WDI Database Population density (people/km2), 2008 638 38 WDI Database Population in cities > 1 million (% of total population) .. 22c Population in largest city (% of urban population) .. 24d Agglomeration index (settlement size: 100,000) 90 63 Uchida and Nelson 2008 Gender parity index for net primary enrollment rate, 2007 1 0.98e UNESCO Structure of the economy Agriculture, value added (% of GDP), 2008 (estimate) 5 11b CIA Factbook/WDI Database Industry, value added (% of GDP), 2008 (estimate) 14 44b CIA Factbook/WDI Database Energy wealth (net energy exports/per capita), ktoe .. 1.28f IEA Governance Voice and accountability, 2008 (percentile rank 0­100, 100 = best) 22.1 23.7 World Bank WGI Connectivity Paved roads (% of total road network), 2006 100 58 CIA Factbook Primary road density (km of paved roads/km2 of arable land), 2006 0.4 0.8 WDI Database/CIA Factbook Primary road density (km of paved roads/1,000 people), 2006 0.3 1.4 WDI Database/CIA Factbook Travel time to city Index (%) 2 28 Author's calculations (source in Appendix) Rural access index (%), 2004 .. 58 Roberts and others 2006 National average distance to capital city (km) .. 436 World Bank 2009 Rail density (km of track/km2 of arable land) 0 0.05 Rail density (km of track/1,000 people) 0 0.09 Mobile cellular subscriptions (per 100 people), 2007 28 58g WDI Database Population covered by mobile cellular network (%), 2007 95 89 WDI Database Internet users (per 100 people), 2007 10 18 WDI Database Spatial Disparities Mean per capita consumption ratio (capital region/national) .. 1.51h Mean per capita consumption ratio (urban/rural), 2005 1.18 1.82i PCBS 2006 Urban Rural Source Population growth (% annual), 2008 3.6 3.1 WDI Database Fertility rate (births per woman), 2006 4.4 4.4 Family Health Survey 2006 Poverty rate (based on national poverty line), 2003 16 15 World Bank 2003 Improved water source (% of population with access), 2006 91 90 Family Health Survey 2006 Improved sanitation (% of population with access), 2006 100 93 Family Health Survey 2006 Stunting (%), 2006 11 9 Family Health Survey 2006 Infant mortality rate, 2006 27 27 Family Health Survey 2006 Under-5 mortality rate, 2006 30 30 Family Health Survey 2006 Appendix: MENA Region 331 Geographic character of Country MENA13 lagging area population (%) Sparse Dense Sparse Dense Far from major cities 0 5 6j 33 j Close to major cities 2 93 5j 56 j West Bank and Gaza: Lagging Area Typology Sources: Elevation, SRTM30 (2000); land cover, GLC2000 (2000); population, Landscan (2005); rail, VMAP0 (1997); rivers, CIA World Data Bank II (1980s); roads, Euro-med (2000) and VMAP0 (1997); slope, in degrees, SRTM30 (2000); urban areas, GPW3-GRUMP (2000); water bodies, Global Lakes and Wetlands Database layer 1 (2004). 332 Poor Places, Thriving People West Bank and Gaza: Spatial Integration Diamond accountability 1.0 0.5 spatial equality connectivity 0.0 in living standards spatial equality in human development country MENA 13 Sources: Accountability: World Bank World Governance Index 2008; spatial equality in living standards: PCBS 2006; spatial equality in human development: Family Health Survey 2006; connectivity: author's calculation based on Nelson (2008). Sources in Appendix. Notes: a. Averages are weighted, whenever possible. b. Data unavailable for Islamic Republic of Iran and the West Bank and Gaza. c. Data unavailable for Djibouti, Tunisia, and the West Bank and Gaza. d. Data unavailable for Djibouti, Jordan, Tunisia, and the West Bank and Gaza. e. Data unavailable for Islamic Republic of Iran, Iraq, Libya, Syrian Arab Republic, and the Republic of Yemen. f. Data unavailable for Djibouti and the West Bank and Gaza. g. Data unavailable for Islamic Republic of Iraq. h. Data unavailable for Algeria, Libya, and the West Bank and Gaza. i. Data unavailable for Algeria, Lebanon, and Libya. j. Data unavailable for Djibouti. .. = not available; MICS = Multiple Indicator Cluster Survey. Because of missing data, the spatial concentration of economic activity maps are not available for Jordan, Lebanon, and Tunisia, and the West Bank and Gaza, and the lagging area map is not available for the Islamic Republic of Iran. Definition of Spatial Integration Diamond Indicators: Accountability: Voice and accountability index, reversed (percentile rank 0­100, 100 = best). Spatial equality in living standards: rural per capita consumption as percentage of urban per capita con- sumption (%). Spatial equality in human development: urban infant mortality as percentage of rural infant mortality (%). Connectivity: Travel time to city index (% of population). See indicator definitions for more details. Appendix: MENA Region 333 REPUBLIC OF YEMEN Country MENA13a Source General Population (millions), 2008 23 334 WDI Database GDP per capita, PPP (constant 2005 international dollars), 2008 2,232 6,198b WDI Database GDP growth (average annual %), 1999­2008 3.9 4.7b WDI database Physical geography Surface area (1,000 km2), 2008 528 8,774 WDI Database Arable land (% of surface area), 2005 3 6 WDI Database Average precipitation (mm/year), 2008 167 137 FAO AQUASTAT Renewable water resources per capita (m3/inhab/year), 2008 (estimate) 92 1,014 FAO AQUASTAT Human geography Population growth (annual), 2008 3 1.9 WDI Database Labor force participation rate (% of population ages 15­64), 2007 45 52 WDI Database Population density (people/km2), 2008 44 38 WDI Database Population in cities > 1 million (% of total population), 2007 9 22c WDI Database Population in largest city (% of urban population), 2007 30 24d WDI Databse Agglomeration index (settlement size: 100,000) 23 63 Uchida and Nelson 2008 Gender parity index for net primary enrollment rate .. 0.98e Structure of the economy Agriculture, value added (% of GDP), 2009 (estimate) 10 11b CIA Factbook/WDI Database Industry, value added (% of GDP), 2009 (estimate) 39 44b CIA Factbook/WDI Database Energy wealth (net energy exports/per capita), ktoe 0.38 1.28f IEA Governance Voice and accountability, 2008 (percentile rank 0­100, 100 = best) 14.9 23.7 World Bank WGI Connectivity Paved roads (% of total road network), 2005 9 58 CIA Factbook Primary road density (km of paved roads/km2 of arable land), 2005 0.4 0.8 WDI Database/CIA Factbook Primary road density (km of paved roads/1,000 people), 2005 0.3 1.4 WDI Database/CIA Factbook Travel time to city index (%) 31 28 Author's calculations (source in Appendix) Rural access index (%), 2004 21 58 Roberts and others 2006 National average distance to capital city (km), 2000 406 436 World Bank 2009 Rail density (km of track/km2 of arable land) 0 0.05 Rail density (km of track/1,000 people) 0 0.09 Mobile cellular subscriptions (per 100 people), 2006 14 58g WDI Database Population covered by mobile cellular network (%), 2006 68 89 WDI Database Internet users (per 100 people), 2007 1 18 WDI Database Spatial disparities Mean per capita consumption ratio (capital region/national), 2005 1.7 1.51h Yemen National HH Survey 2005 Mean per capita consumption ratio (urban/rural), 2005 1.67 1.82i Yemen National HH Survey 2005 Urban Rural Source Population growth (% annual), 2008 4.9 2.1 WDI Database Fertility rate (births per woman), 2002 4.5 6.7 Family Health Survey (PAPCHILD) Poverty rate (based on national poverty line), 1998 30.8 45 World Bank 2002 Improved water source (% of population with access), 2006 74 52 MICS 2006 Improved sanitation (% of population with access), 2006 92 34 MICS 2006 Stunting (%), 2006 56 57 MICS 2006 Infant mortality rate, 2006 55 73 MICS 2006 Under-5 mortality rate, 2006 57 86 MICS 2006 334 Poor Places, Thriving People Geographic character of Country MENA13 lagging area population (%) Sparse Dense Sparse Dense Far from major cities 11 24 6j 33 j Close to major cities 7 57 5j 56 j Republic of Yemen: Spatial Concentration of Economic Activity Source: G-Econ project, Yale, 2010. Republic of Yemen: Lagging Area Typology Sources: Elevation, SRTM30 (2000); land cover, GLC2000 (2000); population, Landscan (2005); rail, VMAP0 (1997); rivers, CIA World Data Bank II (1980s); roads, Euro-med (2000) and VMAP0 (1997); slope, in degrees, SRTM30 (2000); urban areas, GPW3-GRUMP (2000); water bodies, Global Lakes and Wetlands Database layer 1 (2004). Appendix: MENA Region 335 Republic of Yemen: Spatial Integration Diamond accountability 1.0 0.5 spatial equality connectivity 0.0 in living standards spatial equality in human development country MENA 13 Sources: Accountability: World Bank World Governance Index 2008; spatial equality in living standards: Yemen National Household Survey 2005; spatial equality in human development: MICS 2006; connectivity: author's calculation based on Nelson 2008. Sources in appendix. Notes: a. Averages are weighted, whenever possible. b. Data unavailable for Islamic Republic of Iran and the West Bank and Gaza. c. Data unavailable for Djibouti, Tunisia, and the West Bank and Gaza. d. Data unavailable for Djibouti, Jordan, Tunisia, and the West Bank and Gaza. e. Data unavailable for Islamic Republic of Iran, Iraq, Libya, Syrian Arab Republic, and the Republic of Yemen. f. Data unavailable for Djibouti and the West Bank and Gaza. g. Data unavailable for Islamic Republic of Iraq. h. Data unavailable for Algeria, Libya, and the West Bank and Gaza. i. Data unavailable for Algeria, Lebanon, and Libya. j. Data unavailable for Djibouti. .. = not available; MICS = Multiple Indicator Cluster Survey. Because of missing data, the spatial concentration of economic activity maps are not available for Jordan, Lebanon, and Tunisia, and the West Bank and Gaza, and the lagging area map is not available for the Islamic Republic of Iran. Definition of Spatial Integration Diamond Indicators: Accountability: Voice and accountability index, reversed (percentile rank 0­100, 100 = best). Spatial equality in living standards: rural per capita consumption as percentage of urban per capita con- sumption (%). Spatial equality in human development: urban infant mortality as percentage of rural infant mortality (%). Connectivity: Travel time to city index (% of population). See indicator definitions for more details. 336 Poor Places, Thriving People Indicator Definitions Indicator Source Notes Population WDI Database Total population is based on the de facto definition of population, which counts all residents regardless of legal status or citizenship, except for refugees not permanently settled in the country of asylum (generally considered part of the population of their country of origin). The values shown are midyear estimates. Source: United Nations Population Division. GDP per capita, PPP WDI Database GDP per capita is based on purchasing power parity (PPP). PPP GDP is gross domestic product converted to international dollars using purchasing power parity rates. An international dollar has the same purchasing power over GDP as the U.S. dollar has in the United States. GDP at purchaser's prices is the sum of gross value added by all resident producers in the economy plus any product taxes and minus any subsidies not included in the value of the products. It is calculated without making deductions for deprecia- tion of fabricated assets or for depletion and degradation of natural resources. Data are in constant 2005 international dollars. Source: World Bank. GDP growth WDI Database Annual percentage growth rate of GDP at market prices is based on constant local currency. Aggregates are based on constant 2000 U.S. dollars. GDP is the sum of gross value added by all resident producers in the economy plus any product taxes and minus any subsidies not included in the value of the products. It is calculated without making deductions for depreciation of fabricated assets or for depletion and degradation of natural resources. Source: World Bank. Surface area WDI Database A country's total area includes areas under inland bodies of water and some coastal waterways. Source: Food and Agriculture Organization. Arable land WDI Database Arable land includes 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 tem- porarily fallow. Land abandoned as a result of shifting cultivation is excluded. Source: Food and Agriculture Organization. Average precipitation FAO/AQUASTAT Long-term double average over space and time of the precipitation falling on the country in a year is expressed in depth (mm/year). Total renewable FAO/AQUASTAT Sum per capita of internal renewable water resources and external water resources/ actual renewable water resources take into consideration the per capita quantity of flow reserved to upstream and downstream countries through formal or informal agreements or treaties and possible reduction of external flow because of upstream water abstraction. It corresponds to the maximum theoretical yearly amount of water actually available for a country at a given moment. Source: Food and Agriculture Organization. Population growth WDI Database Annual percentage growth rate of total population is used. Labor Force WDI Database Labor force participation rate is the proportion of the population Participation Rate ages 15 and older that is economically active (all people who supply labor for the production of goods and services during a specified period.) Source: International Labour Organization. (continued on next page) Appendix: MENA Region 337 Indicator Source Notes Population density WDI Database Population density is midyear population divided by land area in square kilometers. Population is based on the de facto definition of population, which counts all residents regardless of legal status or citizenship-except for refugees not permanently settled in the country of asylum. Land area is a country's total area, excluding area under inland water bodies, national claims to continental shelf, and exclusive economic zones. In most cases the definition of inland water bodies includes major rivers and lakes. Sources: FAO and World Bank population estimates. Population in cities WDI Database Population in urban agglomerations of more than one million is the > 1 million percentage of a country's population living in metropolitan areas that in 2000 had a population of more than one million people. Source: United Nations, World Urbanization Prospects. Population in largest WDI Database Percentage of a country's urban population living in that country's city largest metropolitan area is used. Source: United Nations, World Urbanization Prospects. Agglomeration index Uchida and Nelson 2008 Percentage of population living in areas with a population density of at least 150 people/km2 that are no more than an hour's travel time from a city of 100,000 people is used. Gender parity index UNESCO Quotient of the number of females and the number of males for net primary enrolled in a primary education is used. Net enrollment ratio is the enrollment rate ratio of children of official school age based on the International Standard Classification of Education (1997) who are enrolled in school to the population of the corresponding official school age. Agriculture, value WDI Database Net output of the agricultural sector after adding up all outputs added and subtracting intermediate inputs is used. It is calculated with- out making deductions for depreciation of fabricated assets or depletion and degradation of natural resources. Agriculture includes forestry, hunting, and fishing, as well as cultivation of crops and livestock production. Source: World Bank and OECD. Industry, value added WDI Database Net output of the industrial sector after adding up all outputs and subtracting intermediate inputs is used. It is calculated without making deductions for depreciation of fabricated assets or deple- tion and degradation of natural resources. Industry comprises value added in mining, manufacturing, construction, electricity, water, and gas. Source: World Bank and OECD. Energy wealth IEA Database Annual difference between a nation's energy exports and energy imports is measured per capita. Voice and Worldwide Governance Extent to which a country's citizens are able to participate in accountability Indicators selecting their government is evaluated as well as freedom of expression, freedom of association, and free media. Paved roads CIA Factbook Roads surfaced with crushed stone (macadam) and hydrocarbon binder or bituminized agents, with concrete, or with cobblestones is measured in length. Source: International Roads Federation. Road density CIA Factbook Length of paved roads (km) (includes crushed stone and hydrocarbon binder or bituminized agents, with concrete, or with cobblestones) is measured per unit area of arable land (km2) or per thousand people. Travel time to city Authors' calculations Percentage of population living over 3 hours' road travel away from index a city of at least 0.5 million inhabitants is used. (continued on next page) 338 Poor Places, Thriving People Indicator Source Notes Rural access index Roberts et al. 2006 In practice, the RAI measures the number of rural people who live within two kilometers (typically equivalent to a walk of 20­25 minutes) of an all-season road as a proportion of the total rural population. An "all-season road" is a road that is motorable all year round by the prevailing means of rural transport (typically a pick-up or a truck that does not have four-wheel drive). National average World Bank 2009 Population weighted average distance to the capital city is used. distance to capital city Rail density WDI Database/ Length of railway route available for train service, irrespective of the Euromed number of parallel tracks, is measured per area of arable land (km2) or per thousand people. Mobile cellular WDI Database Subscriptions to a public mobile telephone service using cellular tech- subscriptions nology provide access to the public switched telephone network. Post-paid and prepaid subscriptions are included. Source: Inter- national Telecommunications Union and World Bank estimates. Population covered WDI Database Percentage of people that live in areas served by a mobile cellular by mobile cellular signal regardless of whether they use it is measured. Source: Inter- network national Telecommunications Union and World Bank estimates. Internet users WDI Database Number of people with access to the worldwide network is used. Source: International Telecommunications Union and World Bank estimates. Mean per capita Multiple sources (mainly Average consumption of goods and services per person is used. consumption national household surveys) Population growth WDI Database Annual population growth rate is used. Population is based on the de facto definition of population, which counts all residents regardless of legal status or citizenship, except for refugees not permanently settled in the country of asylum (generally consid- ered part of the population of the country of origin). Source: United Nations Population Division. Fertility rate Multiple sources (mainly Annual number of births per woman is used. national household surveys) Poverty rate Multiple sources Percentage of national population living below the national (mainly national poverty line is used. household surveys) Improved water Multiple sources Percentage of the population with reasonable access to an adequate source (mainly national amount of water from an improved source, such as a household household surveys) connection, public standpipe, borehole, protected well or spring, and rainwater collection is used. Unimproved sources include vendors, tanker trucks, and unprotected wells and springs. Reasonable access is defined as the availability of at least 20 liters per person a day from a source within one kilometer of the dwelling. Source: World Health Organization and United Nations Children's Fund. Improved sanitation Multiple sources Percentage of the population with at least adequate access to (mainly national excreta disposal facilities that can effectively prevent human, household surveys) animal, and insect contact with excreta is used. Improved facilities range from simple but protected pit latrines to flush toilets with a sewerage connection. For effectiveness, facilities must be correctly constructed and properly maintained. Source: World Health Organization and United Nations Children's Fund. (continued on next page) Appendix: MENA Region 339 Indicator Source Notes Stunting Multiple sources (mainly Percentage of children having a height (or length). national household more than two standard deviations below the median of surveys) the NCHS/WHO growth reference for their age is used. Source: World Health Organization. Infant mortality rate Multiple sources Number of infants dying before reaching one year of age is meas- (mainly national ured per 1,000 live births in a given year. Source: Inter-agency household surveys) Group for Child Mortality Estimation. Under-5 mortality Multiple sources Probability per 1,000 that a newborn baby will die before rate (mainly national reaching age five is used if subject to current age-specific mortality household surveys) rates. Source: Inter-agency Group for Child Mortality Estimation. References Abbasi-Shavazi, M. J., and P. McDonald. 2005. "The Fertility Decline in the Islamic Republic of Iran, 1972­2000." Asian Population Studies, 2 (3): 217­37. El Laithy, H., and K. Abu-Ismail. 2005. "Poverty in Syria: 1996­2004." Unpublished. Haddad, A. 1996. Poverty in Lebanon. Beirut: ESCWA. Mehryar, A. 2004. "Primary Health Care and the Rural Poor in the Islamic Republic of Iran." Paper presented at Scaling-Up Poverty Reduction: A Global Learning Process and Conference, Shanghai, May 25­27, 2004. Ministčre de la Santé. 2008, 2007. Santé en Chiffres (Health Statistics). Rabat, Morocco: Royaume du Maroc, Ministčre de la Santé, Roberts, P., K. C. Shyam, and C. Rastogi. 2006. "Rural Access Index: A Key Development Indicator." World Bank Transport Paper. Shlash, Amal. 2009. Iraq High Committee for Strategy Reduction Draft Strategy Review. Powerpoint presentation. Amman. Uchida, H., and A. Nelson. 2008. "Agglomeration Index: Towards a New Measure of Urban Concentration." World Bank. http://sitere sources.worldbank.org/INTWDR2009/Resources/4231006- 1204741572978/Hiro1.pdf. UNESCO. 2007. Gender Parity Index for Net Primary Enrolment Rate. Institute for Statistics. 340 Poor Places, Thriving People WHO-UNICEF. 2008. "Joint Monitoring Report for Water Supply and Sanitation." http://www.wssinfo.org/resources/documents.html?type= country_files. World Bank. 1999. Democratic and Popular Republic of Algeria; Growth, Employment and Poverty Reduction. Washington, DC: World Bank. ------. 2002. Algeria Low Income Housing Project. Mid-term Review Mis- sion Report. Washington, DC: World Bank. ------. 2004. Jordan: Poverty Assessment. Washington, DC: World Bank. ------. 2006a. Tunisia: Agricultural Policy Review. Washington, DC: World Bank. ------. 2006b. World Development Report 2006: Equity and Development. Washington, DC: World Bank. ------. 2007. Making the Most of Scarcity: Accountability for Better Water Management in the Middle East and North Africa. Washington, DC: World Bank. ------. 2009. World Development Report 2009: Reshaping Economic Geog- raphy. Washington DC: World Bank. Index Boxes, figures, notes, and tables are indicated by b, f, n, and t, respectively. A climate change likely to ABB (Activity-Based Budgeting) affect, 72­74 system, Zambia, 147 cluster approach applied to, 28, absolute numbers living in poverty, 262­68, 264t, 265t, 267t 75­78, 77f, 77t, 78t employment in MENA, 262­63 Activity-Based Budgeting (ABB) extension services, women's access system, Zambia, 147 to, 47 Ades, A., 133 high density in MENA, 67­68 AERI (Agricultural Exports and high-value agricultural products Rural Income) project, (fruits and vegetables), shift Egypt, 267­68 to, 264­66, 265t Africa. See Middle East and North land markets, 196­98, 197f Africa; Sub-Saharan Africa; productivity issues, 39, specific countries 263­64, 264t African Americans, internal migration rainfed agriculture project, the of, 106 Republic of Yemen, xviii Agenda for a Vital Countryside subsidies, 266­67, 267t (Netherlands, 2004), 260 supply chains, 266, 267 agglomeration. See also urban versus urbanization with delayed rural transition from, 4, 5f, 21, economic. See economic 68­74, 70f, 71f, 71t, 73f agglomeration Ahmadinejad, Mahmoud, 35 measurement of, 64­68, 64b, 65t, Akhtar, Shamshad, xviii 66f, 67­68m Algeria, 297­300 Agricultural Exports and Rural agriculture in, 70, 71, 72, 73, 263 Income (AERI) project, Egypt, area-based programs (program 267­68 mode) in, 283 agriculture. See also under specific character of lagging areas in, 118 countries colonial legacy in, 138 341 342 Poor Places, Thriving People Algeria (continued) Argentina cross-border trade, 228, 230 decentralization in, 290 deconcentration in, 285 needs-based spatial allocation of education in, 160b fiscal resources in, 282 energy subsidies in, 180 PMT in, 186 evaluating impact of spatially urbanization and productivity in, 39 targeted public expenditure Armenia, PMT in, 186 in, 250 asabiyyah, 45b fertility rates in, 95, 96m Asia. See East Asia and Pacific Region; fiscal incentives for lagging areas Eastern Europe and Central in, 252, 253b Asia; South Asia; specific housing in, 19, 22, 192 countries investment climate, spatial Association of Southeast Asian disparities in, 15f Nations (ASEAN), 228 land for business in, 195 Ataturk, Kemal, 136 measuring agglomeration in, 66 Australia migration within, 107 territorial approach to cluster multilevel analysis in, 124n15 development in, 262 planned growth poles in, 259 urbanization and productivity proximity advantage in, 214­15 in, 39 regional economic development averages, problem of, 75t programs in, 27, 256 spillovers, power of, 213 B subnational investment climate Bangladesh in, 142, 143f education of girls in, 165 territorial development planning subnational investment climate in, 30, 280 in, 141 transportation in, 24, 217, 221, Belgium, National Solidarity 223, 225, 226, 227b Measure, 281 types of lagging areas in, 10, 113 belt areas, 10­11, 112­13, 119, 213 urbanization process, Berlin method, 282 smoothing, 202b border crossings, 25, 228­30, aménagement du territoire (territorial 229m, 229t development planning), 29­30, BRAC, 165 280, 281b Brakman, S., 250 analysis of spatial disparity. See Brazil typology of spatial disparity decentralization in, 290 Appalachian Regional Commission divergence-convergence cycle, 99 (ARC), U.S., 250 electronic proximity in, 235 Arab League, 228 evaluating impact of spatially Arab Republic of Egypt. See Egypt, targeted public expenditure Arab Republic of in, 251 ARC (Appalachian Regional health care in, 174 Commission), U.S., 250 proximity to economic density area-based programs (program mode), in, 117 30, 283­84 spillover effects in, 216 Index 343 Britain. See United Kingdom colonialism, legacy of, 13­14, 135­39 broadband penetration, 233­34, 239f commercial banks in MENA, 144t bus rapid transport (BRT), 25, 231 conditional cash transfer programs business, land for, 194­96 (CCTs) educational enrollment, effect on, C 18­19, 158, 165 Caribbean. See Latin America and as SSNs, 177, 183­85, 183t, 184f Caribbean, and specific connectivity infrastructure, 3, 23­26, countries 213­45 CCTs. See conditional cash transfer border crossings and regional programs trade facilitation, 25, 228­30, Central Asia. See Eastern Europe and 229m, 229t Central Asia, and specific electronic proximity, 25­26, 26f, countries 54, 232­40, 234t, 236f, 238t, Chandra, A., 53 239f, 240b Child Development Project, the matching appropriate packages to Republic of Yemen, 175 types of lagging area, 3t, 11, Chile 29f, 56t, 57f, 213t electronic proximity, 240 out-migration encouraged by, 53 lagging area catch-ups in, 249 political demand for spatial National Innovation Council, 261 equity with productivity, PMT in, 186 achieving, 52­54 proximity to economic density proximity advantage in MENA, in, 117 214­16, 215m, 215t Regional Development Associations spillovers, power of, 23, 32, in, 293n2 213­16 transportation in, 224 transportation. See transportation urbanization and productivity in, 39 Constantine Plan (1955­63), 138 China convergence, 98­105 divergence-convergence cycle, 99 breaking link between economic lagging area catch-ups in, 249 agglomeration and spatial migration within, 106 disparity, 50­52, 51f subnational investment climate deceptive appearance of, 104­5 in, 141 demographic transitions cities. See entries at urban contributing to, 7 climate change, potential effects of, 4, education leading to, 51 22, 72­74, 199, 263 evaluating impact of spatially clusters targeted public expenditure development of, 26, 260­62 on, 250­51 green (agricultural), 28, 262­68, factors contributing to, 8­10, 9b 264t, 265t, 267t lagging areas, catch-ups in, Cohesion Policy, EU, 282 248­49 Colombia in living standards, 99 child health and infant mortality migration leading to, 105­12 rates, 173 private sector interest contributing education in, 18 to, 26 344 Poor Places, Thriving People Coparticipatión Federal de Impuestos, dependency ratios and spatial Argentina, 282 disparities, 6­7 Corporacion de Fomento, Chile, 249 deruralization, MENA's lack of, cost of living, 97­98, 98t 106­7, 108f Costa Rica DHS (Demographic and Health health care in, 20, 175 Survey), 78, 82 PMT in, 186 Disability-Adjusted Life Years urbanization process, (DALYs), 226 moothing, 201, 202 distant versus near areas, 10­11, 114 cultural explanations for lagging divergence-convergence cycle, 99 areas, 131, 132b Djibouti, 301­2 customs procedures, improvements agriculture, delayed transition in, 230 from, 72 Cyprus, measuring agglomeration cross-border trade, 230 in, 66 health care in, 170 Czech Republic, education and land for business in, 195 infrastructure in, 259 measuring agglomeration in, 65­66 poverty, measuring, 78­79, 80 D separation of spatial disparities into DALYs (Disability-Adjusted Life educational and demographic Years), 226 components, 84, 86, 87f decentralization, 30­31, 259­60, social characteristics of spatial 287­90, 288f disparities in, 7 deconcentration, 30, 283­84, transportation in, 217, 221, 225 285­87, 286­87b Dominican Republic, PEM in, 147 degradation and maintenance of road surfaces, 24, 220­21 E Demographic and Health Survey East Asia and Pacific Region. See also (DHS), 78, 82 specific countries demographics electronic proximity, 237 agglomeration measured by poverty, measuring, 81 population distribution, 65t subnational investment climate as component of spatial disparity, in, 144 82­83, 83t, 84­86, 85f Eastern Europe and Central Asia. See density of population, 10­11, also specific countries 114­18, 116t, 117t correlation between GDP and education and, 6­8, 82­83, 83t poverty levels in, 51 internal migration. See migration, electronic proximity, 237 internal economic agglomeration MENA, population growth in, breaking link between spatial 4, 5f, 72, 73f disparity and, 50­52, 51f relationship between spatial growth in social indicators of disparities and, 6­8 disparity alongside, 99 urbanization and urban slum MENA's political need for, 39­44, growth in MENA, population 41m, 42t, 43­44f growth driving, 191, 192f spatial disparity versus, 2, 36 Index 345 Economic and Social Fund for rural children, enrollment rates Development (ESFD), for, 16t, 153t Lebanon, 122­23 salary incentives, 160b economic efficiency test, 255­56 SDIs, 18, 161 economic geography spatial analysis of, 158 colonialism and, 14, 138 supply-side factors, 16 economies of agglomeration and, TIMSS, 153­55, 157f, 159 40, 141, 213, 275 transportation infrastructure of GCC versus rest of MENA, 58 and, 258­59 spillover effects, ability of lagging unequal education provision, spatial area to benefit from, 214 disparities attributable to, subnational investment climate 157­58, 157f versus, 142, 145 urban versus rural poverty and, Ecuador, PMT in, 186 76, 77t education, 16­19, 152­69. See also education effect (inability to acquire under specific countries human capital in lagging CCTs, use of, 18­19, 158, 165 areas), 97 completion rates for adults older Education for All Program (Sarva than 25, 153, 154f, 155f Shiksa Abhiyan), India, 282 convergence resulting from, 51 EEI (Egyptian Education Initiative), demand-side factors and 169, 235 interventions, 16, 17­18, Egypt, Arab Republic of, 303­5 158, 161­63 agglomeration versus spatial demographics and, 6­8, 83­84 disparity in, 2 fertility rates affected by, agriculture in, 70, 71, 72, 196­98, 164­65, 164f 197f, 263, 264, 267­68 gender disparities in, 7, 17, 18­19, annual Human Development 31­32, 153, 155, 158­59, Reports, 120b 163­65, 164f area-based programs (program growth in disparity with economic mode) in, 283­84 agglomeration, 99 character of lagging areas in, household characteristics, spatial 118, 119 disparities attributable to, community-identified local 155­56 infrastructures in, xviii ICT, use of, 168­69, 168b convergence and divergence inability to acquire human capital in, 103­4f in lagging areas (education convergence trends in, 9b effect), 97 cross-border trade, 230 infrastructure, 19, 166­69, 168b decentralization in, 288, 289 lower enrollment rates for rural deconcentration in, 285, 286­87 children in MENA, 16t demographics of, 6, 82­83, 83t, PISA, 155 84­86, 85f quality of, 18, 158, 159­61, economic agglomeration in, 41 160b education in, 17, 18, 85f, 86, 153, relationship between spatial 155, 161, 163, 164­65, 164f, disparities and, 6­8 167, 169 346 Poor Places, Thriving People Egypt, Arab Republic of (continued) social networks and migration electronic proximity, 25, 235, flows in, 9­10 237, 240 spatial component of inequality energy subsidies in, 180, 181 in, 6 fertility rates in, 94, 164­65, 164f spillovers, power of, 213 fiscal incentives for lagging areas in, subnational investment climate in, 252, 254 142, 143f gender differences in experience of targeting of SSNs in, 186, 188 spatial inequity, 46, 47­48 transportation in, 24, 217, 221, health care in, 19, 171­76 223, 226, 227b, 231 housing in, 192, 193 types of spatial disparity in, 11 housing, land for, 22 urban planning in, 200­201 investment climate, spatial urbanization in, 23 disparities in, 15f urbanization process, land for business in, 195 smoothing, 202b measuring agglomeration in, water supply and sanitation 64­65, 64b in, 175 migration within, 107, 108, 109­10, Egypt Labor Market Panel 110b, 111 Survey, 108 needs-based spatial allocation of Egyptian Education Initiative (EEI), fiscal resources in, 30, 282 169, 235 PEM in, 149t, 150f electronic proximity, 25­26, 26f, planned growth poles in, 259 54, 232­40, 234t, 236f, political demand for spatial 238t, 239f, 240b equity and productivity ELMO index, 80­81 in, 35 Empowerment Zones, U.S., 251 political organization in, energy subsidies, 52, 180­81, 140f, 279b 180t, 182m poverty England. See United Kingdom mapping, 121­22 Enterprise Zones, U.K., 251 measuring, 76, 77t, 79, 81­82 equalization transfers, 148 spatial concentration ERVET (Regional Institution of, 103­4f for Territorial Economic proximity advantage in, 215 Development), Italy, 260­61 public expenditure management ESFD (Economic and Social in, 16 Fund for Development), at regional, governorate, and Lebanon, 122­23 community levels, 120­23, Europe, Eastern. See Eastern 120b, 121t, 123t Europe and Central Asia, separation of spatial disparities and specific countries into educational and European colonialism, legacy of, demographic components, 13­14, 135­39, 204n2 84­86, 85f European Union (EU). See also social characteristics of spatial specific countries disparities in, 6, 7 agricultural exports from MENA social funds, 183 to, 266 Index 347 border crossings in, 228 differences in experience of spatial Cohesion Policy, 282 inequity by gender, 46­48, 47b evaluating impact of spatially educational disparities, 7, 17, targeted public expenditure 18­19, 31­32, 153, 155, in, 250 158­59, 163­65, 164f intraregional trade in, 228 health care, women's awareness lagging area catch-ups in, of, 20 248, 249 household characteristics, 83 LEADER Community Initiative, migration affected by, 10, 47­48, 28, 259­60 47b, 110­11 measuring agglomeration in, 66 public services for lagging PEM in, 148 areas, 46­47 road connectivity problems, transportation, 47, 225 EuroMed review of, 220t GENIE program, Morocco, 240 Geographic Information System F (GIS) technology, 114, 166­68, farming. See agriculture 168b, 214 fertility rates geographical targeting, 185­86 contributing to spatial disparities, Germany, evaluating impact of 6­7, 94­96, 95f, 96m spatially targeted public convergence trends affected by, 9 expenditure in, 251 education affecting, 164­65, 164f Girls' Secondary Education Project, fiscal resources, needs-based spatial the Republic of Yemen, 165 allocation of, 30 GIS (Geographic Information France System) technology, 114, colonial legacy of, 138­39, 204n2 166­68, 168b, 214 convergence in, 51 Glaeser, E. L., 133 Freedom House, 134 global warming, potential effects freight transport, 227b, 232 of, 4, 22, 72­74, 199, 263 fringe areas, 10, 112­13, 131 GRDP (gross regional domestic fruits and vegetables, agricultural shift product), 258 in MENA to, 264­66, 265t Great Britain. See United Kingdom Fund of National Solidarity (FSN), green (agricultural) clusters, 28, Tunisia, 185 262­68, 264t, 265t, 267t gross domestic income (GDI), G agricultural employment/ Gadgil formula, 282 productivity as percentage Gaza. See West Bank and Gaza of, 262 GCC (Gulf Cooperation Council) gross domestic product (GDP) countries, 58, 72, 180, 266 agglomeration and economic GDI (gross domestic income), growth, relationship agricultural employment/ between, 135 productivity as percentage agricultural employment of, 262 and, 69­70 GDP. See gross domestic product energy subsidies as percentage gender issues. See also fertility rates of, 52, 180, 181 348 Poor Places, Thriving People gross domestic product (GDP) Hong Kong SAR, China, urbanization (continued) process in, 201, 202­3 health services as percentage household characteristics of, 19 educational disparities attributable in lagging areas, 50b to, 155­56 needs-based spatial allocation of link between economic problems in fiscal resources based on, 282 lagging areas and, 6­8, 94­97 PEM and, 150 poverty and household welfare, spa- population growth and, 73 tial and nonspatial components poverty levels relative to, 38, 51 of, 77­82, 79b, 80t, 81t public expenditure management housing, land for, 21­22, 192­94, and, 16 193t, 194b rural road investments and, 53 Human Development Reports, subnational estimates of, 50, 59n4 annual, in Egypt, 120b gross regional domestic product (GRDP), 258 I GRUMP database, 205n24 Ibn Khaldun, 40, 45b Gulf Cooperation Council (GCC) ICT. See information and countries, 58, 72, 180, 266. See communications technology also specific countries IFPRI (International Food Policy Research Institute), 73 H IHSES (Iraq Household health care, 19­20, 169­76. See also Socioeconomic Survey), 76 fertility rates; water supply; IMCI (Integrated Management of and under specific countries Childhood Illnesses) program, child health and infant mortality Morocco, 20, 174 rates, 172­74 INDH (Initiative Nationale de convergence, deceptive appearance Développement Humain), of, 105 Morocco, 122, 183 demand-side constraints, 20, India 171, 175­76 divergence-convergence cycle, 99 growth in disparity with economic migration within, 106 agglomeration, 99 needs-based spatial allocation of ICT, use of, 176 fiscal resources in, 282 indicators of spatial disparities in, spillover effects in, 216 169, 1778­179t subnational investment climate mandatory pension and health in, 141­42 insurance plans, 181­82 indicator definitions, 336­39 policy makers, ideas for, 177b Indonesia, spillover effects quality of, 174­76 in, 213, 216 road safety and, 226 information and communications sanitation, 171, 173f, 175 technology (ICT) urban versus rural poverty in education, 168­69, 168b and, 76, 77t electronic proximity, 25­26, 26f, home market effect, 40 54, 232­40, 234t, 236f, 238t, Honduras, PMT in, 186 239f, 240b Index 349 in health care, 176 decentralization in, 288 planned growth poles and, 259 deconcentration in, 285 infrastructure economic agglomeration in, 43, 44 community-identified local education in, 155, 157, 159, 163 infrastructures in Sohag, electronic proximity, 25, 237 Egypt, xviii energy subsidies in, 180 educational, 19, 166­69, 168b fertility rates in, 94, 95f electronic proximity, 25­26, 26f fiscal incentives for lagging areas in, human capital and, 258­59 253b, 254, 255m poles of development, connecting health care in, 20, 172, 174, 175, 176 poor places to. See connectivity housing in, 22, 192, 193, 194b infrastructure lagging area catch-ups in, 248 transport. See transportation land for business in, 195 Initiative Nationale de migration within, 106, 107, 108 Développement Humain PEM in, 149t, 150­51, 151f (INDH), Morocco, 122, 183 political demand for spatial equity institutional arrangements. See public and productivity in, 35 institutions for spatial policies political organization, spatial Integrated Management of Childhood impact of, 139 Illnesses (IMCI) program, poverty, measuring, 77 Morocco, 20, 174 private sector interest, Integrated Sanitation and Sewerage underpinning, 26 and Infrastructure Project provincial and municipal political (ISSIP), Egypt, 175 organization in, 279b internal migration. See migration, proximity advantage in, 216 internal public expenditure management Internal Migration Survey Project, in, 16 Syria, 108 transportation in, 24, 25, 217, 221, International Food Policy Research 223, 225, 226, 227b Institute (IFPRI), 73 urbanization process in, 22­23, Internet. See information and 199­200, 201, 202b communications technology Iraq, 309­11 investment climate, spatial disparities agriculture in, 73, 263 in, 14­15, 15f, 141­45, 142b, CCTs, 184­85, 184f 143f, 144t, 145b character of lagging areas Iran, Islamic Republic of, 306­8 in, 118 agriculture in, 70, 71, 72, 73, 264 colonial legacy in, 138 area-based programs (program cross-border trade, 228 mode) in, 283 deconcentration in, 285 character of lagging areas in, 118 economic agglomeration in, convergence in, 99­101, 40­41, 41m 100f, 104­5 education in, 163 convergence of lagging with leading energy subsidies in, 180, areas in, 26 181, 182m convergence trends in, 9b fiscal incentives for lagging areas cross-border trade, 228, 230 in, 252, 253b 350 Poor Places, Thriving People Iraq (continued) cost of living in, 98 health care in, 170 cross-border trade, 230 land-labor ratios in, 69 decentralization in, 288 poverty, measuring, 76, 77, 77f deconcentration in, 285 proximity advantage in, 215 education in, 17, 153, 155, 156, PRSP for, 76 157, 159, 161, 167, 169 separation of spatial disparities into electronic proximity, 237 educational and demographic energy subsidies in, 181 components, 84 fiscal incentives for lagging areas in, transportation in, 24, 217, 221, 252, 253b, 254 226, 227b health care in, 19, 174 water supply in, 67m, 171, 172f housing in, 22, 192 Iraq Household Socioeconomic investment climate, spatial Survey (IHSES), 76 disparities in, 15f Ireland land for business in, 195 lagging area catch-ups in, 248, 249 measuring agglomeration in, 65, 66 transportation in, 224 PEM in, 149t, 151, 152f irrigation and productivity issues in poverty, measuring, 78­79, 80, 81 MENA, 263­64, 264t provincial and municipal political Islamic Law, on ownership of organization in, 279b undeveloped land, 194b public expenditure management Islamic Republic of Iran. See Iran, in, 16 Islamic Republic of separation of spatial disparities into Israel, exclusion from study, 58 educational and demographic ISSIP (Integrated Sanitation and components, 84, 87­88, 88f Sewerage and Infrastructure social characteristics of spatial Project), Egypt, 175 disparities in, 7 Italy subnational investment climate cluster development in, 260­61 in, 142, 143f connectivity infrastructure in, 53 targeting of SSNs in, 185 evaluating impact of spatially transportation in, 25, 225, 226, 227b targeted public expenditure types of lagging areas in, 10­11, 113 in, 250­51 urbanization process, Territorial Pacts Initiative, 261 smoothing, 202b Jordan Education Initiative (JEI), 169 J Jamaica, PMT in, 186 K Japan Kaufman, D., 135 divergence-convergence cycle, 99 Kemalist ideologies, 136 evaluating impact of spatially Korea, Republic of targeted public expenditure New Industrial Cities program, 261 in, 250 SEZs, 249 JEI (Jordan Education Initiative), 169 spillover effects in, 216 Jordan, 312­14 urbanization process, smoothing, agriculture in, 70, 72, 196, 263 201, 202 character of lagging areas in, 118 Kraay, A., 135 Index 351 L separation of spatial disparities into lagging areas. See also spatial educational and demographic disparities in MENA components, 84, 88­89, 89f border areas as, 228 social characteristics of spatial catch-ups in, 248­49 disparities in, 8 character of, 118­19, 118t, 119f transportation in, 217, 223, 224, characterization of, 63­64 225, 226, 227b cultural explanations for, 131, 132b leveling the playing field for defined, 50b development, 3, 11­23, poverty and, 74­82, 75t, 77f, 77t, 131­212 78t, 79b, 80t, 81t breaking link between spatial land-labor ratio and delayed transition productivity and spatial from agriculture in MENA, 4, disparity, 50­52, 51f 5f, 21, 68­74, 70f, 71f, 71t, 73f education, 16­19, 152­69. See also land markets, 21­22, 191­98, 193t, education 194b, 195t, 197f governance and urbanization, Lao People's Democratic Republic, 133­41, 133t, 134t, 140f, 140t spillover effects in, 216 health care, 19­20, 169­76. See also lateral redistribution toward leading health care provinces, lack of, 107­9, 109t importance of, 31­32 Latin America and Caribbean. See also investment climate, 14­15, 15f, specific countries 141­45, 142b, 143f, 144t, 145b electronic proximity, 237 matching appropriate packages to poverty, measuring, 81 types of lagging area, 3t, 11, subnational investment climate 29f, 56t, 57f, 131t in, 144 PEM, 15­16, 146­51, 148f, LEADER Community Initiative, 149t, 150­52f EU, 28, 259­60 policy package 3 (private sector Lebanon, 315­17 interest) and, 247 agriculture in, 72, 196, 263 political demand for spatial character of lagging areas in, 118 equity with productivity, colonial legacy in, 138 achieving, 50­52 cross-border trade, 228, 230 political history, uneven playing education in, 155 field due to, 12­14 electronic proximity, 25, 237 SSNs, 21­22, 177­89. See also social fiscal incentives for lagging areas safety nets in, 252, 253b urbanization process, smoothing, health care in, 19, 170, 171, 173 21­23, 189­203. See also housing in, 192 urbanization process, land for business in, 195 smoothing measuring agglomeration Libya, 318­20 in, 65, 66 agriculture in, 72, 263 migration in, 106, 112t character of lagging areas in, 118 poverty, measuring, 77, 81­82, cross-border trade, 228 122­23 economic agglomeration rural population growth in, 106 in, 40­41, 41m 352 Poor Places, Thriving People Libya (continued) Le Maroc Compétitif initiative, electronic proximity, 25, 237 Morocco, 261 fiscal incentives for lagging areas Mashreq, 118­19, 119f, 228 in, 252, 253b mass transit, 24­25, 224, 227b, 230­32 health care in, 170 Mastruzzi, M., 135 measuring agglomeration in, 65, 66 MAUP (Modifiable Areal Unit migration within, 107 Problem), 79b poverty, measuring, 75­76, 77 Mauritania, cross-border trade by, 228 proximity advantage in, 215 MENA. See Middle East and North transportation in, 24, 221 Africa, and specific economies literacy. See education Mengistae, T., 143­44 living standards, spatial disparities Mercado Común del Sur or Mercosur in, 3­11 (Southern Cone Common characteristics of, 36, 37t Market), 228 cost of living, 97­98, 98t metropolitanization, MENA's lack definition of lagging area and, 50b of, 106­7, 108f divergence-convergence cycle, 99 Mexico importance of understanding, 3­4 education in, 18, 165 percentage of inequality evaluating impact of spatially attributable to spatial targeted public expenditure disparity, 6 in, 250 resolution of, 8­10, 9b Micro-Regions Strategy, 261 social disparities, spatial disparities needs-based spatial allocation of as, 6­8 fiscal resources in, 282­83 type of spatial disparity in given PMT in, 186 region, discerning, 10­11 rural development policy in, 260 urbanization with delayed transition Middle East and North Africa from agriculture, 4, 5f (MENA). See also specific localization economies, 40 economies Logistics Performance Index (LPI), economies included in, 32n1, 58 228­30, 229t, 241n3 demographics of. See demographics Lower Egypt. See Egypt, Arab economic agglomeration, political Republic of need for, 39­44, 41m, 42t, LPI (Logistics Performance Index), 43­44f 228­30, 229t, 241n3 internal migration in, 7. See also migration, internal M political history of, 12­14 Madagascar, PEM in, 147 spatial disparities in. See spatial Maghreb, 118­19, 228 disparities in MENA maintenance of road surfaces, 24, Middle East and North Africa 220­21 Development Report (World Malta, exclusion from study, 58 Bank, 2010), 1 Management and Planning migration, internal, 8­10, 22 Organisation (MPO), Islamic convergence, leading to, 105­12 Republic of Iran, 285 gender issues, 10, 47­48, 47b, market efficiency in ICT, 25­26 110­11 Index 353 improved connectivity infrastruc- housing in, 22, 192, 193 ture encouraging, 53 investment climate, spatial lateral redistribution toward leading disparities in, 15f provinces, lack of, 107­9, 109t lagging area catch-ups in, 248 metropolitanization and land for business in, 195 deruralization, lack of, measuring agglomeration 106­7, 108f in, 65­66 public services, providing, 198­98 migration within, 107 remittances, 110, 111f needs-based spatial allocation social networks and, 9­10, 109­10, of fiscal resources in, 110b, 111f 30, 282 sorting effect, 6­7, 96­97 political demand for spatial equity Millennium Development Goals, and productivity in, 35 19, 166 poverty, measuring, 76, 77t, 79b, mobile phones, 233, 236, 237, 239f 81, 82, 122 Modifiable Areal Unit Problem private sector interest, (MAUP), 79b underpinning, 26, 28 Mohammed VI (king of Morocco), provincial and municipal political 35, 280 organization in, 279b Morocco, 321­23 proximity advantage in, 214 agglomeration versus spatial separation of spatial disparities disparity in, 2 into educational and demo- agriculture in, 72, 73, 262, graphic components, 84, 263, 264, 268 89­90, 90f, 91f CCTs, 184 social characteristics of spatial character of lagging areas in, 118 disparities in, 8 cluster development in, 261 social funds, 183 colonial legacy in, 138­39 sorting effect in, 96 convergence of nonleading and spatial component on inequality leading areas in, 26 in, 6 cost of living in, 98 subnational investment climate in, cross-border trade, 228, 230 142­44, 143f deconcentration in, 285, 286­87b targeting of SSNs in, 185­86 economic agglomeration territorial development planning in, 42­44, 43f, 44f in, 30, 280 education in, 16­17, 18, 153, 155, transportation in, 25, 222­23, 156, 158, 159, 161, 163, 225, 227b, 231 167­69 urbanization process, smoothing, electronic proximity, 237, 240 202b fertility rates in, 94­95 water supply in, 171, 172f fiscal incentives for lagging areas in, Mozambique, PEM in, 147 252­54, 253b MPO (Management and Planning gender differences in experience of Organisation), Islamic spatial inequity, 46 Republic of Iran, 285 health care in, 19, 20, 171, 172, multilevel analysis of spatial disparity, 173, 174 120­23, 120b, 121t, 123t 354 Poor Places, Thriving People N NRENs (National Research and NAFTA (North American Free Trade Education Networks), 235 Agreement), 228 NRRP (National Rural Roads NAJM customs system, 230 Program), Morocco, 222­23 National Aid Fund (NAF), Jordan, 185 O National Household Survey OECD. See Organisation for (NHHS), 78 Economic Co-operation National Human Development and Development Initiative (INDH), Morocco, Office for Forestry and Livestock 122, 183 Development in the northwest National Innovation Council, (ODESYPANO), Tunisia, 284 Chile, 261 Offices Régionales des Mise en National Research and Education Valeur Agricole (ORMVAs), Networks (NRENs), 235 Morocco, 268 National Rural Roads Program 1,000 Villages program, Egypt, (NRRP), Morocco, 222­23 183, 282 National School Development Plan Organisation for Economic (NDSP), Jordan, 161 Co-operation and National Solidarity Measure, Development (OECD). Belgium, 281 See also specific countries Nazif, Ahmed, 35 agricultural policy, 266 NDSP (National School on evaluating impact of Development Plan), spatially targeted public Jordan, 161 expenditure, 250 near versus distant areas, housing, 192 10­11, 114 PEM, 146, 148 needs-based spatial allocation regional economic development of fiscal resources, 30, programs, new paradigms 281­83 for, 258 Netherlands, Agenda for a Vital ORMVAs (Offices Régionales des Countryside (2004), 260 Mise en Valeur Agricole), New Industrial Cities program, Morocco, 268 Korea, 261 Ottoman empire, legacy of, 13­14, New Zealand, urbanization and 135­39, 204n2 productivity in, 39 NHHS (National Household P Survey), 78 Pacific Region. See East Asia and Nicaragua Pacific Region, and specific education in, 18 countries PMT in, 186 Pakistan North Africa. See Middle East and proximity to economic density North Africa, and specific in, 117 countries subnational investment climate North American Free Trade in, 141 Agreement (NAFTA), 228 transportation in, 224 Index 355 Palestinian territories. See West Bank sociopolitical phenomena arising and Gaza from spatial disparity, 38­39 PEM (public expenditure spatialization, political tendency management), 15­16, 146­51, toward, 38­39, 82 148f, 149t, 150­52f stresses created by spatial pension plans, mandatory, 181­82 disparities, 44­46, 45b Peru, child health and infant mortality uneven playing field resulting rates in, 173 from political history of Philippines, child health and infant MENA, 12­14 mortality rates in, 173 Political Instability Task Force, 134 PISA (Program for International political organization Student Assessment), 155 of MENA states and economies, Plan Maroc Vert, Morocco, 268 278­79, 278f, 279b playing field, leveling. See leveling the provincial and local administrations playing field for development in MENA, 279b PMT (proxy means testing), spatial impact of, 133­41, 133t, 186, 188 134t, 140f, 140t pocket areas, 11, 112­13, 119 population. See demographics point of sale (POS) devices, 235 Population and Family Health poles of development, connecting Survey, 156 poor places to. See connectivity POS (point of sale) devices, 235 infrastructure poverty, 74­82 policy axioms for spatial development, absolute numbers, focusing on, 31­32, 292­93 75­78, 77f, 77t, 78t policy packages for dealing with averages, problem of, 75t spatial disparities, 3t, 10­11, GDP, levels relative to, 38, 51 29f, 49­50, 56t, 57f. See also growth in disparity with economic connectivity infrastructure; agglomeration, 99 leveling the playing field for household welfare, spatial and development; private sector nonspatial components of, interest in nonleading areas, 77­82, 79b, 80t, 81t underpinning urban versus rural, 76­77, 77t, 78t political demand for spatial equity poverty mapping, 121­23, 185­86 with productivity, 1­3, 35­62 Poverty Reduction Strategy Papers characteristics of spatial disparities (PRSPs), 76 in MENA and, 36, 37t private sector interest in nonleading connectivity infrastructure, 52­54 areas, underpinning, 3, 26­28, economic agglomeration, need for, 247­73 39­44, 41m, 42t, 43­44f cluster development, 26, 260­62 gender differences in experience of green (agricultural) clusters, 28, spatial inequity, 46­48, 47b 262­68, 264t, 265t, 267t leveling the playing field for lagging area catch-ups, government development, 50­52 role in facilitating, 248­49 matching policy packages and, 49­50 matching appropriate packages to private sector interest in nonleading types of lagging area, 3t, 11, areas, underpinning, 52­54 29f, 56t, 57f, 247t 356 Poor Places, Thriving People private sector interest in nonleading public institutions for spatial policies, areas, underpinning (continued) 29­31, 275­95 policy package 1 (leveling the area-based programs (program playing field) and, 247 mode), 30, 283­84 political demand for spatial decentralization, 30­31, 259­60, equity with productivity, 287­90, 288f achieving, 54­55 deconcentration, 30, 283­84, regional economic development 285­87, 286­87b programs, new paradigms importance of, 31, 275­77 for, 27­28, 27f, 32, 54­55, needs-based spatial allocation of 258­62, 258t fiscal resources, 30, 281­83 spatially targeted public policy makers, ideas for, 291b expenditure political organization of MENA difficulties in evaluating, 250­51 states and economies, 278­79, lack of success of, 254­58, 257m 278f, 279b widespread use in MENA, territorial development planning, 251­54, 252t, 253­54b 29­30, 280, 281b tax breaks, 252, 253­54b urbanization process, smoothing, productivity, agricultural, 39, 23, 201 263­64, 264t public-private partnerships. productivity, demand for spatial equity See private sector interest with. See political demand for in nonleading areas, spatial equity with productivity underpinning Program for International Student public services Assessment (PISA), 155 gender issues, 46­47 program mode (area-based programs), needs-based spatial allocation of 30, 283­84 fiscal resources for, 282 proximity advantage in MENA, quality versus quantity of facilities, 214­16, 215m, 215t 32, 174­76 proximity to economic density, urbanization process, smoothing, 117­18, 117t 198­201 proxy means testing (PMT), 186, 188 public transportation, 24­25, 224, PRSPs (Poverty Reduction Strategy 227b, 230­32 Papers), 76 public expenditure management Q (PEM), 15­16, 146­51, 148f, quality 149t, 150­52f of education, 18, 158, 159­61, 160b public expenditure, spatially targeted of health care, 174­76 difficulties in evaluating, 250­51 public service facilities, quality lack of success of, 254­58, 257m versus quantity of, 32, 174­76 regional economic development programs, new paradigms R for, 27­28, 27f, 32, 54­55, regional economic development 258­62, 258t programs, new paradigms widespread use in MENA, 251­54, for, 27­28, 27f, 32, 54­55, 252t, 253­54b 258­62, 258t Index 357 Regional Institution for Territorial SEZs (special economic zones), 249, Economic Development 251, 254 (ERVET), Italy, 260­61 short message service (SMS) regional trade facilitation, 25, 228­30, systems, 235 229m, 229t Shrouk program, Egypt, 283­84 remittances, 110, 111f Sidi Abdallah Technopole, Reproductive and Child Health Algeria, 259 Program, India, 282 SIGAD customs system, 230 reproductive health. See fertility rates SINDA customs system, 230 Republic of Korea. See Korea, Singapore, health care in, 20, 175 Republic of Smart Village Technoparks, Egypt, Republic of Yemen. See Yemen, the 240, 259 Republic of SMS (short message service) Reshaping Economic Geography (World systems, 235 Development Report, 2009), Social Change Agents program, xvii, 1, 3, 49, 106 Egypt, 175­76 response test, 255 social disparities, spatial disparities roads. See transportation as, 6­8 Romania, transportation in, 224 social networks and migration flows, rural areas. See urban versus rural 9­10, 109­10, 110b Russian Federation Social Priorities Program Basic divergence-convergence cycle, 99 Health Project, Morocco, 174 evaluating impact of spatially social safety nets (SSNs), 21­22, targeted public expenditure in 177­89 Soviet Union, 251 CCTs as, 177, 183­85, 183t, 184f PMT in, 186 commodity subsidies, 180­81, 180t, 181f, 182m S mandatory pension and health safety, transportation, 25, 225­26f insurance plans, 181­82 salary incentives, 160b PMT, 186, 188 sanitation, 171, 173f, 175 policy makers, ideas for, 190b Sapir Report, EU, 148, 248 social funds, 182­83, 183t Sarva Shiksa Abhiyan (Education for targeting, 185­88, 187m, 189m All Program), India, 282 types utilized in MENA, 180t Saudi Arabia social services. See public services gender differences in experience of sorting effect, 6­7, 96­97. See also spatial inequity, 46 migration, internal transportation and gender in, 47 South Africa, transportation in, 221 School Development Initiatives South America. See Latin America (SDIs), 18, 161 and Caribbean, and specific School Quality Initiative, Morocco, 161 countries schools generally. See education South Asia. See also specific countries SDIs (School Development electronic proximity, 237 Initiatives), 18, 161 subnational investment climate sectoral versus territorial approach to in, 144 cluster development, 261­62 transportation in, 24, 222 358 Poor Places, Thriving People Southern Cone Common Market public institutions for dealing with, (Mercado Común del Sur or 29­31, 275­95. See also public Mercosur), 228 institutions for spatial policies Soviet Union, evaluating impact self-resolution of, 8­10, 9b of spatially targeted public as social disparities, 6­8 expenditure in, 251. See also spillovers, power of, 23, 32 Russian Federation typology of, 10­11, 63­129. See also Spain typology of spatial disparity agricultural productivity spatially targeted public expenditure. in, 21, 69, 71t See public expenditure, colonial legacy of, 138­39 spatially targeted divergence-convergence cycle, 99 special economic zones (SEZs), 249, spatial convergence. See convergence 251, 254 spatial disparities in MENA, spillovers, power of, 23, 32, 213­16. xvii­xviii, 1­33 See also connectivity agglomeration versus, 2 infrastructure connectivity infrastructure, as Sri Lanka policy package for, 3, 23­26, health care in, 20, 175 213­45. See also connectivity PMT in, 186 infrastructure SSNs. See social safety nets leveling the playing field, as policy Sub-Saharan Africa. See also specific package for, 3, 11­23, countries 131­212. See also leveling the cross-border trade, 230 playing field for development electronic proximity, 237 living standards as measure of, land for business in, 195 3­11. See also living standards, poverty, measuring, 81 spatial disparities in subnational investment climate methodological approach in, 144 to, 57­58 transportation in, 24, 221 percentage of inequality subnational investment climate, attributable to, 6, 38­39, 14­15, 15f, 141­45, 142b, 38f, 82, 98 143f, 144t, 145b policy axioms aimed at leveling, subsidies 31­32, 292­93 agricultural policy and, 266­67, 267t policy packages for dealing with, 3t, commodity subsidies as SSNs in 10­11, 29f, 49­50, 56t, 57f MENA, 180­81, 180t, 181f political demands regarding, 1­3, energy, 52, 180­81, 180t, 182m 35­62. See also political housing, 193 demand for spatial equity with production and price based, 28, 32 productivity supply chains, agricultural, 266, 267 poverty distribution and, 38­39, 38f sustainability test, 255 private sector interest, underpin- Sweden, divergence-convergence ning, as policy package for, 3, cycle in, 99 26­28, 247­73. See also private Switzerland, needs-based spatial sector interest in nonleading allocation of fiscal resources areas, underpinning in, 281­82 Index 359 Syrian Arab Republic, 324­26 TIMSS (Trends in International agriculture in, 70, 71, 72, 74, Mathematics and Science 263, 264, 266, 267 Study), 153­55, 157f, 159 CCTs, 184 TIR (Transports Internationaux character of lagging areas in, 118 Routiers), 232 colonial legacy in, 138 TME (total managed cross-border trade, 228, 230 expenditure), 147 electronic proximity, 25, 237 Tobler, Waldo, 213 energy subsidies in, 180 total fertility rates (TFR), 164­65 fiscal incentives for lagging areas total managed expenditure in, 254b (TME), 147 land for business in, 195 trade facilitation, regional, 25, measuring agglomeration 228­30, 229m, 229t in, 65­66 Transport and Trade Facilitation migration within, 107, 108­9 Audit (TTFA), 231 poverty, measuring, 75, 79, 80, 81 transportation, 23­25, 52­54, regional economic development 217­26 programs in, 27, 256­57 border crossings and regional trade separation of spatial disparities into facilitation, 25, 228­30, 229m, educational and demographic 229t components, 84, 90­93, 92f degradation and maintenance of social characteristics of spatial road surfaces, 24, 220­21 disparities in, 8 education and road infrastructure, sorting effect in, 96­97 258­59 targeting of SSNs in, 185, 186­88, efficiency of road use, 25, 226, 227b 187m freight transport, 227b, 232 transportation in, 222, 224 gender issues regarding, 47, 225 types of lagging areas importance of, 52­53 in, 11, 112­13 improving spillover connectivity, 230­32, 233b T LPI, 228­30, 229t, 241n3 Tanzania, health care in, 174 main connectivity problems in targeting SSNs, 185­88, 187m, 189m MENA, 220t tax breaks for lagging areas in MENA, migration, improved connectivity 252, 253­54b encouraging, 53 Telecommunications Development paved roads in MENA, 217, Fund, Chile, 240 218­20, 218f, 219f territorial development planning, percentage of MENA population 29­30, 280, 281b within 3 hours of urban area, territorial versus sectoral approach to 217, 219t cluster development, 261­62 public, 24­25, 224, 227b, 230­32 TFR (total fertility rates), 164­65 regulation of, 226, 227b, 232 Thailand, proximity to economic rural connectivity deficits, 24f, density in, 118 221­23, 222f Theil index, 78­80 safety, 25, 225­26f Thompson, F., 143­44 urban congestion, 24­25, 223­25 360 Poor Places, Thriving People Transports Internationaux Routiers subnational investment climate (TIR), 232 in, 142b Trends in International Mathematics targeting of SSNs in, 185 and Science Study (TIMSS), territorial development planning 153­55, 157f, 159 in, 30, 280, 281b TTFA (Transport and Trade transportation in, 222, 226, 227b Facilitation Audit), 231 urbanization process, Tunisia, 327­29 smoothing, 202b Agency for Rural Development in ZDRs (Zones de Développement northwest of, xviii Régional) in, 257m, 257­58 agglomeration versus spatial Turkey disparity in, 2 cross-border trade, 228 agriculture in, 70, 71, 72, 196, PMT in, 186 266, 267, 268 transportation in, 224 area-based programs (program typology of spatial disparity, mode) in, 284 10­11, 63­129 character of lagging areas agglomeration, measuring, 64­68, in, 118 64b, 65t, 66f, 67­68m colonial legacy in, 138 belt areas, 10­11, 112­13, 119, 213 convergence in, 9b, 101­3, 102f character of lagging areas, 118­20, cross-border trade, 228, 230 118t, 119f decentralization in, 288 characterization of lagging areas, deconcentration in, 285, 286 63­64, 112­13 economic agglomeration in, 41­42, convergence, 98­105 42t, 43f, 44 cost of living, 97­98, 98t education in, 155, 163 density of population and, 10­11, electronic proximity, 235 114­18, 116t, 117t energy subsidies in, 180 economic opportunity and human evaluating impact of spatially development influences, 82­94 targeted public expenditure country case studies, 84­94, 85f, in, 251 87­93f fertility rates in, 94, 96 demographics, role of, 82­83, 83t fiscal incentives for lagging areas education, role of, 83­84 in, 252, 253b, 254 methods for distinguishing, 84 gender differences in fertility effect, 6­7, 94­96, experience of spatial 95f, 96m inequity, 46, 47b, 48 fringe areas, 10, 112­13, 131 health care in, 19, 170­71, 174 inability to acquire human capital in housing in, 22, 192, 193, 194b lagging areas (education lagging area catch-ups in, 248 effect), 97 land for business in, 195 lagging versus nonlagging areas, migration within, 111 10­11, 115 PEM in, 147, 148f, 149t link between household provincial and municipal political characteristics and organization in, 279b economic problems in proximity advantage in, 215 lagging areas, 94­97 Index 361 low- and middle-income countries, health care in, 20, 175 MENA patterns typical of, PEM in, 147 36, 37t United Nations Development matching policy packages to specific Programme (UNDP), 120b disparity patterns, 3t, 10­11, United Nations World Food 29f, 49­50, 56t, 57f, 291t Programme (WFP), 235­36 MENA map of distribution of, United States 12­13m, 114­15m convergence in, 51 multilevel analysis of, 120­23, 120b, divergence-convergence 121t, 123t cycle, 99 near versus distant areas, education in, 161 10­11, 114 evaluating impact of spatially pocket areas, 11, 112­13, 119 targeted public expenditure poverty, measuring. See poverty in, 250, 251 proximity to economic density, internal migration of African 117­18, 117t Americans within, 106 sorting effect, 96­97 lagging area catch-ups in, 248 lateral redistribution toward leading U provinces in, 107 Uganda needs-based spatial allocation of evaluating impact of spatially fiscal resources in, 282 targeted public expenditure Upper Egypt. See Egypt, Arab in, 250 Republic of health care in, 174 Urban Land Organization (ULO), Ukraine, child health and infant Islamic Republic of mortality rates in, 173 Iran, 194b ULO (Urban Land Organization), Urban Upgrading and Renovation Islamic Republic of Iran, 194b Agency, Tunisia, 194b UN-WIDER, 39 urban versus rural underpinning private sector interest. agriculture, delayed transition See private sector interest from, 4, 5f, 21, 68­74, 70f, 71f, in nonleading areas, 71t, 73f underpinning convergence trends and, UNDP (United Nations Develop- 8­10, 9b ment Programme), 120b cost of living, 97­98, 98t uneven playing field. See leveling the economic importance of playing field for development urbanization process, United Arab Emirates, measuring 39­44, 41m, 42t, 43­44f agglomeration in, 65 governance and urbanization, United Kingdom 133­41, 133t, 134t, colonial legacy of, 204n2 140f, 140t convergence in, 51, 105 lateral redistribution toward divergence-convergence cycle, 99 leading provinces, lack of, evaluating impact of spatially 107­9, 109t targeted public expenditure measuring agglomeration, 64­68, in, 251 64b, 65t, 66f, 67­68m 362 Poor Places, Thriving People urban versus rural (continued) diversion of water, programs for, 28 metropolitanization and settlement patterns shaped by, deruralization, lack of, 67­68m 106­7, 108f spatially adapted delivery systems migration from rural to urban for, 175 areas. See migration, internal Water Users' Associations poverty, measuring, 76­77, (WUAs), 264 77t, 78t West Bank and Gaza, 330­32 public institutions for spatial agriculture in, 71, 72, 196 policies, 23 CCTs, 184 transportation character of lagging areas in, 118 congestion in urban areas, decentralization in, 288, 289 24­25, 223­25 electronic proximity, 236 rural connectivity deficits, 24f, fiscal incentives for lagging areas in, 221­23, 222f 252, 254b urbanization economies, 40 health care in, 20, 176 urbanization process, smoothing, investment climate, spatial 21­23, 189­203 disparities in, 15f growth of urban areas and urban PEM in, 149t slums in MENA, 190­91, provincial and municipal political 191t, 192f organization in, 279b initiatives in MENA region, 201, subnational investment climate in, 203t 142, 143f land markets, 21­22, 191­98, 193t, transportation in, 24, 217, 221 194b, 195t, 197f Western Europe. See European Union policy makers, ideas for, 201, 202b (EU), and specific countries population growth, urbanization, WFP (World Food Programme), and slum growth in MENA, UN, 235­36 191, 192f WHO (World Health public institutions for, 23, 201 Organization), 170 public services and urban planning, World Bank 198­201 on educational enrollment advances Uruguay, urbanization and in MENA, 157­58 productivity in, 39 LPI, 228­30, 229t, 241n3 MENA projects, involvement in, V xviii Vietnam Middle East and North Africa divergence-convergence cycle, 99 Development Report (2010), 1 transportation in, 52­53 political organization, urbanization in, 40 agglomeration index used to assess spatial impact of, 135 W Reshaping Economic Geography water supply (World Development Report, access to, 171, 172f 2009), xvii, 1, 3, 49, 106 agricultural productivity and, road connectivity problems, review 263­64, 264t of, 220t Index 363 World Economic Forum, 169 measuring agglomeration in, 66, 68 World Food Programme (WFP), migration within, 107, 110, 111m UN, 235­36 PEM in, 149­50, 149t World Health Organization poverty, measuring, 76, 77t, 79, (WHO), 170 80, 81 World Values Survey (WVS), 132b provincial and municipal political World War II, 55 organization in, 279b WUAs (Water Users' proximity advantage in, 215­16 Associations), 264 rainfed agriculture project, xviii WVS (World Values Survey), 132b separation of spatial disparities into educational and demographic Y components, 84, 93f Yemen, the Republic of, 333­35 social characteristics of spatial agriculture in, 71, 72, 196, 263 disparities in, 8 CCTs, 184 spatial component on inequality character of lagging areas in, 118 in, 6 cross-border trade, 230 targeting of SSNs in, 186, 188, decentralization in, 288 189m education in, 17, 18, 159­60, 161, transportation in, 222, 223 165, 166 water supply and sanitation in, 171, fertility rates in, 95 173f, 175 fiscal incentives for lagging areas in, 253b Z health care in, 170, 171, 173, 175 Zambia, ABB system for PEM housing in, 22, 192, 193 in, 147 land for business in, 195 Zimbabwe, PMT in, 186 Eco-Audit Environmental Benefits Statement The World Bank is committed to preserv- Saved: ing endangered forests and natural resources. The Office of the Publisher has · 14 trees chosen to print Poor Places, Thriving People: · 384 lbs. of solid waste How the Middle East and North Africa Can · 6,325 gallons of water Rise Above Spatial Disparities on recycled · 1,313 lbs. of CO2 paper including 50% post-consumer recy- equivalent of net cled fiber in accordance with the recom- greenhouse gases mended standards for paper usage set by · 4 million BTUs of the Green Press Initiative, a nonprofit total energy program supporting publishers in using fiber that is not sourced from endangered forests. For more information, visit www.greenpressinitiative.org. G eographical differences in living standards are a pressing concern for policy makers in the Middle East and North Africa (MENA) region. Production is most efficient when concentrated in leading areas. So how can the region reduce spatial disparities in well- being without compromising growth? Poor Places, Thriving People addresses this compelling issue. The solution to spatial disparities lies in matching the policy package to a lagging area's specific characteristics. Is the lagging area problem really as serious as one thinks? Is it a problem of low economic opportunity or of poor human development? Are lagging area populations close enough to agglomerations to benefit from spillovers? Is there a realistic chance of private investment? Drawing on the World Bank's 2009 World Development Report: Reshaping Economic Geography, the authors propose three policy packages. First, all lagging areas can benefit from a level playing field for development and by investing in people. Geographic disparities in the policy environment are a legacy of MENA's history, and gaps in human development are a major component of spatial disparities. Smart policies for the investment environment, health, education, social transfers, and urban development can, therefore, close spatial gaps in living standards. Second, lagging areas that are connected to the poles of development can benefit from spillovers. MENA's expenditure priority is not necessarily long-distance primary connections, but maintaining infrastructure and short-distance connections such as rural roads and peri-urban networks. Public-private partnerships can also bring electronic connectivity to lagging areas. Third, policy makers can make the most of private investment in nonleading areas by shifting regional development policy away from wasteful subsidies towards the facilitation of cluster- based growth. Poor Places, Thriving People will be of interest to policy makers in government and in international organizations and individuals in research institutions and universities who work in economic policy, regional and rural policy, and social development. ISBN 978-0-8213-8321-6 90000 9 780821 383216 SKU 18321